Opportunities for Reducing Vegetative Ozone Exposure through U.S. Power Plant Carbon Standards

Shannon L. Capps, Charles T. Driscoll, Habibollah Fakhraei, Pamela H. Templer, Kathleen F. Lambert, Kenneth J. Craig, Stephen B. Reid

Shifts in fuels and technologies used to produce electricity will be required to meet carbon dioxide emissions standards for U.S. power plants and will simultaneously affect emissions of nitrogen oxides, which influence ambient ozone formation. We assess the potential ecosystem co-benefits through reduced ozone exposure of crops and trees in accordance with three alternative scenarios for U.S. power plant CO2 emissions standards. Estimates of emissions in 2020 from 2,417 fossil fuel-fired power plants were developed with the Integrated Planning Model. These emissions were used in Community Multiscale Air Quality (CMAQ) model to project impacts on ozone concentrations and species-specific potential productivity loss of crops and trees at 12-km x 12-km resolution. The potential productivity losses of soybean and cotton crops as well as eastern cottonwood and black cherry trees are non-negligible in the reference scenario. The greatest potential productivity gains for these species exist in the more stringent policies with the slightly greater reductions in crop and tree potential productivity losses achieved through a carbon tax-like mechanism.

Developments in the 2014 National Emissions Inventory

Rich Mason, Jennifer Snyder, Rhonda Thompson, Tesh Rao, Ron Ryan, Laurel Driver, Madeleine Strum, Sally Dombrowski

The 2014 NEI has been released to the public and includes numerous updates to many key sectors such as Oil and Gas, Residential Wood Combustion, Fires, Commercial Marine Vessels, Solvents, Industrial and Commercial/Institutional Fuel Combustion, Dust sources, and Agricultural Livestock and Fertilizer. We discuss the methodology updates, provide comparisons to the 2011 NEI, and mention new tools in our Emissions Inventory System that have enabled for more transparency and improved communication with reporting agencies during the NEI process. We will also mention issues that we will be focusing on for version 2 of the 2014 NEI and big picture ideas for the next triennial inventory in 2017.

Impacts of Technology-Driven Transportation Emissions on Future U.S. Air Quality in a Changing Climate

Patrick Campbell, Yang Zhang, Fang Yan, Zifeng Lu, and David Streets

Warming of the climate system is unequivocal, and emissions from the transportation sector are rapidly changing worldwide. In this study, we investigate the impact of advanced technology-driven changes in emissions from the transportation sector (i.e., on-road vehicles, non-road engines, aircraft, rail, and ship) on future U.S. air quality in the face of climate change. Simulations are performed over an area in North America using an offline-coupled Weather Research and Forecasting/Community Multi-scale Air Quality modeling system (WRF/CMAQ) for current-year and future-year periods of 2001 - 2005 and 2046 - 2050, respectively. The initial and boundary conditions are based on simulation results from a dynamically-downscaled climate model. Current-year emissions are based on the U.S. Environmental Protection Agency's 2005 National Emission Inventory Version 4, and the emission growth factors for the transportation sector are derived from a dynamic technology model, the Speciated Pollutant Emission Wizard (SPEW)-Trend under the IPCC A1B emission scenario. By 2046 - 2050, the annual domain-average transportation emissions of carbon monoxide (CO), nitrogen oxides (NO_x), volatile organic compounds (VOCs), ammonia (NH₃), and sulfur dioxide (SO₂) are projected to decrease over the continental U.S. The decreases in gaseous emissions are mainly due to reduced emissions from on-road vehicles and non-road engines, which exhibit spatial and seasonal variations across the U.S. Although particulate matter (PM) emissions widely decrease over the continental U.S., some areas across the domain experience relatively large increases due to increases in ship emissions. The on-road vehicle emissions dominate the emission changes for CO, NO_x, VOC, and NH₃, while emissions from both the on-road and non-road modes have strong contributions to PM and SO₂ emission changes. The evaluation of the baseline 2005 WRF simulation indicates that annual biases are close to or within the acceptable criteria for meteorological performance in the literature. There is an overall good agreement in the 2005 annual statistics and spatial distributions of chemical concentrations from CMAQ against both surface and satellite observations. There are domain-average decreases in future concentrations of CO, NO₂, and PM_2.5, which are dominated by decreases in transportation emissions. However, there are widespread increases in daily maximum 8-hr ozone (O₃) across the U.S., which are due to enhanced methane (CH₄) and carbon dioxide (CO₂) concentrations, and larger reduction in NO_x emissions compared to VOC emissions over regions with VOC-limited O₃ chemistry. We find that climate change leads to 1) increases in 2-meter temperature (T2) and biogenic VOC (BVOC) emissions that further enhance the O₃ levels over VOC-limited regions, 2) decreases in precipitation that increase PM_2.5 levels in spite of decreases in emissions of primary PM and precursor gases, and 3) increases in downward shortwave radiation, T2, and BVOC emissions that further enhance PM_2.5 levels in the southeast U.S. The results of this study imply that incorporation of dynamic technology-driven emissions into climate/air quality modeling has important policy implications on emission control to help attain, maintain, or improve future air quality.

Development of 2011 Hemispheric Emissions for CMAQ

A. Eyth, G. Pouliot, J. Vukovich, M. Strum, P. Dolwick, C. Allen, J. Beidler, B.H. Baek

Emissions for a year 2011 case on a hemispheric grid was prepared for CMAQ v5.1.1. The emissions included a 4 month spin-up and were processed through an updated version of SMOKE. The input emissions for the United States, Canada, and Mexico were based on EPA's 2011v6.2 platform emissions which rely on data from the 2011 National Emissions Inventory, Version 2. Fire emissions for the modeling period were derived from the FIre INventory from NCAR (FINN) data set. Methods were developed to process onroad mobile source emissions in Alaska, Hawaii, Puerto Rico and the Virgin Islands. Emissions for areas outside of North America were taken from the task force on the Hemispheric Transport of Air Pollution (HTAP) version 2 dataset. Double counting of emissions, including those from commercial marine vessels, between the two main emission inventory sources was prevented. Spatial surrogates were developed for the 108km resolution hemispheric grid. New features were built into SMOKE 4.0 to facilitate the development of the hemispheric emissions, including the ability to zero out and apply factors to emissions by country. Ancillary data files for speciation and temporal allocation outside of the United States were formatted for use in SMOKE 4.0. The computation of hourly emissions took into account time zones around the world.

Impact of the Biomass Burning Aerosols on the Regional Climate of the Southeastern U.S.

Peng Liu, Yongtao Hu, Athanasios Nenes, Armistead G. Russell

Biomass burning (BB), as one of the largest sources of trace gases and aerosols in the atmosphere, can play an important role in air quality, weather and climate by influencing the chemistry and radiative forcing. In the southeastern U.S., it has been demonstrated that biomass burning is a main source to atmospheric aerosols, especially during the main biomass burning season. Furthermore, emissions from biomass burning in the southeastern U.S. are expected to increase in the future. On one hand, wildfires are expected to increase due to climate change. On the other hand, prescribed burning will become more frequent and extensive due to land use change. However, few studies have placed a particular focus the potential impact of biomass burning aerosols on the climate of this region. This study, which serves as an attribution study, used the coupled WRF-CMAQ, to understand if aerosols from all emission sources do affect the current climate in the southeastern U.S, how much of the impacts can be attributed to aerosols from biomass burning.

To account for the uncertainty in assessing the climate impact of BB, the ensemble method, referred to as "short ensembles" (Wan et al., 2014), is applied, in which the continuous multi-year long climate simulation is replaced by the ensembles of short-term simulations, as long as the ensembles are able to represent the climate.

A variety of climatic fields that are important to regional climate are investigated, including radiation, temperature, regional circulations, clouds and precipitation. The results showed that the total aerosols do have a significant influence on the climate in the southeastern U.S. during the season investigated. Though biomass burning is an important contributor to the total aerosol loading in the SEUS, BB aerosols contribute little to the impact on regional climate from total aerosols in terms of the ensemble mean of the regional average. In other words, perturbation induced by biomass burning is in general too weak to develop into features at larger scale and finally affect the regional climate. In addition, though the wide variance occurs in the aerosol loadings from biomass burning, the investigated climatic fields are not sensitive to the wide change in biomass burning emissions.

Emissions Reconciliation Analyses in Californias South Coast Air Basin

Stephen Reid¹, Hilary Hafner¹, Michael McCarthy¹, Yuan Du¹, Timothy French²

¹Sonoma Technology, Inc., 1450 N. McDowell Blvd., Suite 200, Petaluma, CA 94954
²EMA, 333 West Wacker Drive, Suite 810, Chicago, IL 60606

Emissions inventories are important components of air quality planning and are key inputs for photochemical grid models that support air quality assessments. Several methods are available to evaluate and improve emissions estimates, including comparisons between emissions inventories and ambient monitoring data. These comparisons, which are often called "emissions reconciliation," are used to identify omissions or inaccuracies in an emissions inventory, leading to further investigation and inventory improvement. The basic approaches used to perform emissions reconciliation analyses include selective, quantitative comparisons of emissions inventory- and ambient-derived molar pollutant ratios (e.g., VOC/NO_x or CO/NO_x). This paper presents an evaluation of air quality model-ready emissions inputs developed to support the latest Air Quality Management Plan (AQMP) for California's South Coast Air Basin (SoCAB). The study includes an assessment of trends in ambient VOC and NO_x concentrations and VOC/NO_x ratios for several monitoring sites in the SoCAB during the years from 1995 through 2015. The study also includes an assessment of VOC/NO_x ratios derived from emissions data used as inputs to the Community Multiscale Air Quality (CMAQ) model for several simulations conducted as part of a dynamic evaluation of the AQMP modeling system. Study results feature comparisons between ambient- and emissions-derived data over several years (e.g., 2000, 2005, 2008, and 2012) and will provide insights into potential issues with VOC and NOx emissions estimates used for air quality planning in the SoCAB region.

Air Quality Impacts of Electrification in tandem with Intermittent Renewable Resources

Michael MacKinnon, Ph.D., Siavash Ebrahimi, Marc Carreras-Sospedra, Ph.D., Jack Brouwer, Ph.D., Donald Dabdub, Ph.D., G.S. Samuelsen, Ph.D.,

Renewable resources, including intermittent wind and solar power, are a key strategy to reduce the environmental impacts of power generation, including reducing emissions of greenhouse gases (GHG) and pollutants towards improving regional air quality (AQ). Furthermore, the concomitant electrification of additional energy sectors can replace traditional fuel combustion with low carbon and efficient electric technologies. With this premise, electrification strategies could result in net reductions in GHG emissions, particularly as electrical grids become integrated with higher levels of renewable power. Similarly reductions in pollutant emissions could achieve AQ benefits by reducing atmospheric levels of primary and secondary pollutants, including ozone and fine particulate matter (PM2.5). On the hand, electrification of reduce emissions but may also increase power sector emissions from new electrical demand or have grid dynamic impacts increasing emissions (e.g., ramping, cycling, start/stop). The result is the potential for both increases and decreases in emissions and atmospheric pollutant concentrations. Furthermore, assessing AQ impacts is not as simple as quantifying total emissions. The complexity of pollutant formation and fate requires an understanding be gained of how perturbations in pollutant emissions from both the power and any electrified sectors yield changes in spatial and temporal emissions, and how these are converted by atmospheric chemistry and transport into ground-level pollutant concentrations. Thus, there is a need for more information regarding how renewable resources in tandem with increased electrification of energy sectors could impact emissions and AQ. The goal of this work is to quantify the impacts on primary and secondary air pollutants and GHG from the electrification of energy sectors in tandem with high levels of renewable resources in California. A set of scenarios are analyzed, developing spatially and temporally resolved emissions and simulating the resulting AQ accounting for the electrification of various energy sectors in tandem with renewable resource integration in the power sector. The resulting output is assessed for ground-level concentrations of ozone and PM2.5 to determine overall AQ impacts. Cases are assessed for both a summer and winter episode in order to capture high ozone (summer) and high PM (winter) episodes. The results show that electrification will largely result in AQ improvements, but could yield areas of worsening represented by localized increases in ozone and PM2.5 resulting from increased power sector emissions from new electrical demand and impacts of grid dynamics. Furthermore, impacts depend on multiple factors and vary markedly by pollutant, sector, horizon year, season, and location. The results have implications for both renewable resource deployment and regional AQ improvement planning.

Incorporate Traffic Demand Model Data in SMOKE-MOVES Processing for Denver Ozone Modeling

Tejas Shah, Yesica Alvarez, Michele Jimenez, and Ralph Morris

Ramboll Environ, Novato, CA

Amanda Brimmer and Ken Lloyd

Denver Regional Air Quality Council (RAQC)

The Denver Regional Air Quality Council (RAQC) is in the process of developing an ozone State Implementation Plan (SIP) to address attainment of the March 2008 0.075 ppm ozone National Ambient Air Quality Standard (NAAQS). Photochemical Grid Model (PGM) modeling was used to demonstrate the Denver Metropolitan Area (Denver Metro) and North Front Range (NFR) ozone nonattainment area (NAA) will attain the ozone NAAQS by 2017. On-road mobile source emissions is one of, if not the, most important source of ozone precursors in the Denver Metro/NFR NAA. Thus, providing an accurate representation of mobile source emission accounting for spatial and temporal variations in fleet mix, starts, reactivity and other factors is important for producing an accurate and reliable attainment dmeonstartion. The purpose of this paper is to demonstrate how detailed link-based Traffic Demand Model (TDM) data can be used in SMOKE-MOVES processing to generate day-specific hourly gridded speciated on-road mobile source emissions for PGM modeling.

SMOKE-MOVES is typically applied using county-level VMT, vehicle population and speeds that are allocated to grid cells using an appropriate spatial surrogate (e.g., roadway locations) and then applies MOVES emission factors from a lookup table using hourly gridded meteorological data (temperature and humidity). SMOKE-MOVES can not directly use link-based TDM activity data as input. In this project, we developed an approach to use link-based data from the TDM models as input to the SMOKE-MOVES processing. We treated each grid cell and speed class as a pseudo-county in the processing and developed spatial surrogate that was one-to-one mapping of a pseudo-county to respective grid cell. The off-network emissions were spatially allocated using surrogates developed from trip starts (start exhaust) and trip ends (for evaporative processes) by Traffic Analysis Zone (TAZ). We developed a diurnal temporal profile for each pseudo-county (i.e. grid cell intersecting the network) using the TDM data and apply them on a pseudo-county basis in the temporal processing. This approach takes advantage of detailed link-based spatial and temporal varying activity data available from the TDM model and the robust emissions calculation methodology of SMOKE-MOVES that takes into account temporal and spatial variations in meteorology and has EPA approval and support.

Air Pollution Externalities and Energy Choices: Linking Electricity Dispatch, Air Quality and Health Impact Models

Michael D. Moeller¹, Frank A. Felder²,Kirk Baker³, Annmarie G. Carlton¹

Energy consumption is the largest controllable source of emissions that negatively impact air quality and subsequently human health. Electric generating units (EGUs), a major subset of energy consumption, contribute the most to primary PM_2.5 in the northeast and sulfate mass in the continental U.S. The associated societal health costs are large. Emissions from fossil-fueled EGUs alone are estimated to cause more than 38,200 heart attacks and 554,000 asthma attacks per year in the United States with direct annual health costs of $167 billion. Though substantial, the air-pollution related health costs are not completely accounted for in conventional energy planning and operations, but rather occur as non-priced market externalities subsidized by the public. Electricity dispatch is optimized across the power grid based upon minimizing costs subject to reliability and demand-supply constraints. In this study we link day-ahead and hourly electricity unit and commitment dispatch from a regional power market forecast model (DAYZER), used by electricity planners and day traders, with air quality modeling from the Community Multiscale Air Quality Model (CMAQ). We evaluate an alternative energy scenario currently pending before the Bureau of Ocean Energy Management (BOEM). The proposed "Atlantic Wind Connection" is an off-shore wind farm integrated into the Mid-Atlantic power grid. Together these models generate realistic spatial and temporal air quality impacts from the inclusion of wind-farm electricity capacity relative to the existing power grid. Analysis by the Environmental Benefits Mapping and Analysis Program (BenMAP) model assesses health incidence impacts and associated societal health costs. Results suggest 20-60 avoided pre-mature mortalities annually in the effected Mid-Atlantic and Midwest regions with an overall estimated cost savings range of $230 - $520 million per year. Uncertainties in health cost savings are explored in terms of payback period (i.e., the time at which failure to construct the wind farm becomes more expensive than construction, operation and maintenance). Coupled modeling systems, such as those employed here, that reflect actual electricity markets and link to their environmental impacts can inform policymakers and enable design of a more sustainable U.S. energy system that reduces peak concentrations of multiple pollutants to protect human health.

Development of an Emission Uncertainty Inventory and Modeling Framework: Case Study of Residential Wood Combustion

Rabab Mashayekhi¹, Shunliu Zhao¹, Sahar Saeednooran¹, Amir Hakami¹, Richard Menard², Michael D. Moran², and Junhua Zhang²

¹ Department of Civil and Environmental Engineering, Carleton University, Ottawa

² Air Quality Research Division, Environment and Climate Change Canada

Model-ready emissions are often developed without formal quantification of uncertainties associated with emission inventory development and emissions processing. This lack of uncertainty characterization has important implications in downstream applications of model results, where costly decisions are made based on models' "best estimates". In this study as a proof-of-concept example we have accounted for the uncertainties in the underlying emission data from residential wood combustion (RWC) in the U.S. and Canada and developed a framework that allows for propagation of uncertainties in various stages leading to the generation of emissions through the SMOKE (Sparse Matrix Operator Kernel Emissions) emissions processing system. An effort has been made to identify all important underlying raw information, including emission factor and activity data that contribute to the estimation of total annual RWC emissions. The uncertainty for each individual inventory parameter is characterized based on existing information sources, including the American Housing Survey (AHS) from the U.S. Census Bureau, Timber Products Output (TPO) surveys from the U.S. Forest Service, the Canadian Facts survey, and the AP-42 emission factor document from the U.S. EPA. The propagation of uncertainties is based on the Monte Carlo simulation code external to SMOKE. Latin Hypercube Sampling (LHS) is implemented to generate a set of random realizations of each parameter, for which the uncertainty is assumed to be a normally distributed random variable. The uncertainties associated with three emissions processing steps temporal allocation, spatial allocation, and speciation are also considered. Random realizations are obtained for each temporal and speciation profile and spatial surrogate field external to SMOKE by assigning a standard deviation to each single coefficient for each profile followed by generation of realizations using the LHS approach. SMOKE output for primary emissions (e.g., CO) with emission inventory uncertainty accounted for shows a relative uncertainty of about 30-50% across the domain with the exception of a few locations where the raw data (especially wood density) are very uncertain. Spatial allocation contributes significantly to the overall uncertainty (particularly in Canada) as the relative uncertainty approaches 80% over areas where low RWC activities are reported. By applying this framework we can produce random realizations of model-ready gridded emissions that along with available meteorological ensembles can be used to propagate uncertainties through CTMs. The approach developed here provides an effective means for formal quantification of uncertainties in estimated emissions from various source sectors and for continuous documentation, assessment, and reduction of emission uncertainties.

Quantifying the effect of natural variability on the assessment of climate policies' health benefits compared to costs
Rebecca K. Saari, Yufei Mei, Erwan Monier, Fernando Garcia-Menendez

We quantify the effect of natural variability on the health benefits of climate policy. Natural variability can significantly influence climate predictions; however, most studies that simulate the influence of the future climate on air quality use 1-5 year simulations that are unable to capture natural variability, and they have not examined its effect on the health impacts of climate change or climate policy. Here, we employ 30-year simulations around 2050 and 2100 and initial condition ensembles to account for inter-annual and multi-decadal variability in ozone and fine particulate matter. For each year, we use 5000 Monte Carlo simulations to estimate uncertainty in health outcomes and valuations from two climate stabilization policies. Self-consistent estimates of policy costs and health benefits are derived using the MIT Integrated Global System Model (IGSM) framework to develop integrated economic and climate projections. These projections are used to drive the global chemical transport model CAM-Chem and obtain ambient concentrations for the health impacts model BenMAP-CE. We find that uncertainty from natural variability and health impacts yield a range of annual health benefits from climate policy between -$0.5 and $1 trillion by mid-century, representing -30% to 30% of annual policy costs. We compare the components of uncertainty that contribute to this range in 2050 and 2100 to inform best practices for estimating the health benefits of climate policy.

SPATIAL DISTRIBUTION OF PARTICULATE MATTER EMISSION FROM RESIDENTIAL COMBUSTION IN LATIN AMERICA, AFRICA, AND ASIA

Ekbordin Winijkul, Tami C. Bond

This study describes a framework for attributing national-level particulate matter emissions from residential combustion to 5km x 5km gridded cells and examines the effects of different pollution mitigation policies on these emissions. This study covers regions where solid biomass fuel provides more than half of total residential energy: Latin America, Africa, and Asia. First, national-level residential energy usage is divided into cooking, heating, lighting, and other uses. Second, five land types are classified using global nightlight, population density, and land cover maps: urban; electrified rural with forest access; electrified rural without forest access; non-electrified rural with forest access; and non-electrified rural without forest access. Third, energy consumption is apportioned among all land-types and end-uses. Finally, appropriate stove technologies and emission factors are assigned to each combination of end-use and land-type. The final product is a spatially-distributed map of particulate matter emissions.

The effects of pollution mitigation policies on the particulate matter emissions from residential combustion are examined considering realistic factors that constrain mitigation, including the intended end-uses, spatial constraints involving proximity to forest or electricity, existing technology, and assumptions about user behavior. We explore three mitigation scenarios: (1) use of the cleanest current stove, where existing cleanest technology in each land type is applied to burn existing fuels; (2) use of stove standards, where all available stoves are designed to meet performance standards; and (3) use of clean fuels, where users adopt the cleanest fuels plausible in each land type.

Modeled Source Apportionment of Reactive Nitrogen in the Greater Yellowstone Area
Tammy Thompson

Coming soon

A combined line-point-source model for ship emissions in the port of Hamburg, Germany

Armin Aulinger, Volker Matthias, Johannes Bieser

Pollutant exhaust from ships is an important factor influencing the
air quality in cities with large ports like the city of Hamburg,
Germany. Therefore, municipalities are concerned with measures aimed
at mitigating air pollution by ships, in particular with NOx and
PM. For planning these measures and evaluating their effectiveness, it
is essential to have an emission model that can be run in any temporal
and spacial resolution, and above all, is able to simulate different
emission scenarios.

At Helmholtz-Zentrum Geesthacht we developed a model that simulates
the activities sailing, maneuvering and berthing of every single
registered ship in the port of Hamburg and calculates their emissions
of amongst others NOx, SO2 and elemental carbon. The vessels'
activities can be derived either from Automatic Identification System
(AIS) signals or from arrival-departure data. This means that the
least that must be known is time of entry into the port, arrival at
the quay, departure from the quay and time of leaving the
port. Emissions during sailing and maneuvering are represented as line
sources while berthing emissions are point sources. These emissions
are then gridded using a freely selectable grid definition and stored
in an user defined temporal resolution. Because emissions are
calculated for every single ship before aggregation, factors that
represent abatement strategies or even emission increases can be
applied to single ships, ships of certain types or of certain
sizes. This makes the model a highly flexible tool for creating
emission scenarios.

Currently, the output format of the data follows the M3IOAPI standard
so they can be directly used in SMOKE or CMAQ. The model is written
and documented in a way that makes it easily adaptable to other ports.

Quantifying co-benefits of CO2 emission reductions in Canada and the US: An adjoint sensitivity analysis

Marjan Soltan zadeh, Amanda Pappin, Shunliu Zhao, Amir Hakami (Carleton University); Matt D. Turner, Shannon L. Capps, and Daven K. Henze (University of Colorado); Peter B. Percell (University of Houston);Jaroslav Resler (ICS Prague); Jesse O. Bash, Kathleen Fahey,Sergey L. Napelenok (USEPA); Rob W. Pinder; Armistead G. Russell and Athanasios Nenes (Georgia Tech); Jaemeen Baek, Greg R. Carmichael, and Charlie O. Stanier (University of Iowa); Adrian Sandu (Virginia Tech); Tianfeng Chai (University of Maryland); Daewon Byun (NOAA)

The topic of air quality co-benefits of Greenhouse Gas (GHG) reductions has been the subject of many scenario-based research studies. Scenario-based (SB) studies evaluate air quality co-benefits by adopting collective measures introduced in a climate policy scenario. Being sector-specific, i.e., SB analysis considers emissions changes in specific sectors and across all locations; however, SB studies cannot distinguish between benefits accrued from GHG reductions among sources in different locations. Location dependence, also known as source-specificity, is an important factor that can be captured in an adjoint-based analysis. The present study aims to quantify how the ancillary benefits of reducing criteria co-pollutants vary spatially and by sector. It also aims to compare their spatial distribution with that of abatement costs or the social cost of carbon.

In an earlier study we applied the adjoint of CMAQ to quantify the health benefits associated with emission reduction of criteria pollutants (NO_X) in a limited (mobile on-road) sector on a location-by-location basis across the US. These health benefits are then converted to CO₂ emission reduction co-benefits by accounting for source-specific emission rates of criteria pollutants in comparison to CO₂. We further expand the scope of our study by including Canada and by considering co-benefits resulting from reduced PM_2.5 emissions. PM_2.5 marginal benefit is defined as the monetized health benefit of reducing primary PM_2.5 components including organic carbon (POC), black carbon (PEC), sulfate (PSO₄), and nitrate (PNO₃) from a given source. We integrate the results from the adjoint of CMAQ with emission estimates from 2011 National Emission Inventory (NEI 2011) at the county level (aggregated up to 36-km resolution), point source data from EPA's Air Markets Program Data (AMPD 2011), National Pollutant Release Inventory (NPRI), and GHGenius sectoral estimates for Canada. In our analysis we consider a number of sectors including transportation, EGUs, and other major point source sectors in Canada and the US.

Our preliminary results show that the monetized health benefits associated with reductions in 1 ton of CO₂ emissions is up to $65/ton CO₂ for mobile gasoline light duty, and $55/ton CO₂ for mobile diesel heavy duty in Canada. For mobile emissions the calculated co-benefits through gaseous pollutants including NO_x is larger than the calculated co-benefits through PM_2.5 due to smaller emission factors for primary PM emissions from these sectors. Calculated co-benefits show a great deal of spatial variability across emission locations for different sectors and sub-sectors. Implications of such spatial variability in devising control policy options that effectively address both climate and air quality objectives will be discussed.

Improving Air Quality Modeling Performance and Capabilities in Bogot, Colombia

Pachon, Jorge. Associate Professor Universidad de La Salle, Bogota, Colombia

Galvis, Boris. Associate Professor Universidad de La Salle, Bogota, Colombia

P rez, Maria Paula. Research Assistant Universidad de La Salle, Bogota, Colombia

Ramirez, Jhonathan. Research Assistant Universidad de La Salle, Bogota, Colombia

Castro, Luisa. Research Assistant Universidad de La Salle, Bogota, Colombia

Palacios, Cesar. Research Assistant Universidad de La Salle, Bogota, Colombia

Henderson, Barron. Assistant Professor University of Florida, Gainesville, FL, USA.

Nedbor-Gross, Robert. Graduate Research Assistant, University of Florida, Gainesville, FL, USA.

In 2012 the Bogota s Environmental Agency (SDA) along with University of La Salle and University of Florida implemented a CMAQ based air quality (AQ) modeling system. The modeling system puropose was to understand the fate and transport of atmospheric pollutants in the city and assess AQ improvement strategies. This first attempt to implement the model showed a general good agreement between model estimates and observations with a relative overestimation of PM10, CO, SO₂, O₃ and NO_x concentrations at specific areas of the city. The city government is currently working to improve AQ, devising plans and applying strategies for which Bogota's AQ model has been of great usefulness. In the last two years, our efforts have been focused on improving modeling results and capabilities, updating emissions inventories to 2014, improving estimates of resuspended particle matter from paved and unpaved roads by considering mitigation factors (precipitation, vegetation, transport), testing different methodologies to spatially interpolate traffic and dust loadings, adjusting mobile source emission factors to consider fleet turnover and changes in fuel characteristics, using different terrain databases to improve meteorological performance, and automating emission processing, among others. These activities have been accompanied by AQ modeling training to recent graduated professionals strengthening modeling capabilities in the country. This work shows the main results of our AQ modelling efforts in the last two years in Bogota, experience that can be of potential interests for other cities in developing countries.

Estimating source attribution from oil and gas extraction on nitrogen deposition at western national parks using CAMx-PSAT

Michael Barna, National Park Service

Tammy Thompson, Colorado State University

Tom Moore, Western States Air Resources Council/Western Regional Air Partnership

Todd McDonnell, E&S Environmental Chemistry

Tim Sullivan, E&S Environmental Chemistry

Over the last ten years, oil and natural gas extraction has rapidly increased in the Intermountain West states of Wyoming, Colorado, Utah and New Mexico.  These operations often occur near sensitive wilderness areas and national parks.  Ecosystems within these areas are often near or above a "nutrient nitrogen critical load", meaning that additional nitrogen deposition may foster unwanted changes in plant communities.  This study uses the 2011 WAQS (Western Air Quality Study) CAMx (Comprehensive Air Quality Model with Extensions) platform to simulate nitrogen deposition at several western National Parks, and to determine the contribution of emissions from oil and gas sources to this deposition estimate.  A fine scale (4km) grid was employed to assess within-park deposition gradients.  A detailed emission inventory was developed for the WAQS, including a comprehensive survey of the oil and gas development sector. CAMx predicts both wet and dry deposition of a full suite of nitrogen-containing gases and particles, including ammonia (NH3), nitric acid (HNO3), nitrogen oxides (NOx), peroxyacetyl nitrates (PAN), and particulate ammonium (NH4+) and nitrate (NO3-).  National Park Units that were significantly affected with regard to nitrogen deposition from oil and gas sources include Mesa Verde National Park in southwestern Colorado (0.61 kg N/ha/yr, 19% of total deposition), Dinosaur National Monument in northeastern Utah/northwestern Colorado (0.58 kg N/ha/yr, 20% of total deposition), Hovenweep National Monument in southeastern Utah/southwestern Colorado (0.43 kg N/ha/yr, 14% of total deposition), and Rocky Mountain National Park in northern Colorado (0.26 kg N/ha/yr, 8% of total deposition).  The bulk of the impact is in the form of oxidized nitrogen deposition, suggesting that NOx controls for oil and gas development equipment (e.g., drill rigs, compressor engines) would be most effective for reducing harmful downwind impacts at National Parks.

Impacts of climate change on photochemical pollutants and allergenic pollen in the United States

Ting Cai^1,3, Allison P. Patton¹, Yong Zhang^1,2, Zhongyuan Mi^1,3, Panos Georgopoulos^1,2,3,4

¹Environmental and Occupational Health Sciences Institute (EOHSI), Rutgers University, Piscataway, NJ 08854, USA

²Department of Chemical and Biochemical Engineering, Rutgers University, Piscataway, NJ 08854, USA

³Department of Environmental Sciences, Rutgers University, New Brunswick, NJ 08901, USA

4Department of Environmental and Occupational Health, RBHS-SPH, Piscataway, NJ 08854, USA

The prevalence of Allergic Airway Disease (AAD) has grown globally resulting in increased numbers of emergency department visits and hospitalizations. Clinical studies show that AAD can be exacerbated by the synergistic action of bioaerosols such as pollen and fungi, and atmospheric pollutants such as ozone. Furthermore, climate change has critically affected atmospheric processes involved in the dynamics of air pollution systems and emissions of natural pollutants such as pollen and spores. Previous studies, involving data from nationwide observations of airborne pollen counts of selected plant species in conjunction with climatic factors, indicated that the start date and length of pollen season, the average peak value and annual total of daily counted airborne pollen have been affected substantially by the changing climate. The present study investigates the co-occurrences of peak ozone concentrations and peak pollen counts across the Contiguous United states (ConUS). The concentrations of pollen and ozone in future years (2047-2050) are simulated with an adapted version of CMAQ (The Community Multiscale Air Quality) and compared with "base" years concentrations (2001-2004) employing a grid with 50 km by 50 km horizontal resolution. The meteorological conditions for the simulations of future years are modeled with WRF (Weather Research and Forecasting model) based on IPCC projections. Spatiotemporal correlation analysis is employed to examine relationships between ozone and pollen concentrations. The outcomes of this study would be useful in supporting development of strategies for managing health-impacts of co-occurring  photochemical pollutants and allergenic pollen.

Concentrations of individual fine particulate matter components in the United States around the 4th of July
Elizabeth A. W. Chan, Adam Benson, Barbara Buckley, and Aisha Dickerson

Fireworks emit particulate matter (PM) air pollution. Laboratory and epidemiologic studies have linked exposure to PM with cardiovascular and respiratory effects. Although it was recently reported that the mass of PM with a nominal mean aerodynamic diameter less than or equal to 2.5 μm (PM2.5) is elevated on July 4th and 5th, no studies to date have used national, multi-year air quality monitoring data to determine which individual PM2.5 components increase due to July 4th fireworks. To evaluate this, we compiled and analyzed daily average PM2.5 air quality data collected by Environmental Protection Agency's Chemical Speciation Network monitors positioned at 379 urban sites across the United States (US) over the years 2000 to 2014. By combining all individual daily mean PM2.5 concentrations recorded and viewing the arithmetic mean concentrations over time, we observed sharp and statistically significant increases in the concentrations of the firework-related chemicals barium (Ba), copper (Cu), chlorine (Cl), magnesium (Mg), potassium (K), and strontium (Sr) on July 4th, which persisted through July 5th. There were also small, but statistically significant, increases of the firework-related components aluminum (Al), arsenic (As), antimony (Sb), chromium (Cr), phosphorous (P), sulfur (S), titanium (Ti), and zinc (Zn), but not of elemental or organic carbon (EC or OC), calcium (Ca), cesium (Cs), iron (Fe), nickel (Ni), or sodium (Na) on July 4th.

Association of trends in US ambient air quality and cardiovascular mortality for 2000-2010

Anne E Corrigan, Michelle Becker, Lucas Neas, and Ana Rappold

With the implementation of the Clean Air Act's National Ambient Air Quality Standards, air quality in the United States has notably improved.  Here we investigate whether declining levels of air pollutants are associated with improvements in human health.  We examine the relationship between the long term change in fine particulate matter (PM_2.5) and cause-specific mortality rates, adjusted for age, across the United States.  Data was compiled from the EPA's Air Quality System, National Center for Health Statistics, and Census Bureau for PM_2.5, mortality, and covariate information, respectively, for 613 U.S. counties with air quality monitoring from 2000 to 2010.  For all analyses, linear mixed-effects models were employed with fixed effects for changes in income, race, Hispanic status, education, and prevalence of cigarette smoking.  Our preliminary findings indicate a strong association between improvements in air quality and improvements in health, with the strongest associations related to cardiovascular deaths.  With random slopes and intercepts, we also examine the heterogeneity of these effects with respect to geographic regions and the impact of regulatory action. 

This abstract of a proposed presentation does not necessarily reflect the policies of the U.S. Environmental Protection Agency.

Exposure to Fine Particulate, Black Carbon, and Particle Number Concentration in Transportation Modes in Bogota

B. Galvis. Department of Environmental and Sanitary Engineering, Universidad de la Salle, Bogota, Colombia

R. Morales Betancourt. Department of Civil and Environmental Engineering, Universidad de Los Andes, Bogota Colombia

S. Balachandran. College of Engineering & Applied Science, University of Cincinnati, Cincinnati, OH

We determined concentrations of aerosols to which commuters are exposed in Bogota, Colombia. We measured fine particulate, equivalent black carbon, and number of sub-micron particles concurrently on different roads and modes of transportation using portable instruments. We selected three roads because of their dissimilar traffic composition, loads and street geometry. We performed measurements in the available modes at each road segment, including two active modes (walking and cycling) and three motorized modes (bus, car and taxi). We also ran tests during two bike-or-walk only street events. We found the lowest exposures for pedestrians and cyclists at an open-street with an exclusive bike lane built in the sidewalk. We found the highest average concentrations of fine particulate, equivalent black carbon, and sub-micrometer particle number concentration, inside Bogota's BRT system buses. These concentrations were up to 6.5 times greater than those for pedestrians and cyclists in the same street. We measured equivalent black carbon concentrations that account for 50% to 70% of the fine particulate across all modes. We observed elevated number concentration of sub-micrometer particles, with average concentration ranging from 60 x 10³ cm³ for pedestrians in the open street configuration, to nearly 200 x 10³ cm³ inside the BRT buses. The presence of dedicated bike lanes showed an impact on reducing the exposure concentration for bike users and pedestrians. The dose inhaled by commuters in motorized vehicles during the time spent inside the vehicle was higher than for those using active modes, the cyclist being the lowest.

Studying Aerosol Indirect Effects on Grid and Subgrid Scale Clouds using the two-way Coupled WRF-CMAQ

Jian He, Kiran Alapaty, Timothy Glotfelty, Xiaoliang Song, Guang Zhang, Shaocai Yu, and Daiwen Kang

Air pollution is one of the environmental process impacting climate which in turn impacts air pollution. Thus, it is one of the complex and uncertain aspect of climate change. As per the Intergovernmental Panel on Climate Change Fifth Assessment report (IPCC AR5), in particular, aerosols indirect effects on clouds, especially subgrid-scale clouds, are one of the highly uncertain forcings in air quality and climate modeling studies. In many climate models, aerosol indirect effects on subgrid scale clouds are not treated, increasing the uncertainties in climate predictions and consequent effects on air pollution. An advanced treatment of aerosol indirect effects on grid scale clouds have been previously implemented into the two-way coupled WRF-CMAQ. In this work, an explicit treatment for aerosol indirect effects on subgrid scale clouds are implemented into the two-way coupled WRF-CMAQ.

To estimate aerosol indirect effects on clouds, two types of simulations are conducted in this work: (1) using actual aerosol concentrations predicted by CMAQ; and (2) using background aerosol concentrations (natural aerosol concentrations + 1/10 of the actual anthropogenic aerosol concentrations predicted by CMAQ). The differences between the two simulations can represent the impacts of anthropogenic aerosols on clouds. Numerical simulations are conducted at a horizontal resolution of 108 km and 40 layers over Northern Hemisphere for the summer (June, July, and August) of 2006. Our preliminary results indicate that high aerosol concentrations reduce the subgrid scale convective precipitation by a domain average of 25 mm with significant decreases over tropics (e.g., > 80mm). Grid scale cloud liquid water path (LWP) and ice water path (IWP) are increased by a domain average of 0.7 and 0.4 g m^-2, respectively, whereas subgrid scale cloud LWP and IWP are increased by 96.7 and 1.7 g m^-2, respectively. As a result, shortwave cloud radiative forcing is increased by a domain average of 3.3 W m^-2 and downward surface solar radiation is decreased by a domain average of 4.0 W m^-2. These results demonstrate the significant impacts of anthropogenic aerosols on clouds as well as radiation predictions contributing to climate change.

Development of the GAINS-Korea for Integrated Assessment of Greenhouse gas Air pollutant Management in Korea
Younha Kim 1, Jung-Hun Woo 1,2, Ki-Chul Choi 1, Jinsu Kim 2 , Jinseok Kim 2 , Chanjong Bu 1 , Yungu Lee 1, Young-Hwan Ahn 3, Sangkyun Kim 4
1 Department of Advanced Technology Fusion (DATF), Konkuk University, 120 Neungdong-ro, Gwangjin-gu, Seoul 05029, Korea
2 Division of Interdisciplinary Studies, Konkuk University, 120 Neungdong-ro, Gwangjin-gu, Seoul 05029, Korea
3 Korea Energy Economic Institute, Ulsan, Korea
4 National Institute of Environmental Research (NIER), Inchen, Korea

We have developed the GAINS-Korea, as an integrated modeling framework for greenhouse gas-air pollution management for Korea. The GAINS-Korea was designed to have 17 regions, 100+ source sectors, and 10 pollutants. For the emissions module, we developed the mapping algorithm between CAPSS and GAINS sector-activity structures that includes emission activities, emission factors, penetrations, and etc. The policy scenarios and control technologies for South Korea are also included in the GAINS-Korea model. The CAMx, as a regional-scale chemistry transport model, was used to develop the source receptor (S-R) matrix for both of ozone and particulate matter concentrations that describe the response of a range of air quality indicators to changes in the emissions of the various pollutants in each of the source regions. For each developed modules including impact assessment are now integrated into GAINS-Korea, so that it covers scenario analysis for S. Korea. Based on the GAINS-Korea model, we also developed several representative scenarios for S. Korea. For the baseline, the BAU (Business as usual) scenario was developed until the year 2030. The air pollutant reduction scenario to support air quality plan and the scenario under GHGs reduction plan for S. Korea were also created in the GAINS-Korea model. To investigate the effect of Co-control for both air pollutants and GHGs, we tried to create the co-control scenario and analysis the co-benefit through the GAINS-Korea model. We will present integrated impact analysis from emission pathways to health impact/cost analysis for future scenario of South Korea.

Acknowledgement : This subject is supported by Korea Ministry of Environment as "Climate Change Correspondence Program" and is financially supported by Korea Ministry of Environment(MOE) as "Graduate School specialized in Climate Change"

Real-Time Air Quality Forecasting over Southeastern United States using Updated Emissions and Satellite-Constrained Boundary Conditions

Qi Li, Chinmay Jena, and Yang Zhang

Three-dimensional air quality model is a powerful tool for real-time air quality forecasting (RT-AQF). Accurate emissions and boundary conditions (BCONs) can reduce uncertainties associated with model inputs thus potentially improve RT-AQF skill. The online-coupled Weather Research and Forecasting model with Chemistry with the Model of Aerosol Dynamics, Reaction, Ionization, and Dissolution (WRF/Chem-MADRID) has been applied for RT-AQF over southeastern U.S. since May 2009. Previous evaluations of RT-AQF during 2009-2015 showed consistently good skills for ozone (O₃) and fine particles (PM_2.5), but there are increased overpredictions in forecasted PM_2.5 over years. In addition, for all years large biases are found for some secondary PM_2.5 species against some surface observations and for column mass abundance of trace gases against satellite retrievals. Those biases are mainly attributed to out-of-date emissions and inaccurate BCONs. In this work, WRF/Chem-MADRID with updated emissions and adjusted BCONs is applied for RT-AQF for O₃ season (May to September) in 2016.  The updated emissions are based on the 2011 national emission inventory (NEI). The adjusted BCONs are based on satellite-constrained values for CO, SO₂, NO₂, O₃, and HCHO. By using observations from available surface networks and satellites, a comprehensive evaluation is performed, including domain average discrete and categorical performance statistics, spatial and temporal distributions, and column mass abundances. The results will be compared with those without updates during 2009-2015. Two sensitivity simulations are conducted in July 2016 to test the model sensitivity to these two updates in model input: (1) using the old emission while keeping the updated satellite-constrained BCONs, and (2) using the old BCONs while keeping updated emissions. Preliminary results show that the forecast skill for PM_2.5 has been improved by using updated emissions and satellite-constrained BCONs. Compared to the results in May 2015, the mean bias for PM_2.5 forecast in May 2016 decreases from -1.3 to -0.6 g m^-3. Large biases remain for some variables. For example, the normalized mean biases for precipitation in May are over 73%. These biases may be due to uncertainties in emissions, inaccurate meteorological predictions, missing and/or inaccurate model treatments (e.g., cloud processes and cloud-radiation feedbacks). The results from this work illustrate the benefit of improved model inputs and help identify future areas of improvement for the RT-AQF model. 

Effects of aerosol feedback on aircraft-attributable surface O3 and PM2.5 concentrations using the two-way coupled WRF-CMAQ modeling system

Chowdhury Moniruzzaman, Jared Bowden and Saravanan Arunachalam

Aircraft's landing and take-off (LTO) emissions contribute to poor local air quality and atmosphere's radiative forcing (RF) imbalance. Aircraft emitted particulate matter (PM) scatters and absorbs radiation from the sun which changes temperature, wind speed, relative humidity and planetary boundary layer (PBL) height in the atmosphere. The changes also affect atmospheric chemistry which causes changes in PM concentrations. This change in PM concentration in the atmosphere in turn affects meteorology. The process is called aerosol feedback, and such real-world feedback effects are neglected in traditional air quality models (where meteorology is used as input and not affected by chemistry). A two-way coupled meteorology-chemistry model can simulate this aerosol feedback effect on concentrations of surface layer ozone (O3) and PM having size less than 2.5 microns (PM2.5). In this study, a coupled meteorology-chemistry model WRF-CMAQ was used to determine the impacts of aircraft's LTO emission's contribution to surface layer O3 and PM2.5 concentrations as well as surface layer temperature for continental USA at a 36x36-km horizontal resolution. Aircraft's LTO emission contribution to surface layer ozone O3 and PM2.5 concentrations were determined for both 1) without aerosol feedback (similar to traditional air quality modeling where meteorology is used as input) and 2) with aerosol feedback (where both meteorology and chemistry affect each other). The aircraft emissions inputs were taken from a global chorded inventory of aircraft activity modeled by the FAA's Aviation Environmental Design Tool (AEDT). Preliminary results from our model simulations are that when aerosol feedback effect was considered, the LTO emission contribution to total concentration of O3 and PM2.5 are 48% and 55% lower respectively in January 2005 than those when feedback effect was not considered. The model simulation also shows that aircraft LTO emissions contribute to 0.0004 K domain average temperature increase in January 2005 in continental USA. We are performing longer-term simulations for an annual period, and results from this annual simulation will be presented. Using coupled meteorology atmospheric model such as WRF-CMAQ may give improved air quality prediction for better understanding the atmospheric impacts of this key emissions sector.

Estimating Environmental Co-benefits of U.S. GHG Reduction Pathways Using the GCAM-USA Integrated Assessment Model

Yang Ou, Wenjing Shi, Dan Loughlin, Chris Nolte, Steven J. Smith, Catherine Ledna, Jason West

Previous studies have shown that mitigating climate change through curbing greenhouse gas (GHG) emissions can bring about substantial environmental co-benefits, such as for air quality and reductions in energy-related water demand. A variety of pathways are available for reducing GHG emissions, however, including a transition to low-carbon fuels, carbon capture technologies, renewable energy, and energy efficiency and conservation. These pathways can have very different environmental co-benefits. The cost and reliability of energy may also differ. Development of a sustainable climate mitigation strategy thus benefits from simultaneous consideration of climate change, environmental, and energy objectives.

 

Integrated Assessment Models (IAMs) have the potential to support coordinated climate, environmental, and energy decision making. Here we illustrate how the Global Change Assessment Model-USA (GCAM-USA) - an IAM with state-level resolution for the U.S. - can be applied to explore the environmental co-benefits of alternative GHG reduction pathways. We compare an idealized mitigation pathway that focuses on nuclear energy and carbon capture with another that focuses on renewable energy to 2050. Environmental metrics such as emissions of multiple air pollutants and energy-related water demands are evaluated for various GHG mitigation targets.  The resulting pollutant projections could be used for more detailed emission and air quality modeling through SMOKE and CMAQ.

Incorporating Air Pollutant Emission Factors and State-Level Controls and Energy Policies within the GCAM-USA Integrated Assessment Model

Wenjing Shi^1,2, Yang Ou^1,2,3, Dan Loughlin², Chris Nolte², Steve Smith⁴, Catherine Ledna⁴

¹ Oak Ridge Institute for Science and Education

² Office of Research and Development, U.S. Environmental Protection Agency, RTP, NC

³ Environmental Sciences and Engineering, University of North Carolina at Chapel Hill

⁴ Joint Global Change Research Institute, Pacific Northwest National Laboratory, College Park, MD 

The Global Change Assessment Model (GCAM) is an integrated assessment model that links representations of economy, energy sector, land use, and climate within an integrated modeling environment. GCAM-USA, which is an extension of GCAM, provides U.S. state-level resolution within a regionalized global modeling framework. The overall goal of our research is to explore how GCAM-USA can be used to address air, climate, and energy system goals simultaneously and cost-effectively.

Here we incorporate characterizations of U.S. air pollutant emission factors and air pollutant controls as well as state-level air pollutant, climate and energy policies into GCAM-USA. Emission factors for technologies of future vintages are obtained from a number of different sources, depending on the sector. State-level emission caps are implemented, approximating the Cross State Air Pollution Rule (CSAPR) and the Clean Power Plan (CPP). The CSAPR representation places constraints on state-level, electric sector NOx and SO2 for a subset of U.S. states. The CPP representation places caps on state-level electric sector CO2 for all U.S. states except Hawaii. CSAPR and CPP caps were derived from EPA analyses of each rule. Pollutant controls for NOx and SO2 are represented using marginal abatement curves.

We show the resulting projections of emissions and electricity generation, and compare them with state-level results of other EPA models. The results indicate general agreement at both the national level and for specific states. We conclude with examples of how the extended GCAM-USA model can be used to support coordinated, long-term air, climate and energy planning.

Development of GUIDE (GHG and air pollutants Unified Information Design system for Environment) system
Jung-Hun Woo 1, Younha Kim 1, B.H. Baek 2, Seungjick Yoo 3, Yoonkwan Kim 4
1 Department of Advanced Technology Fusion (DATF), Konkuk University, 120 Neungdong-ro, Gwangjin-gu, Seoul 05029, Korea
2 CMAS Center, University of North California, USA
3 Graduate school of International Service, Sookmyung Women's University, Seoul, Korea
4 GreenEcos, Seoul, Korea

With the support of Korea Environmental Industry & Technology Institute (KEITI), the GUIDE (GHG and air pollutants Unified Information Design system for Environment) system is under the development. It is an unified environmental information design system that connects air pollution emissions control and costs with air quality benefits, health and economic impacts to assist researches and environmental policy makers with a user-friendly GUI. The main features of the GUIDE are; 1) the new macro economy-based Benefit-Cost(B-C) model for decision making, 2) state-of-science source-receptor relationship surface which can examine impacts of emissions control in realtime even with non-linear chemical reactions, 3) implementation of integrated GHGs and Air Pollutants(APs) emissions inventory for Korea, and 4) incorporation of China and North Korea emission inventories to quantify out-of-region contribution. The simultaneous optimization for bi-directional co-benefits (i.e. co-benefits of APs and GHGs control) is the ultimate function of GUIDE system. In the conference, we will present design issues and initial development progress of the GUIDE.

Acknowledgement : This subject is supported by Korea Ministry of Environment as "Climate Change Correspondence Program.

Exploring Conditions Leading to Wintertime Ozone Episodes in Natural Gas Fields

Yuling Wu

Arastoo Pour Biazar

Bernhard Rappenglueck

Exploiting the WRFchem single column model to study factors contribute to wintertime ozone production in natural gas fields

Air Quality and Acid Deposition Forecast of South Athabasca Oil Sands Development Applying CMAQ Model

Wen Xu¹ and Xin Qiu²

1 Alberta Environment and Parks, Edmonton, Alberta, Canada

2 Novus Environmental Inc., Guelph, Ontario, Canada

In situ oil sands development is expected to dominate bitumen production in the coming decades and much of it will be located in the south Athabasca oil sands area (SAOS). In order to assess impact of environmental footprints of oil sands production in the future as the air portion of SAOS regional strategic assessment, air quality in SAOS for baseline case in year 2010 was simulated and the modeling results were evaluated with monitoring data as a foundation of this assessment (presented as phase 1 in CMAS 2014 annual conference). Based on developed emission inventories and modelling inputs for forecasted future development scenarios, this study (phase 2) applies CMAQ model to simulate the ground level concentrations of ozone, PM_2.5, PM₁₀, NO₂, SO₂, CO, and acid deposition in SAOS for future development scenarios of high and low production levels in year 2020, and high production level in year 2050.

The CMAQ predictions for 2020 low production, 2020 high production and 2050 high production scenarios demonstrate the estimated impacts of future development scenarios on ambient air quality in the SAOS. By comparing air quality forecasts of the future scenarios with baseline year 2010, CMAQ model predicts almost no change of ground level SO₂ concentrations, slight increases of ground level NO₂ and CO concentrations, slight decrease of O₃, smaller increase of PM_2.5 and PM₁₀ concentrations in the major part of SAOS area and larger increase of PM_2.5 and PM₁₀ concentrations in the southern portion of SAOS area. In addition, areas in SAOS exceeding provincial standard of annual total deposition for all the future year scenarios shrink but annual total deposition increases locally nearby sites of newly commissioned central processing plants in the future scenarios.

With the comparisons amongst future scenarios including 2020 low production, 2020 high production and 2050 high production, the model predicts no significant difference between future scenarios of 2020 low production and 2020 high production for O₃, PM_2.5, PM₁₀, SO₂, NO₂, CO, annual nitrogen deposition, annual sulphur deposition and annual total deposition. The model also predicts insignificant difference between 2020 scenarios and 2050 high production for O₃, SO₂, NO₂, CO, but increase particularly in the southern part of SAOS for PM_2.5 and PM₁₀, and locally significant increase of annual total deposition near newly commissioned central processing plants in the 2050 high production scenario.

Development of an activity-based marine emission inventory using AIS data

Bruce Ainslie, Monica Hilborn and Robert Nissen

A spatially- and temporally-resolved marine emission inventory has been created over Pacific Northwest's Georgia Basin for the year 2015 through the use of AIS data. Over 11 million ship position records were collected and used to construct the inventory. Each record consisted of vessel position (latitude and longitude), vessel speed and direction, vessel name and International Marine Organization (IMO) number and draught. By matching vessel IMO number against a master database of vessel engine characteristics (e.g. maximum continuous power rating, vessel service speed, auxiliary engine size, etc.) a vessel's combustion emissions were estimated every 15 minutes while within the Georgia Basin. For tankers, vessel draught was also used to estimate petroleum product load and hence fugitive VOC emissions as well as fugitive loading emissions. Fugitive VOC emissions from the movement of petroleum products by barge was also estimated. Emissions estimates were evaluated using fuel sales data from BC Ferries vessels operating on a number of routes within the Georgia Basin. The marine emission inventory was also tested in Environment and Climate Change Canada's AURAMS photochemical model.

Use of SMOKE model outside of USA: Mobile sources emission inventory using area type approach.

Igor Baptista de Araujo and Taciana Toledo de Almeida Albuquerque

SMOKE was used to prepare spatially and temporally (hourly) averaged vehicular emissions from 117 roads in a metropolitan region in Brazil, using an "Area" approach (sources of pollutants were represented by area sources, instead use of USEPA MOVES, due to lack of data). Area sources file ($ARINV) were composed of regions exposed to various materials and urban roads, with their chemical nature (speciation) represented by a unique (but composed), profile derived from Speciate data.

MEGAN vs BEIS in Texas: A biogenic model showdown
Doug Boyer, Miranda Kosty, Jim Smith, Ph.D., Marvin Jones, Ph.D.

The TCEQ has developed biogenic emission inventories for 2012 using the MEGAN and BEIS models for photochemical model input. This presentation will showcase the development of the model inputs using recent land cover and emission factor research. Biogenic emissions from the two models will be compared to Auto-GC measurements in Texas. CAMx results will also be presented to illustrate how current photochemical models respond to changes in biogenic emission inputs.

Effects of including nitrogen oxides emissions due to lightning on CAMx model performance in Texas

Shantha Daniel

Andrea Zuzack

The Texas Commission on Environmental Quality (TCEQ) primarily uses the Comprehensive Air Quality Model with Extensions (CAMx) for attainment demonstration modeling. However, the TCEQ currently does not include nitrogen oxides emissions due to lightning (LNO_X) in its modeling emissions inventory. The Community Multi-Scale Air Quality Model (CMAQ) contains an inline option for calculating and including LNO_X. CMAQ's inline option derives LNO_X from user derived flash counts and outputs a diagnostic file with LNO_X emissions. This poster illustrates an approach used by the TCEQ to include LNO_X inventory derived from flash counts in a CAMx simulation. A comparative analysis is presented for both models, with and without LNO_X inventories.

Improved wildfire smoke modeling, AIRPACT-Fire, for enhanced communication of human health risk
Yunha Lee 1, Joseph K. Vaughan 1, Serena H. Chung 1, Adam Kochanski 2, Susan O'Neill 3, Farren Herron-Thorpe 4, Matt Kadlec 4, Brian Lamb 1
1 Laboratory for Atmospheric Research, Washington State University
2 Atmospheric Science Department, University of Utah
3 Pacific Wildland Fire Sciences Laboratory, USDA Forest Service
4 Air Quality Program, Washington State Department of Ecology

AIRPACT-5 is an operational forecasting framework for regional air-quality run by Washington State University. It uses CMAQ for chemistry and dispersion; SMARTFIRE, BlueSky and SMOKE for emissions; and Google Maps web display for dissemination to the public. WRF-SFire is a coupled atmospheric-wildfire model that simulates detailed, coupled atmospheric and fire behavior modeling and improves wildfire emissions modeling. AIRPACT-Fire, the AIPRACT-5 framework running with WRF-SFire, is funded by The Joint Fire Science Program to demonstrate delivery of enhanced wildfire related air-quality (AQ) forecast and nowcast (i.e., short-term forecasting) information. It will be the state-of-the-science for modeling the effect of wildfire smoke on AQ. Target audiences include the public and AQ and medical professionals. Messaging will be delivered to the target audiences by various mediums such as the web, smart-phones, emails and/or SMS texts and automated phone messaging. Besides the WRF-SFire, the new AIRPACT-Fire system will include the following enhancements:

• The NOAA National AQ forecast will be used to obtain boundary conditions.
• NOAA Earth System Modeling Laboratory High Resolution (3-km) Rapid Refresh (HYYR) WRF results will be used to produce nowcasts, delivering updated smoke results every few hours during regional wildfire events.
• Visual Range (VR) will be explored through a linked smart phone app and on-line co-laboratory providing for collection and comparison of VR estimates, VR derived from digital imagery, VR from monitors and VR from model results.

An analysis of sensitivity of MOVES emissions estimates to traffic data and comparison to grid-cell estimates and near-road measurements from the Las Vegas field study

R. Chris Owen, Heather Simon, Alison Eyth, Sue Kimbrough, Sharon Phillips, Jeff Vukovich, Michelle Snyder

A near-road measurement campaign was conducted along I-15 in Las Vegas, NV, from December, 2008 through January, 2010. Measurements included CO, NO, NOx, BC, PM2.5, and meteorological parameters  and were collected at 100 m west of the roadway (upwind) and at 20, 100, and 300 m each of the roadway (downwind). Traffic data was also collected at the site, which included speed and length bins for each lane along I-15. This work uses the traffic field data to develop site-specific running exhaust emissions rates based on  the Motor Vehicle Emission Simulator (MOVES) model and tests the sensitivity of emissions estimates to various MOVES run specifications, including county-specific vs national default fleet-mixes and age distributions. The site-specific emissions are also compared with gridded emissions developed as inputs for a national CMAQ simulation. The comparison with CMAQ input emissions focuses on diurnal and weekday and weekend traffic activity and total highway running exhaust emission rates from the on-road sector. The impact of various emissions rates on ambient concentrations are also evaluated by comparing local measurements with project-level dispersion model simulations. 

The 2014 National Emission Inventory for Rangeland Fires and Crop Residue Burning

George Pouliot and Venkatesh Rao

The National Emissions Inventory is developed on a triennial schedule. In this paper, we summarize the methods used, challenges, and results in the development of the 2014 National Emissions Inventory (NEI) for these 4 biomass burning sectors: wildfires, prescribed fires, rangeland, and crop residue burning. In the 2011 NEI, Biomass Burning accounts for approximately 1/3 of the total PM2.5 in the NEI, and is the top emitting data category for PM2.5 emissions. In addition, Biomass Burning contributes significantly to total VOC and a number of Hazardous Air Pollutants (HAPS) including formaldehyde, acrolein, and acetaldehyde. We will summarize the methods for these 4 sectors as well as the challenges of developing the inventory for these highly variable emission sources.

Incomplete sulfate aerosol neutralization despite excess ammonia in the eastern US: a possible role of organic aerosol

Rachel F. Silvern, Daniel J. Jacob, Patrick S. Kim, Eloise A. Marais, and Jay R. Turner

Acid-base neutralization of sulfate aerosol (S(VI) = H2SO4(aq) + HSO4- + SO42-) by ammonia (NH3) has important implications for aerosol mass, hygroscopicity, and acidity. Surface network and aircraft observations across the eastern US show that sulfate aerosol is not fully neutralized even in the presence of excess ammonia, at odds with thermodynamic equilibrium models. The sulfate aerosol neutralization ratio (f = [NH4+]/2[S(VI)]) averages only 0.51 +/- 0.11 mol mol-1 at sites in the Southeast and 0.78 +/- 0.13 mol mol-1 in the Northeast in summer 2013, even though ammonia is in large excess as shown by the corresponding [NH4+]/2[S(VI)] ratio in wet deposition fluxes. There is in fact no site-to-site correlation between the two quantities; the aerosol neutralization ratio in the Southeast remains in a range 0.3-0.6 mol mol-1 even as the wet deposition neutralization ratio exceeds 3 mol mol-1. While the wet deposition neutralization ratio has increased by 4.6% a-1 from 2003 to 2013 in the Southeast US, consistent with SO2 emission controls, the aerosol neutralization ratio has decreased by 1.0-3.2% a-1. Thus the aerosol is becoming more acidic even as SO2 emissions decrease. One possible explanation is that sulfate particles are increasingly coated by organic material, retarding the uptake of ammonia. The ratio of organic aerosol (OA) to sulfate increases over the 2003-2013 period as sulfate decreases. We implement a kinetic mass transfer limitation for ammonia uptake to sulfate aerosols in the GEOS-Chem chemical transport model and find improved agreement with surface and aircraft observations of the aerosol neutralization ratio. If sulfate aerosol becomes more acidic as OA/sulfate ratios increase, then controlling SO2 emissions to decrease sulfate aerosol will not have the co-benefit of suppressing acid-catalyzed secondary organic aerosol (SOA) formation.

The predicted impact of VOC emissions from Marijuana cultivation operations on ozone concentrations in Denver, CO.

Chi-Tsan Wang, Christine Wiedinmyer, William Vizuete

Colorado is the first state to legalize the industrial-scale cultivation of marijuana plants. As a result, thousands of marijuana cultivation operations are present throughout the greater Denver area. The plants found in these cultivations have the potential to release a significant amount of biogenic VOCs, such as monoterpene(C10H16), alpha-pinene, and D-limonene. Further, many cultivations are located in the metropolitan area resulting in these plumes interacting with other urban emission sources resulting in the potential to impact ozone concentrations. This is critical for the city of Denver as they are designated as "Moderate" status (86ppb~100ppb) in non-attainment 8-hour ozone of the National Ambient Air Quality Standard (NAAQS). The little research done on marijuana has focused on the impact of smoke on indoor air quality (Martyny, Serrano, Schaeffer, & Van Dyke, 2013), or the spatial and temporal issues associated with the smell of marijuana chemical compounds (Rice & Koziel, 2015). There have been no studies, however, that have identified or quantified emission rates from marijuana growing cultivations and their impact on ambient ozone. This work will use a regulatory air quality model developed by ENVIRON, UNC-IE and the air quality management agencies of Colorado, Utah, and Wyoming, to predict the impact on ozone by the addition of marijuana cultivation operation emissions. The Comprehensive Air Quality Model with Extensions, CAMx, was applied based on the Three-State Air Quality Modeling Study (3SAQS) for the entire year of 2008. For all marijuana emissions, we assumed they consisted entirely of monoterpenes (C10H16). The location and type of facility data were provided by the Department of Revenue, Colorado. The magnitude of emission rates for each facility was based on the plant count limitation of state law and the emission database based on Wiedinmyer et al., 2008. Due to the high uncertainty in these emission rates, this work will present ozone impacts due to a variety of sensitivity runs where these rates will be varied. Changes in predicted surface ozone concentrations will be quantified as well as impacts on ozone chemical processes.

High-resolution emission inventories of agricultural fugitive dust in China
Ruimin Li 1, Weiwei Chen 1, Daniel Q. Tong 2,3, Hongmei Zhao 1, Shichun Zhang 1, Xuelei Zhang 1, Aijun Xiu 1,4 1Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences 2Air Resources Lab., NOAA Center for Weather and Climate Prediction, College Park, MD 3Cooperative Institute for Climate and Satellite, University of Maryland College Park, MD 4Institute for the Environment, University of North Carolina at Chapel Hill, NC

Emissions from agricultural sources contribute to air quality problems in China substantially. In this study agricultural PM10 and PM2.5 emission inventories at the county level in China are developed. Based on a bottom-up method, we collected the agricultural activity data in each county in China for the year of 2012. These data include conventional farming operations, i.e., agricultural burning, land preparation, harvesting, grain processing, fertilizer application, livestock production, and farming equipment. The emission factors of these operations and crop areas are calculated at the county level. Then the county-level PM10 and PM2.5 emission inventories from agricultural operations are developed and their spatial-temporal distributions analyzed. The results show the total amount of agricultural PM10 and PM2.5 emissions in China were 522.5104 ton and 322.9104 ton per year, respectively. We divide China into seven regions and ranked total amount of PM10 and PM2.5 emissions in these regions from the largest to the smallest as: East, Northeast, Northwest, Central, Southwest, North, and South. The contribution of emissions from agricultural burning is the largest, which accounted for 41.7% and 76.7% (mean distribution from 2001 to 2012) of the total PM10 and PM2.5 emissions, respectively. In addition, the land preparing practice contributes 46.8% and 13.4% to the total PM10 and PM2.5 emissions, respectively. The spatial distribution of agricultural PM emissions shows the regions with the largest PM emissions are in the East and Northeast China, e.g. in the provinces of Zhejiang, Shandong, Henan and Heilongjiang. The temporal profiles of agricultural emissions of PM10 and PM2.5 in China have the largest peak in the fall months and another smaller peak in the summer months. This newly developed agricultural PM emission inventories will be incorporated with the emission inventories from other sectors and input into the CMAQ model to improve air quality modeling in China.

Development of Current and Future-year Point Source Air Emissions Inventories for Alberta Province of Canada
Fuquan Yang, Xin Qiu, Jenny Vesely, Shannon Testart, Hamish Hains.

Abstract:


This study focused on developing an updated provincial scale point source air emissions inventory for 2010 base year across the Province of Alberta, Canada. A comprehensive method was developed by merging the existing 2010 inventory with other recent data provided by the Province, collecting and compiling new data to fill any data gaps. Additionally, the 2010 inventory for Alberta industrial sources has also been updated to a more recent year (i.e., 2013).

The study also provides an approach to overcoming challenges of how to harmonize data from various sources with different levels of QA/QC process throughout the inventory development. Furthermore, based on the updated baseline emission and the existing Alberta province-wide future emissions, we developed the future-year scenarios point source emissions for the future year of 2030 and 2045 periods for Alberta province of Canada. The process included reviewing the recent existing province-wide forecast point emissions and analyzing their strengths and weakness. We also introduced several improvements as more information became available. In this paper we will also describe the technical approach for developing the 2030 and 2045 inventories focusing on regions and province-wide scale.

Development of 2014 Georgia Wildland Fire Emission Inventory

Tao Zeng, Di Tian, And James Boylan

Wildland fires burned about 1.38 million acres in Georgia during 2014 and emitted large amounts of air pollutants such as particulate matter, nitrogen oxide (NOx), volatile organic compounds (VOCs), and carbon monoxide (CO). These 2014 emissions have been estimated by the Georgia Environmental Protection Division (EPD) using the same method as used in previous Georgia wildland fire emission inventories. Also, they have been submitted to the U.S. Environmental Protection Agency (EPA) to be included as part of 2014 National Emission Inventory (NEI). The 2014 NEI requires emissions to be reported by two combustion phases (i.e. flaming and smoldering) as compared to previous NEIs that only required total emissions. NOx emissions are usually higher during the flaming combustion phase (more complete combustion), while emissions of CO, VOC and NH3 are usually higher during the smoldering phase due to incomplete combustion. Therefore, the emissions during each combustion phase should be estimated using emissions factors and fuel consumption factors specific to the combustion phases. In addition, new speciation profiles and plume rise parameters should be developed for each combustion phase when used in air quality modeling.

To develop the emissions by combustion phases, EPD used the CONSUME model to separate fire emissions into 3 stages:  flaming, smoldering, and residual smoldering. The emissions from both the flaming and smoldering stages in CONSUME were combined to represent emissions during the flaming phase since emissions from these two combustion stages co-exist and measured emissions factors often represent mixed emissions from these two combustion stages.  The emissions from the residual smoldering stage in CONSUME was used to represent emissions during the smoldering phase.  In comparison, EPA used the CONSUME model in a different way to develop the wildland fire inventory in the draft 2014 NEI.  Emissions during flaming in CONSUME was used to represent emissions during the flaming phase, and the combined smoldering and residual smoldering emissions in CONSUME were used to represent emissions during the smoldering phase.  EPD compared the two wildland fire inventories.

EPD did not use the satellite remote sensing burn records from EPA.  Instead, EPD collected local fire event data from the state and federal forestry agencies and military databases. In Georgia, the burned area in 2014 was 1.38 million acres, with 1.36 million acres from prescribed fires and the remainder from wildfires.  This is 58% more acres than was estimated by EPA using satellite remote sensing. EPD estimated 2014 annual fire emissions of PM2.5, PM10, CO, NOx, SO2, and VOCs.  Except for VOCs, the EPD estimates are close to EPA's estimates due to higher emission factors adopted by EPA which compensates for their lower burn acres. However, a much higher VOC emission factor led to 2.8 times more VOC emissions in the EPA draft 2014 NEI compared to EPD's VOC estimates.

Canadian Anthropogenic Methane and Ethane Emissions: A Regional Air Quality Modeling Perspective

Junhua Zhang¹, Michael D. Moran¹, Qiong Zheng¹, and Steve Smyth²

¹Air Quality Research Division, Environment and Climate Change Canada, 4905 Dufferin Street, Toronto, Ontario, M3H 5T4, Canada
²Pollutant Inventories and Reporting Division, Environment and Climate Change Canada, 351 Saint-Joseph Blvd., Gatineau, Quebec, K1A 0H3, Canada

Methane (CH4) is one of the most prevalent greenhouse gas (GHG) species, and tonne for tonne its global warming potential is more than 25 times that of CO2. However, due to its low photochemical reactivity, methane, along with ethane, is usually not classified as a volatile organic compound (VOC) and is treated crudely in air quality models. In 2013, 15% of Canada's anthropogenic GHG emissions (CO2 equivalent) were from methane, 43% of which were from the natural gas and petroleum sector. For the U.S., in 2014 CH4 accounted for about 11% of all U.S. anthropogenic GHG emissions, 33% of which were from the natural gas and petroleum sector. Due to the importance of methane emissions from the natural gas and petroleum sector, the Canadian, U.S., and Mexican governments recently committed to taking joint action to reduce methane emissions from this sector by 40-45%below 2012 levels by 2025. Since the natural gas and petroleum sector is also a large source of VOC emissions that affect air quality, the proposed methane emissions reductions should have direct or indirect benefits for both climate and air quality.

In this study, Canadian anthropogenic methane emissions have been estimated from the VOC emissions inventory compiled for air quality modeling purposes. VOC-to-TOG (Total Organic Gas) conversion factor and weight-percentage-of-methane fields from TOG speciation profiles obtained from the U.S. EPA''s SPECIATE database were used to calculate sector-specific methane emissions based on the total VOC emissions for various anthropogenic source sectors, including the natural gas and petroleum sector. This approach can then be used to calculate self-consistent estimates of the magnitude of methane and VOC emissions reductions associated with various proposed control strategies and it allows the spatial and temporal distribution of these reductions to be examined. These estimated methane emissions will also be compared with the methane emissions in the Canadian national GHG inventory compiled for climate research purposes. This comparison will provide information on the internal consistency between different types of inventories and on the uncertainty of methane emissions estimates. The same approach has been applied to estimate emissions of ethane, which is a co-pollutant from some but not all methane source sectors.

THE MODEL FOR SIMULATING THE ROCKET EXHAUST FORMATION AND DISPERSION AND ITS INTEGRATION WITH CMAQ FOR LONG RANGE ASSESSMENT

Erick Giovani Sperandio Nascimento, Davidson Martins Moreira, Taciana Toledo de Almeida Albuquerque

During the launch of rockets and spacecrafts, huge and hot clouds are generated near the ground level, and are composed by buoyant exhaust products, such as alumina, carbon monoxide and hydrogen chloride. This process takes a few minutes to occur, and generally populated areas nearby the launching center may be exposed to high levels of hazardous pollutant concentrations within few minutes to less than a couple of hours. Due to the specificity of the representation of the source term -- which is the rocket exhaust cloud - - and since a receptor can be impacted in less than one hour, common air quality models were not designed to deal with such a unique problem. Furthermore, the cloud may be transported to farther distances and impact receptors in longer time and space scales. Thus, the launching centers around the globe, like spaceports, need to operationally assess the short and long range impacts of rocket launch events in the environment through meteorological and air quality modeling. For this end, this work presents the development of a new model called Model for Simulating the Rocket Exhaust Dispersion - MSRED, which was designed to simulate the formation, rise, expansion, stabilization and dispersion of rocket exhaust clouds for short range assessment, which is able to directly read meteorological data from WRF model output. And, for the long range modeling, the MSRED was built to generate a ready-to-use initial conditions file to be input to the Community Multi-scale Air Quality (CMAQ) model, since it represents the state-of-the-art in regional and chemical transport air quality modeling. This hybrid, modern and multidisciplinary system is the basis of a modeling framework ready to be operationally used at the Alcantara Launching Center (the Brazilian gate to the space), for pre and post-launching simulations of the environmental effects of rocket operations.

Implementation of Canopy Reduction mechanism to CMAQ

Jan A. Arndt, Volker Matthias, Armin Aulinger, Johannes Bieser, Matthias Karl

Nitrogen oxide and nitrogen dioxide (NOx) emissions from biogenic sources are responsible for five to ten percent of European total NOx emissions. These biogenic emissions escape mainly as nitrogen oxide from microbial processes in soil. It is oxidized in the lower layers of the earth's atmosphere to nitrogen dioxide, which affects amongst others, the natural ozone cycle. This nitrogen dioxide is better water soluble which leads to an partly uptake of the pollutant by the leaves of plants.

At Helmholtz-Zentrum Geesthacht we use the CMAQ-SMOKE modeling system to study the impact of natural and anthropogenic emissions on the coastal areas of northern Europe. The Biogenic Emission Inventory System is used to generate biogenic emissions of VOCs and nitrogen oxide, which is emitted in the lowest model layer. In CMAQ, the interaction between the pollutants and the vegetation is taken into account in the calculation of the dry deposition velocity in form of a stomatal resistance. But after the chemical transformation to nitrogen dioxide, there is no further consideration of the uptake by leaves.

Based on the stomatal resistance calculation for the dry deposition velocity, we developed an air concentration canopy reduction module that works comparable to an extra depositional loss module. To investigate the impact on model results with respect to the uptake of nitrogen dioxide by leaves, we implemented a "Canopy reduction module" in CMAQ after the gas-phase chemistry subroutine. To test the sensitivity of the model to the position of the "Canopy reduction module", we created versions with the module in the emission and the dry deposition subroutines as well.

Lightning NOx Production in CMAQ: Part II - Parameterization Based on Relationship between Observed NLDN Lightning Strikes and Modeled Convective Precipitation Rates

Daiwen Kang, Kristen Foley, Nicholas Heath, David Wong, Rohit Mathur, Shawn Roselle, and Jon Pleim

This study is a continuation of Part I of our work that used hourly NLDN lightning data to produce lightning NO_X in CMAQ, which was developed for retrospective model applications. However, for modeling exercises where the observed lightning strikes are not generally available (e.g., real-time air quality forecasts), we have developed a lightning NOx parameterization that is based on the relationship between the observed NLDN lightning strikes and model predicted convective precipitation rates. The scheme was specifically developed using data over the continental United States for a time period spanning over a decade. Here, we present a detailed analysis describing how the model-predicted convective precipitation rates related with the observed NLDN strikes in space and time over the decade analyzed. Preliminary results show distinctive and unique spatial patterns for the relationship between observed lightning strikes and modeled convective precipitation rates over the continental United States. Based on these results, a new parameterization for lightning NO_X production in CMAQ is developed and evaluated. Comparisons with the previous CMAQ parameterization (which used monthly lightning data and convective precipitation rates) show that the new scheme is comparable to the previous one. These results are encouraging because the new scheme doesn't depend on knowledge of the observed lightning strikes (as the previous one did) and is therefore suitable for real-time air quality forecasts.

New Particle Formation and Growth in CMAQ-NPF: Application of Comprehensive Modal Methods to Observations during CalNex and CARES

Benjamin N. Murphy¹, Francis S. Binkowski², Havala O. T. Pye¹, Tinja Olenius³, Ilona Riipinen³ and Jon Pleim¹

1National Exposure Research Laboratory, Research Triangle Park, North Carolina

²Institute for the Environment, University of North Carolina, Chapel Hill, North Carolina

³Department of Analytical Chemistry and Environmental Sciences, Stockholm University, Stockholm, Sweden

Secondary formation and subsequent growth of ultrafine atmospheric particles is an important source of particles to the atmosphere, in addition to direct emissions from combustion sources. If we are to use regional chemical transport models like CMAQ to assess the impact of anthropogenic gas and particle emissions on human health and cloud microphysical properties, we must be able to predict the strength and variability of secondary particle formation events. In a traditional model, these events rely principally on sulfuric acid and, more recently, a stabilizing molecule (e.g. NH₃, amine, etc) to begin. However, emerging evidence shows that organic compounds are capable of participating with sulfuric acid or, when sulfuric acid concentrations are low, generating particles on their own. Moreover, because they are so abundant, organic vapors drive the growth of particles to larger sizes in many environments.

We implement into the CMAQ model a new aerosol processing module designed for robust prediction of particle number concentrations, sources and sinks, while accounting for the participation of inorganic and organic compounds. The new module leverages the speed and flexibility of modal aerosol techniques with state-of-the-art schemes for treating new particle formation (via results from recent laboratory observations and the highly detailed Atmospheric Cluster Dynamics Code), coagulation, and intermodal transference. We apply the new model to observations made during the CalNex and CARES 2010 campaigns and evaluate model performance against observed number concentrations and size distributions.

Halogen chemistry in the CMAQ model
Golam Sarwar, Kristen Foley, Heather Simon, Kathleen Fahey, Jia Xing, Rohit Mathur

Halogens (iodine and bromine) emitted from oceans alter atmospheric chemistry and influence atmospheric ozone mixing ratio. We previously incorporated a representation of detailed halogen chemistry and emissions of organic and inorganic halogen species into the hemispheric Community Multiscale Air Quality model. We performed simulations without the halogen chemistry as well as with the halogen chemistry without and with the photolysis of higher iodine oxides. The halogen chemistry without the photolysis of higher iodine oxides lowered summertime mean ozone by ~15% over marine environments; while the halogen chemistry with the photolysis of higher iodine oxides lowered ozone by ~48%. Here, we revise the halogen chemistry by (1) updating all photolytic reactions involving halogen species, (2) incorporating several heterogeneous reactions, and (3) revising organic and inorganic halogen emissions. We perform model simulations without and with the halogen chemistry for the summer months. We use the first month as model spin-up and analyze the results for the remaining months. Our model results confirm that the halogen chemistry effectively reduces ozone not only over surface marine environments but also aloft. However, the spatial impact on ozone varies substantially, and the accompanying paper presents a detailed analysis of the spatial impacts. We compare our current model results to those obtained with previous simulations and also with available observations.

A new version of the Community Multiscale Air Quality Model: CMAQv5.2
Jonathan Pleim and the CMAQ development Team, USEPA

A new major version of CMAQ has been developed and evaluated and will be released to the community in the Fall of 2016. The major new features of CMAQv5.2 include new treatment of organic aerosols, new windblown dust model, and a new gas-phase chemical mechanism (CB6). The new organic aerosol treatment and wind-blown dust address know deficiencies in earlier model versions. There have also been modifications to the PX LSM and ACM2 PBL model in WRFv3.8 that affect CMAQ results. In addition, CMAQv5.2 will include the instrumented capabilities: the Decoupled Direct Method (DDM) and Sulfur tracking. The presentation will briefly describe the major new features and introduce other talks and posters that go into greater detail. Highlights of evaluation of CMQv5.2 and comparisons to earlier versions will also be shown. Progress and future developments of the Next Generation Air Quality model will be summarized.

Predicting PM2.5 Concentrations that Result from Compliance with National Ambient Air Quality Standards

James T. Kelly, Adam Reff, and Brett Gantt

Office of Air Quality Planning & Standards, US EPA, RTP, NC 27711

PM_2.5 concentrations that correspond to just meeting existing or potential alternative National Ambient Air Quality Standards (NAAQS) are needed to inform risk assessments conducted during periodic NAAQS reviews. Previously, ambient PM_2.5 concentrations have been adjusted to lower values according to prescribed spatial patterns for this purpose. Although there is some justification for previously used approaches, they do not directly reflect the processes known to determine PM_2.5 concentrations (i.e., emissions, transport, chemistry, and deposition). In this study, a new technique is demonstrated for adjusting PM_2.5 concentrations to correspond to just meeting existing and potential alternative NAAQS. The technique involves developing site specific PM_2.5 adjustment factors by combining results of photochemical grid modeling with the Community Multiscale Air Quality (CMAQ) model and ambient measurements using the Speciated Modeled Attainment Test-Community Edition software. In this presentation, the PM_2.5 adjustment approach will be illustrated by describing results of the method for representative case studies and by comparing with results based on a previously used method.

Enhancements to an Agriculture-land Modeling System - FEST-C and Its Applications

Ellen Cooter¹, Limei Ran¹, Verel Benson², Dongmei Yang³, Ruoyu Wang⁴, Yongping Yuan⁴

¹Computational Exposure Division

ORD NERL/USEPA, Research Triangle Park, NC

²Verel W. Benson, Benson Consulting, 200 Haywood Ct, Columbia, MO 65203, USA

³Institute for the Environment

University of North Carolina at Chapel Hill, NC USA

⁴Systems Exposure Division

ORD NERL/USEPA, Research Triangle Park, NC

The Fertilizer Emission Scenario Tool for CMAQ (FEST-C) system was originally developed to simulate daily fertilizer application information using the Environmental Policy Integrated Climate (EPIC) model across any defined conterminous United States (U.S.) CMAQ domain and grid resolution. This EPIC output information is a required input for CMAQ modeling with the bi-directional NH₃ option. Since the first release of FEST-C v1.0 in October 2013, the system has gone through many updates and enhancements up to the recent update release of FEST-C V1.2 in May 2016. As part of this process, the system has been adapted to support applications for air quality studies using CMAQ with the bi-directional NH₃ option in China. We have also developed methods to support scenarios of agricultural land management change, i.e., reallocating grid-cell agricultural land amongst specific crops in response to hypothetical economic and policy changes. Currently, we are enhancing the system to better integrate bi-directional CMAQ air quality simulations with the Soil and Water Assessment Tool (SWAT) modeling system to improve our understanding of hypoxia in the Gulf of Mexico. These enhancements have advanced FEST-C capabilities from simply supporting CMAQ simulations, to becoming a valuable tool for integrated, one-biosphere assessments of air, land and water quality in light of social drivers and human and ecological outcomes.

This presentation will focus on demonstrating system enhancements since the version 1 release. We will also present results illustrating China air quality simulations, SWAT integration and agricultural cropland reallocation.

Source apportionment of biogenic contributions to ozone formation over the United States
Daniel Cohan, Rui Zhang, Alex Cohan, Arastoo Pour Biazar

Biogenic emissions are the leading source of volatile organic compounds (VOCs) and key to ozone formation. A key uncertainty in BVOC emission modeling comes from the estimation of photosynthetically active radiation (PAR) reaching the canopy. Satellite insolation retrieval data has been shown to better capture the spatial and temporal variations of PAR than meteorological models. Here, we compare source apportionments of ozone formation from BVOC with and without satellite retrievals of insolation and PAR and comparing the BEIS and MEGAN biogenic models. The CAMx photochemical model was applied with Ozone Source Apportionment Technology (OSAT) and brute force sensitivity to quantify BVOC contributions to ozone formation during May to September 2011. Satellite insolation/PAR reduced isoprene emission estimates by an average 3%-4% in BEIS and 9%-12% in MEGAN, since the retrievals showed more clouds than WRF. BEIS with satellite retrieved insolation achieved better performance for ozone simulations (R=0.75 and NME=18%) than the simulations with MEGAN, which tended to over-predict ground-level isoprene concentrations. The higher BVOC in MEGAN compared to BEIS counter-intuitively leads OSAT to apportion more of the ozone formation to BVOC (since MEGAN's higher BVOC makes incremental ozone formation more NO_x limited), but brute force shows the opposite pattern by zeroing out MEGAN's larger BVOC inventory. OSAT and brute force also differ in the extent to which source apportionment peaks near NO_x-rich (VOC-limited) urban regions. Domain-wide biogenic emissions on average contribute 18% of ozone concentration at nonattainment receptor sites during episode days, with strongest contributions in the southeastern U.S.

Lightning NOx Production in CMAQ: Part I - Using Hourly NLDN Lightning Strike Data

Daiwen Kang, Nicholas Heath, David Wong, Jon Pleim, Shawn Roselle, Kristen Foley, and Rohit Mathur

Lightning-produced nitrogen oxides (NO_X=NO+NO₂) in the middle and upper troposphere play an essential role in the production of ozone (O₃) and influence the oxidizing capacity of the troposphere. Despite much effort in both observing and modeling lightning NO_X during the past decade, considerable uncertainties still exist with the quantification of lightning NO_X production and distribution in the troposphere. Further, it is challenging for regional chemistry and transport models to accurately parameterize lightning NO_X production and distribution in time and space. In this study, we use observed lightning strikes from the National Lightning Detection Network (NLDN), aggregated as hourly mean values and gridded to our modeling domain, for lightning NOx production in retrospective CMAQ applications. Total column values of lightning produced NO_X using this new hourly NLDN data are compared with those generated by the previous CMAQ parameterization in which the lightning NO emissions are calculated using local scaling factors adjusted by the convective precipitation rate. Several annual WRF-CMAQ simulations were made to test the new update to lightning NOx (previous lightning NO_X parameterization, NLDN hourly gridded data, and no lightning NO_X). Evaluation results are presented for each of the three cases.

Dynamic Evaluation of Modeled Ozone Response to Emission Changes and Improvement of Future Year Ozone Projections in the South Coast Air Basin

Prakash Karamchandani¹, Ralph Morris¹, Andrew Wentland¹, Julia Lester², Timothy French³

¹Ramboll Environ, 773 San Marin Drive, Suite 2115, Novato, CA 94998

²Ramboll Environ, 707 Wilshire Boulevard, Suite 4950, Los Angeles, CA 90017

³EMA, 333 West Wacker Drive, Suite 810, Chicago, IL 60606

Regional-scale photochemical grid models (PGMs), such as CMAQ and CAMx, are used as part of the decision-making process of developing emission control strategies for attaining ozone and PM_2.5 air quality standards in a future year. In these applications, the model is applied for a "base" year and a future year and changes in modeled values between the base and future year are used to project future year air quality under different emission scenarios. Typically, the applications of PGMs in this regulatory mode only evaluate model performance in an operational sense, i.e., they compare model estimates of ozone concentrations for the base year with measurements. They do not include dynamic evaluation, i.e., the evaluation of the ability of the modeling system (the model inputs and the model itself) to respond correctly to historical and recent changes in precursor emissions. However, dynamic evaluation is critically important for regulatory applications because of the economic and technology implications of control measures that are guided by the model response to these measures. This paper presents the results from a study designed to understand how well the CMAQ modeling system used in the latest Air Quality Management Plan (AQMP) for the South Coast Air Basin (SoCAB) predicts historical and recent trends in measured ozone design values. The study includes the application of the model for several historical and recent years in "forecast" mode as well as simulations to understand model sensitivity to meteorology, boundary conditions, and emissions. The results of using published statistical techniques to "correct" model forecasts of design values are also presented.

Updates on Soil NOx parametrization in CMAQ v5.1

Quazi Ziaur Rasool¹, Jesse Bash², Rui Zhang¹, Ellen Cooter², Benjamin Lash³, Daniel S. Cohan¹ and Lok N. Lamsal^4,5

¹Department of Civil and Environmental Engineering, Rice University, Houston, TX

²Atmospheric Modeling and Analysis Division, National Exposure Research Laboratory, Office of Research and Development, US Environmental Protection Agency, RTP, NC, USA

³ School of Natural Sciences, University of California, Merced, CA

⁴Goddard Earth Sciences Technology and Research, Universities Space Research Association, Columbia, MD 21046, USA

⁵NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA

NO_x has been well established to impact the formation of ozone and particulate matter. Soil NO emissions comprise approximately 20% of the global NO_x budget and are a leading source of NO_x in rural and remote areas. NO is emitted from soil as a result of complex biogeochemical interactions of soil N with specific microbial niches. Accurate estimation of soil NO could enhance simulations of ozone, particulate matter, and atmospheric deposition flux in regional air quality models like CMAQ. Uncertainty in the temporal and spatial distribution of soil NO emissions arises from a lack of dynamic representation of soil properties, land use classification and mineral nitrogen availability in the soil.

This study presents an update to the Berkley Dalhousie Soil NOx parametrization (BDSNP) scheme in CMAQ v5.1 to enhance its consistency with CMAQ-EPIC representation of the nitrogen cycle and bidirectional ammonia exchange. This updated parametrization constrains soil NO emissions by incorporating detailed agricultural fertilizer inputs from EPIC and CMAQ modeled N deposition into the soil N pool. We implement fractional land use definitions rather than dominant land use classification for more accurate sub-grid cell soil NO estimates. This updated scheme is currently implemented for the continental US and could be extended to a global scale. Model results are evaluated against observed aerosol and ozone concentrations, satellite observations of NO₂, and deposition fluxes. The impact on total dry and wet nitrogen deposition will be discussed.

Assessment of Intrastate Contributions to Ozone Nonattainment Monitors in Atlanta, GA

Byeong-Uk Kim, Marcus Trail, Di Tian, and James Boylan

When the U.S. Environmental Protection Agency (EPA) revises the National Ambient Air Quality Standards (NAAQS) for a criteria air pollutant (PM, ozone, SO₂, NO₂, CO, lead), states must submit designation recommendations (attainment, nonattainment, unclassifiable) for their jurisdictional areas within 2 years after the promulgation of the new NAAQS. Often, nonattainment areas include not only areas with monitors violating the NAAQS but also neighboring areas contributing to the NAAQS violation. To develop designation recommendation for the 2015 ozone NAAQS, EPA provided guidance which includes the consideration of five factors: air quality, emissions, meteorology, geography, and jurisdictional boundaries. In addition, EPA recommends that state agencies conduct Weight-of-Evidence analysis such as source apportionment modeling with process-based photochemical grid models to quantify contributions from neighboring areas (e.g. counties). In this study, the Comprehensive Air Quality Model with extensions (CAMx) and the Community Multiscale Air Quality model (CMAQ) are used to quantify the county-by-county contributions to ozone monitors in the Metro Atlanta area that are currently violating the 2015 ozone NAAQS. First, we will utilize the Anthropogenic Precursor Culpability Assessment tool implemented in CAMx. Second, we will conduct brute force runs with CAMx and CMAQ by zeroing out NOx and VOC emissions in each county near the Metro Atlanta area. For meteorology, we use the 2011 Weather Research Forecast model outputs that were prepared by EPA for its ozone Transport Rule (TR) modeling. For emissions, we use a subdomain based on the "eh" version of EPA's 2011/2017 TR modeling platform with revisions to some 2017 EGU emissions in the southeastern states. In the presentation, we will discuss modeling results and how these results can be used to develop state designation recommendations for nonattainment areas.

Enhancements to Land Surface Processes for WRF/CMAQ with PX LSM

Limei Ran, Jonathan Pleim, Robert Gilliam, Ellen Cooter

Computational Exposure Division

ORD NERL/USEPA, Research Triangle Park, NC

Land surface processes in land surface models (LSMs) influence the exchange of heat, moisture, momentum, and trace atmospheric chemicals between the land surface and the atmosphere in meteorology and air quality (AQ) modeling systems such as WRF/CMAQ. Among the LSMs available in WRF, PX and Noah LSMs are commonly used for retrospective simulations with PX LSM mainly designed for CMAQ. Unlike climate LSMs with complex dynamic vegetation and hydrology to model processes over decadal to century future periods, the PX and Noah LSMs have relatively simple vegetation and soil models. Both LSMs rely heavily on data initialization and assimilation for high accuracy over relatively short periods. Surface representation including vegetation and surface albedo are typically generated from parameters specified in LSM land use look-up tables. This presentation highlights the use of Moderate Resolution Imaging Spectroradiometer (MODIS) surface products and an advance photosynthesis-based vegetation model to improve land surface processes for WRF/CMAQ PX LSM. With an updated PX LSM WRF/CMAQ, MODIS vegetation input reduces bias of the 2 m Q estimation during the growing season from April to September, but increases surface O₃ simulation bias in April, August, and September in areas where MODIS vegetation is much less than the current model vegetation. A coupled photosynthesis-stomatal conductance model is developed and evaluated for WRF/CMAQ PX LSM. The evaluations of the new vegetation model against FLUXNET and EPA latent heat and ozone flux measurements will be presented. In addition, we are developing a simple irrigation scheme in PX LSM by incorporating MODIS Irrigated Agriculture Dataset for the United States (MIrAD-US) into the updated WRF/CMAQ with MODIS input. We will demonstrate preliminary results from WRF/CMAQ simulations with the simple irrigation scheme.

Source apportionment of fine particulate matter in Yunlin County in Taiwan

Yi-Ju Lee and Fang-Yi Cheng

Yunlin is located in central-southern portion of western Taiwan. The local industrial emissions (Mailiao industry), vehicle exhausts, and burnings of agriculture wastes all contribute to the poorer air quality in Yunlin. Besides, the emissions from nearby power plants, Taichung metropolitan area, and Changhua industrial park also contribute to the local air pollution problem in Yunlin County. The local circulation such as the land-sea breeze might transport the air pollutants toward the inland areas and induce high concentration. From 2014 to 2015, the averaged PM_2.5 concentration in Douliou and Lunbei (the surface monitoring stations located in Yunlin County) are ranked as the top highest stations in Taiwan.

Source apportionment (SA) based on observation data or emission data can provide the relationship between emission sources and concentration of pollutants. This study was conducted to investigate the main emission source that contributes to the PM_2.5 concentration in Yunlin County using CMAQ source apportionment technique.

The WRF version 3.7.1 and CMAQ version 5.0.2 are conducted and the observation nudging technique is applied in WRF modeling to nudge surface observed temperature, wind speed and wind direction data from Taiwan Central Weather Bureau and Environmental Protection Agency to improve the meteorological conditions. The emission inventory is from Taiwan Emission Data System version 8.1. This study focused on the emission from Yunlin and Taichung power plants through BFM (Brute Force Method) and ISAM (Integrated Source Apportionment Method) technique.

The high pollution episode on Nov 8 and Nov 9, 2015 was selected. Nov 8 was associated with a weak synoptic weather condition and PM_2.5 mainly came from local emissions released in Yunlin County. However, Nov 9 was affected by a continental anticyclone and pollutants were transported by a weak northeasterly wind. The PM_2.5 concentrations in Yunlin County are mainly contributed from the emissions released in the upwind Taichung power plants.

A new physically-based windblown dust emission parametrization in CMAQ

Hosein Foroutan, Jeff Young, Limei Ran, Peng Liu, Jonathan Pleim, Rohit Mathur

Dust has significant impacts on weather and climate, air quality and visibility, and human health; therefore, it is important to include a windblown dust emission module in atmospheric and air quality models. In this presentation, we summarize our efforts in development of a physics-based windblown dust emission scheme and its implementation in the CMAQ modeling system. The new model incorporates the effect of the surface wind speed, soil texture, soil moisture, and surface roughness in a physically sound manner. Specifically, a newly developed dynamic relation for the surface roughness length in this model is believed to adequately represent the physics of the surface processes involved in the dust generation. Furthermore, careful attention is paid in integrating the new windblown dust module within the CMAQ to ensure that the required input parameters are correctly configured. The new model is evaluated for the case studies including the continental United States and the Northern hemisphere, and is shown to be able to capture the occurrence of the dust outbreak and the level of the soil concentration. We discuss the uncertainties and limitations of the model and briefly describe our path forward for further improvements.

Modeling the Impacts of Prescribed Burns for Dynamic Air Quality Management

M. Talat Odman, Aditya Pophale and Yongtao Hu

School of Civil and Environmental Engineering, Georgia Institute of Technology, Atlanta, GA, 30332

Our reliance on prescribed burning is expected to increase as a means for managing forested lands and a measure for preventing wildfires. However, increased smoke emissions from burns may degrade air quality in areas already burdened by other emissions. Here, dynamic management is a paradigm that considers the burning needs together with the consequences of burn emissions and maximizes burn capacity while minimizing the impacts on air quality. Through permitting systems already in place in several states, burns can be restricted on poor air quality days and encouraged when meteorological conditions are favorable.

To facilitate dynamic management, we developed an air quality forecasting system that can predict the impacts of prescribed burns. The system consists of WRF version 3.6 for meteorology, SMOKE version 3.5 for emissions other than prescribed burns and CMAQ version 5.0.2 for chemical transport. The Decoupled Direct Method (DDM-3D), available in CMAQ-5.0.2 for sensitivity analyses, is used to predict the burn impacts. To perform well DDM-3D requires an accurate forecast of burn emissions. For this, first the location and acreages of the burns are forecast based on the weather forecast and geographic burning patterns identified by mining a burn permit database. Then, burn emissions are forecast using estimates of understory fuel loads, fuel consumption rates and field-measured and laboratory tested emission factors.

In this presentation, the modeling system, its impact forecasting capability, and the burn emissions forecasting methods will be described. The improvements made from the initial operation in 2015 to 2016 will be reviewed. Results for the 2016 burning seasons will be presented. The forecasting performance will be evaluated through comparisons to satellite observations and ground-based accounts of fires as well as smoke-induced peaks in observed pollutant levels at ground monitors. The ongoing integration of the burn-impact forecasts with the prescribed burning operation in Georgia and plans for expansion to the Southeastern USA will be discussed.

Direct Radiative Effect of Dust Aerosols and Biomass Burning Over East Asia
Xinyi Dong, Joshua S. Fu, Kan Huang
Department of Civil and Environmental Engineering,The University of Tennessee, Knoxville, Tennessee, 37996, USA

Direct radiative effect (DRE) of biomass burning aerosols is usually considered as negative, which means a cooling enforcement on the atmosphere. While IPCC AR4 and AR5 both indicated that the overall radiative effect of biomass burning and mineral dust aerosol may be positive above the cloud, many of the studies tend to estimate DRE of the aerosol in clear sky only. To help improve the understanding, this study evaluated DRE of biomass burning and dust over East Asia under both clear and all sky conditions. Using a regional coupled climate and chemical transport model WRF/CMAQ with heterogeneous dust chemistry module, we found that biomass burning has negative DRE of -20W/m2 at surface layer (SFC) at source region over Indochina under both clear and all sky conditions, but has significant positive DRE of 10w/m2 at top of atmosphere (TOA) at downwind area over South China under all sky conditions. The perturbation of upwelling short wave radiation (SWR) at TOA introduced by biomass burning aerosol was significantly greater under all sky than under clear sky condition, indicating that biomass burning aerosols, if they exist, are always on top of the cloud over South China and absorbing more SWR than scattering back. Therefore, the impact of Southeast Asia biomass burning on regional climate at South China may need to be re-considered. We are still analyzing the DRE of mineral dust and expecting to report the findings in the coming CMAS meeting.

Improving Regional PM2.5 Modeling along Utahs Wasatch Front

Chris Pennell and Nancy Daher - Utah Division of Air Quality

It is well documented that Utah frequently exhibits large winter-time PM_2.5 concentrations. Utah's winter-time exceedances of the 24-hour PM_2.5 National Ambient Air Quality Standards (NAAQS) are attributed to strong temperature inversions. These temperature inversions can last several days during persistent high-pressure, low surface-wind conditions. In turn, low-level mixing layers trap fine particulate matter near the surface and create a public-health concern.

Utah's complex topography creates a unique challenge. Air quality models struggle to obtain adequate performance or accurately portray regional PM_2.5 composition. During winter-time exceedances, particulate nitrates (PNO₃) often represent the largest portion of PM_2.5 mass, contributing more than organic carbon. To achieve adequate performance, past PM_2.5 modeling required disabling the vertical advection scheme and adding additional ammonia.

To address the challenges of simulating PM_2.5 along the Wasatch Front, the Utah Division of Air Quality (UDAQ) has greatly advanced their photochemical modeling platform. Currently, we are using the Comprehensive Air Quality Model with Extensions (CAMx), version 6.3. This recent release of CAMx includes UDAQ-funded updates to the CB6 chemistry mechanism, snow-cover treatment, and surface model. We considerably increased the number of vertical layers (41) and horizontal grid resolution (1.33 km) of the modeling domain. Finally, meteorological modeling inputs have greatly improved to better capture persistent near-surface air stability.

For this presentation, we compare the differences between our current modeling approach and prior efforts that featured the Community Multiscale Air Quality (CMAQ) model, version 4.7. We present results from our recent modeling of the January, 2011 Persistent Cold Air Pool Study (PCAPS) intensive field campaign. Subsequently, we demonstrate a marked improvement in simulating peak PM_2.5 concentrations, PM_2.5 composition, and temporal correlation with observations.

Updating CMAQ secondary organic aerosol properties relevant for aerosol water interactions

Havala O. T. Pye, Benjamin N. Murphy

US Environmental Protection Agency

T. Khoi Nguyen, Annmarie G. Carlton

Rutgers University

Lu Xu, Nga L. Ng

Georgia Institute of Technology

Properties of secondary organic aerosol (SOA) compounds in CMAQ are updated with state-of-the-science estimates from structure activity relationships to provide consistency among volatility, molecular weight, degree of oxygenation, and solubility/hygroscopicity. These updated properties are used to predict interaction of organic aerosol with particle-phase liquid water. Model predictions are compared to observations over the southeast United States during the Southern Oxidant and Aerosol Study. Based on model-predicted OM/OC, SOA is frequently well-mixed with the inorganic phase throughout most of the Eastern US except in urban areas where POA concentrations are high and OM/OC is low. As a result, aerosol liquid water contributes to the partitioning medium for semivolatile organics. Including water in the partitioning medium of OA leads to overestimates in OC at night and reasonable performance during the daytime unless deviations in ideality are considered. Property updates discussed here will be available in CMAQ v5.2.

Predicting the Impact of a Wood-Stove Change-Out Program on Ambient Particle Levels in Utah's Airshed

Nancy Daher, Christopher Pennell

Utah is susceptible to elevated levels of fine particulate matter (PM_2.5) along the Wasatch Front during winter-time inversions. PM_2.5 levels often exceed the national ambient air quality standards during winter, with residential wood-burning combustion accounting for about 20% of total primary PM_2.5 emissions in Salt Lake County. To guide emissions control strategy development, the impact of a wood-stove change-out program on reducing ambient PM_2.5 levels was investigated using the Comprehensive Air Quality Model with extensions (CAMxv6.3). Modeling was performed for a typical cold air pool episode in January 2011. To improve model performance, a two-way nested 4/1.33 km with 41 vertical layers modeling domain was defined and carbon Bond 6 (CB6r3) mechanism was used for gas-phase chemistry calculations. Meteorological input data was retrieved from the Weather Research Forecast (WRF) model, which was configured to better simulate inversion events by implementing improved land use, snow cover and snow albedo characterizations that more closely represent field observations. The response of ambient PM_2.5 levels to a reduction in wood-smoke emissions was estimated by considering different sensitivity simulations that assume a replacement of wood-burning devices by low-emission ones. Model performance was evaluated by comparing CAMx results to speciated ambient particle measurements.

Impacts on Ambient Particulate Matter by Changing Particle Size Distribution from Emissions Using the Community Air Quality Model (CMAQ): A Case Study of Commercial Aircraft emissions from Landing and Take-off
Jiaoyan Huang, Lakshmi Pradeepa Vennam, Francis Binkowski and Sarav Arunachalam
Institute for the Environment, University of North Carolina, Chapel Hill 100 Europa Drive, Suite 490, Chapel Hill, NC 27517

Community Multi-Scale Air Quality model (CMAQ) has been widely used to understand air pollutant transport and fate, such as particles and ozone. However, only a uniform particle size distribution, geometric mean diameter (GMD) and standard deviation (GSD) from anthropogenic emissions is applied in CMAQ which might improperly represent particles emitted into the atmosphere. In this study, a new module has been developed to read emissions from a specific emission sector (in this case, aircraft emissions). This new module assigns various emission split factors (Aitken and accumulation modes) for different species, and has the ability to calculate number concentrations emitted from airport grids using different GMDs and GSDs for different species. The goal of this study is to to investigate impacts of various particle size distributions from aircraft emissions on ambient fine particulate matter concentrations.

Using this modification, significant increased particle number emissions from aircrafts near airports are found, however, mass emissions stay same as default CMAQ process. We hypothesize that ambient number concentrations in Aitken mode near airports would increase. This would impact atmospheric chemical and physical processes near airports. For example, increased surface areas of particles might enhance partition to particles. Overall, modified ESF, GMD, and GSD represent more reliable particle size distributions from aircraft emissions than default CMAQ setting, and Aitken mode particles are important for human health research related to ultrafine particles near airports.

ASSESSMENT OF CUYYENT AND FUTURE IMPACTS OF AIR POLLUTION ON HUMAN HEALTH
U. Parra Maza, P. Suppan
Institute of Meteorology and Climate Research, Atmospheric Environmental Research (IMK-IFU), Karlsruhe Institute of Technology (KIT), Germany
Presenting author's e-mail: peter.suppan@kit.edu

A modeling study have been conducted using the Weather Research and Forecasting (WRF; Skamarock and Klemp, 2008) model coupled with chemistry (WRF-Chem; Grell et al., 2005) in order to estimate ambient concentrations of air pollutants for the baseline year 2009, and for the future emission scenarios in southern Germany. The model set-up included a nesting approach, where three domains with horizontal resolution of 18 km, 6 km and 2 km were defined. The innermost domain captures local scale features including pollution in urban areas of interest. Anthropogenic emissions for the baseline year 2009 are derived from the emission inventory provided by the Netherlands Organization of Applied Scientific Research (TNO) (Denier van der Gon et al., 2010). For Germany, the TNO emissions were replaced by gridded emission data with a high spatial resolution of 1/64 x 1/64 degrees. Future air quality simulations are carried out under different emission scenarios, which reflect possible energy and climate measures in year 2030.

To investigate health effects associated with air pollution concentrations a local-scale health impact assessment (HIA) have been conducted. The simulation results for the baseline year 2009 are used to quantify present-day health burdens. Concentration-response functions (CRFs) for PM2.5 and NO2 from the WHO Health Risks of Air Pollution in Europe (HRAPIE) project were applied to population-weighted mean concentrations to estimate relative risks and hence to determine numbers of attributable deaths and associated life-years lost. Future health impacts of projected concentrations are calculated taking into account different emissions scenarios.

Recent Updates made for SMOKE version 4.0
B.H. Baek, C. Spennan, A. Eyth, G. Pouliot, C. Allen, J. Beidler, J. Vukovich

In cooperation with the U.S. EPA OAQPS and the California Air Resource Board (CARB), we have created the SMOKE version 4.0. It includes various major updates: 1) Processing pregridded global emissions outside of North America from the task force on the Hemispheric Transport of Air Pollution (HTAP) version 2 dataset through SMOKE modeling system as a part of support for the recent EPA Northern Hemispheric modeling work. 2) Enabling SMOKE modeling system to support CARB unique 20-characters Source Category Code (SCC) and Source Industry Code (SIC), 3) introducing new 12-character geographical code (GEOCODE) approach that can support more than 10 countries which is a limit to current SMOKE modeling system, and 4) various updates to support the latest U.S. EPA National Emission Inventory (NEI) 2011 modeling platform version 6.2.

Comparison of human exposure model estimates of PM2.5 exposure variability using fine-scale CMAQ simulations from the Baltimore DISCOVER-AQ evaluation

Yadong Xu, Janet M. Burke, K. Wyat Appel, Shawn J. Roselle

US EPA Office of Research and Development, National Exposure Research Laboratory, RTP, NC

Human exposure models estimate population distributions of exposure to air pollutants by combining ambient (outdoor) concentration data with human activity patterns to account for the time people spend in different locations (e.g., outdoors, indoors, in vehicles) and the various factors influencing concentrations in those locations. Although measured concentrations at monitoring network sites are typically used as inputs, exposure model results may be improved by taking advantage of the greater spatial and/or temporal resolution of air pollutant concentration fields available from air quality model simulations.

An evaluation of simulations using the Community Multiscale Air Quality (CMAQ) model version 5.1 employing 12x12 km, 4x4 km and 1x1 km horizontal grid-cell resolution performed for the Baltimore DISCOVER-AQ study in July 2011 showed significant improvement in operational model performance for the 4-km vs. 12-km simulations compared to available measurements, but a relatively small change in model performance between the 4-km and 1-km simulations. However, the spatial pattern of PM_2.5 concentrations across the Baltimore-Washington DC area differed between the 4-km and 1-km simulations, with more defined areas of high PM_2.5 concentrations for the 1-km simulation.

The Stochastic Human Exposure and Dose Simulation for Particulate Matter (SHEDS-PM) model was applied using three different hourly PM_2.5 concentrations for the Baltimore-Washington DC area (i.e., 4-km CMAQ, 1-km CMAQ, and available measurement data) to evaluate the impact of the different input concentrations on PM_2.5 exposure variability. Distributions of PM_2.5 exposures from SHEDS-PM showed the influence of capturing the fine-scale spatial and temporal variability, as well as commuting patterns, on the distributions of PM_2.5 exposures.

Organic Aerosol Sources and Partitioning in CMAQv5.2

Benjamin N. Murphy, Matthew C. Woody and Havala O. T. Pye

National Exposure Research Laboratory, US EPA, Research Triangle Park, North Carolina

We describe a major CMAQ update, available in version 5.2, which explicitly treats the semivolatile mass transfer of primary organic aerosol compounds, in agreement with available field and laboratory observations. Until this model release, CMAQ has considered these compounds to be nonvolatile, an assumption that has been shown to lead to inaccurate predictions of OA diurnal profiles during focused measurement campaigns. We also introduce an OA surrogate into the model, called Potential Combustion Secondary Organic Aerosol (pcSOA), which accounts for missing sources of primary organic vapor precursors to aerosol formation. These sources could include emissions of reactive intermediate volatility vapors, semivolatile vapor products that have been previously undetected in laboratory experiments (e.g. from sorption to chamber walls), or multigenerational aging of semivolatile vapors. We evaluate the new model against observations at several scales, showing notable improvement at single sites during short, 1-2 month simulations and slight improvement at regional/continental scales during annual simulations.

Assessing the impact of grid resolution on forward and backward sensitivity results

Melanie Fillingham & Amir Hakami (Department of Civil and Environmental Engineering,

Carleton University)

The accuracy of model predictions depends on the grid resolution, often creating a trade-off between model accuracy and computational cost. The remaining question is to what extent does model resolution impact predicted concentrations and health related burdens. There is not a single optimal resolution that best models every domain; locations with high spatial gradients (emissions, populations, concentrations, often seen in urban areas) may see more of an impact related to grid resolution than those without (often seen in rural areas). These inaccuracies arise because coarse resolutions do not adequately represent the spatial variability as averaging occurs, causing dilution within a grid. The purpose of this study is to analyze the magnitude of the impact grid resolution has on concentration and health impact predictions found from air quality models paired with epidemiological models. We do so using two different approaches; the first in which the relationship will be analyzed through forward sensitivity and the second through adjoint (backward) sensitivity.

The first approach involves the usage of the decoupled direct method (DDM) applied to CMAQ v5.0.2, in which the sensitivity between model input (grid size) and the model output (predicted concentrations) is determined through parametric differentiation with respect to the horizontal grid size in the model. The species modelled are ozone (O₃) and nitrogen dioxide (NO₂). Only the atmospheric functions that depend on the assigned grid resolution (extensive parameters), such as horizontal and vertical advection, and horizontal and vertical diffusion, are differentiated. The second approach involves simulating the health impacts, in particular mortality, related to O₃ and NO₂ exposure on a regional scale, at progressively refined resolutions. The adjoint of gas-phase CMAQ will be used, paired with Health Canada's Air Quality Benefits Assessment Tool (AQBAT), to estimate the spatiotemporal influences of emission sources on human health. Preliminary results from forward differentiation indicate that latitudinal and longitudinal resolutions may have varying impacts on predicted concentrations and health impacts.

CAMx Overview and Recent Updates

Christopher Emery, Bonyoung Koo, Gary Wilson, Greg Yarwood

Ramboll Environ, Novato, CA

An active sponsored research program at Ramboll Environ has led to a multitude of CAMx development activities and public release of two model versions since 2015. The presentation will start with a brief overview of CAMx, followed by a summary of recent specific model updates and new features, and close with plans for additional releases into 2017. Chemistry updates implemented in 2015 and 2016 include: a 1.5-dimensional Volatility Basis Set (VBS) framework for gas-aerosol partitioning and chemical aging of primary and secondary organic aerosols; the addition of oceanic halogen chemistry (I, Br, Cl) to the Carbon Bond 6 mechanism; the addition of temperature- and pressure-dependent branching ratio between ozone and organic nitrate formation pathways in the Carbon Bond 6 mechanism; and the addition of the SAPRC07TC mechanism. Other core model updates include: the addition of explicit top boundary conditions from global models, new map projections; an improved snow cover treatment; and several modifications to improve model speed. Probing Tool updates include: improved ozone source apportionment chemistry that tracks odd oxygen and nitrogen through the NOy cycle to account for NOx recycling of ozone; and the ability to define source apportionment regions by fractional cell coverage rather than whole grid cells. Plans to release a new CAMx version in late 2016 will include: a TCEQ-sponsored condensed iodine-only halogen chemistry mechanism in Carbon Bond 6: the extensions of chemical process analysis to all new chemistry mechanisms; and numerous EPA-sponsored chemistry enhancements for Carbon Bond 6, secondary organic aerosols, aqueous chemistry, and wet deposition. A CAMx release planned for 2017 will include a "Cloud-in-Grid" treatment for sub-grid shallow and deep convective transport, entrainment/detrainment and explicit cloud-scale aqueous chemistry and wet deposition.

Local to regional scale modeled wildland fire impacts on O3 and PM

Kirk R Baker, M. Woody, L. Zhou

Highly instrumented field studies provide a unique opportunity to evaluate multiple aspects of photochemical grid model representation of fire emissions, dispersion, and chemical evolution. Fuel information and burn area for a specific fire coupled with near-fire and downwind chemical measurements provides information needed to constrain model predicted fire plume transport and chemical evolution of important pollutants such as ozone (O3) and particulate matter (PM2.5) that have deleterious health effects. Most local to regional scale field campaigns to date have made relatively few transects through plumes from well characterized fire events (e.g. fuel type, area burned, etc.). While more comprehensive field studies are being planned (FASMEE, FIREX, and FIREChem), existing measurement data is used from multiple field campaigns and routine surface networks to provide some indication about how a regulatory modeling system captures fire impacts on O3 and PM2.5. Comparison of model estimates with large fire impacts against routine surface measurements at rural locations in regional proximity to the 2011 Wallow and Flint Hills fires suggest the modeling system (BlueSky, WRF, SMOKE, and CMAQ) is over-estimating hourly O3 and daily average PM2.5 organic aerosol. Sensitivity simulations where solar radiation and photolysis rates are more aggressively attenuated by fire aerosol and also treating PM2.5 organic aerosol as semi-volatile tended to reduce fire impacts but not ameliorate the over prediction bias. Finer resolution was used to assess primary and secondary impacts from the 2013 Rim fire (4 km) and 2013 agricultural fires (2 km). Aircraft measurements made downwind of the Rim fire in central California and agricultural fires in the Pacific Northwest provide spares data but could help toward improvement of model representation of local to regional scale plume transport and dispersion. Hypothetical fire simulations have been performed using this modeling system to support site and time period selection for future planned field studies that are part of the Fire and Smoke Model Evaluation Experiment (FASMEE) program and provide an illustrative assessment of potential O3 and PM impacts due to planned prescribed burns in the southeast and western U.S.

Development and Applications of Next-Generation Integrated Air Quality Decision Support System (ABaCAS)

Carey Jang, Office of Air Quality Planning and Standards, USEPA
Zac Adelman, Dongmei Yang, Institute of Environment, Univ. of North Carolina at Chapel Hill
Yun Zhu, South China Univ. of Technology, China
Shuxiao Wang, Tsinghua Univ., China

A series of collaborative efforts in the development of a next-generation air quality decision support system, or "Air Benefit and Cost and Attainment Assessment System" (ABaCAS), by a team of U.S. and international scientists have been undertaken since 2013. The objective of this ABaCAS system is to provide scientists and policy makers with a user-friendly framework for conducting integrated assessments of air pollution emissions control cost and their associated air quality, health and economic benefits and attainment goals. The "ABaCAS" system includes five key components: (1) Streamlined edition for integrated cost/benefit and attainment assessment for policy analysis (ABaCAS-SE); (2) International control cost estimate tool (ICET); (3) Real-time air quality response to emissions control tool (RSM/CMAQ); (4) Air quality attainment assessment tool (SMAT-CE); (5) Health and economic benefit tool (BenMAP-CE). A series of ABaCAS pilot applications over the Yangtze River Delta (YRD)/Shanghai region and the Pearl River Delta (PRD)/Guangzhou region in China have been undertaken to conduct assessment of emissions control strategies and their air quality, health, economic and air quality attainment benefits. These ABaCAS case studies and ongoing efforts in expanding applications to USA and other Asian countries will be presented. Further details of the ABaCAS system can be found and the ABaCAS software package can also be freely downloaded at "abacas-dss.com".

Assessment of Air Quality Impacts from the 2013 Rim Fire

Matthew Woody and Kirk Baker

Wildfires account for a significant fraction of PM_2.5 emissions in the U.S., the majority of which are organic aerosols. This work aims to quantify modeled impacts of wildfires, specifically the 2013 Rim Fire, and focuses on how recent organic aerosol updates in CMAQ v5.2 effect biomass burning organic aerosol predictions (i.e. non-volatile vs. semi-volatile primary organic aerosol emissions). We also leverage instrumented field campaigns collected on-board mobile platforms (SEAC4RS, AJAX) and traditional ground-based measurements collected at routine monitors to evaluate model performance during this event. Finally, sensitivity analysis is performed to examine what further model updates/research are needed to more accurately model the impacts from biomass burning events (e.g. volatility distribution of biomass burning primary organic aerosols, biomass burning specific organic aerosol aging).

Evaluation of a pending upgrade of the CTM of NAQFC from CMAQ version 4.6 to 5.0.2 together with a refined treatment to initialize wildfires-related PM

Pius Lee1, Jeff McQueen2, Ivanka Stajner3, Li Pan1,4, Jianping Huang2,5, Daniel Tong1,4,6, Ho-Chun Huang2,5, Hyun-Cheol Kim1,4, Perry Shafran2,5, and Sikchya Upadhayay3,7

1. NOAA Air Resources Laboratory, 5830 University Research Court, College Park, MD 20740
2. NOAA National Centers for Environmental Prediction, 5830 University Research Court, College Park, MD
3. NOAA National Weather Service, Office of Science and Technology Integration, Silver Spring, MD 20910
4. Cooperative Institute for Climate and Satellites, University of Maryland, College Park, MD 20740
5. IMSG, Rockville MD 20852
6. Center for Spatial Information Science and Systems, George Mason University, Fairfax, Virginia, VA 22030
7. Syneren Technologies, Arlington, VA 22201

The National Air Quality Forecasting Capability (NAQFC) is being prepared for a major upgrade for various modeling components. Upgrades includes the meteorological driver, emission projection based on the latest U.S. EPA NEI data, methodologies in capturing intermittent sources such as from wildfires and wind-blown dust, and replacement of NAQFC's CTM from its current EPA released version 4.6 to 5.0.2. This presentation focuses on the CTM and the initialization of fine particulate (PM2.5) fields due to wildfire originated within the NAQFC model domain. Some of the other aforementioned upgrades are discussed in other presentations in this conference. In February 2016, NAQFC has begun to provide PM2.5 surface concentration forecast to the nation twice daily with a 48 h forecast length. Forecasting PM2.5 has been a challenge as NAQFC exhibits notoriously persistent seasonal biases with overestimation in the winter and underestimation in the summer. Moreover, from epideminology studies inhalation of PM2.5 is by far more detrimental to the human health than that by ozone. With the improved aerosol science such as the explicit modeling of mineral species and the production rates of secondary organic aerosols in CMAQ version 5.0 or newer, the pending CTM upgrade in NAQFC leverages those advances. We will show a few studies when and where those upgrades were believed to be helpful. Another aspect to improve PM2.5 forecast is to refine the initialization of wildfire-related PM2.5 fields. NAQFC started to use a pre-analysis model simulation to take advantage of the completed "after-the-fact" wildfire report of yesterday to reconstruct wildfire as reported and feed that to CMAQ to calculate the PM fields in analysis or hind-cast mode for one day to retain the fire-related PM2.5 field for initialization of the PM2.5 fields by ingesting that to start NAQFC's 48 h forecast. The actual fire report helped PM2.5 forecast considerably during the fire seasons. We will share forecast accuracy statistics across various seasons and specific sensitivity cases.

STILT-ASP: A Trajectory-Based Modeling Tool for Assessing the Impacts of Biomass Burning on Air Quality

C. M. Brodowski, M. J. Alvarado, C. R. Lonsdale, J.M. Henderson, J.C. Lin, A. K. Kochanski

The transport of ozone, PM2.5, and their precursors from local and remote wildfires can lead to poor air quality in excess of the National Ambient Air Quality (NAAQS) standards. We have developed a trajectory-based modeling tool that can assess the impact of wildfires on ozone, PM2.5, and other air pollutants at specific receptors. The tool, called STILT-ASP, combines the Stochastic Time Inverted Lagrangian Transport (STILT) model with AER's Aerosol Simulation Program (ASP). STILT is a Lagrangian particle dispersion model that has been used to assess the impacts of different emission sources on measurements of long-lived pollutants, such as greenhouse gases. ASP simulates the detailed gas-phase, aerosol-phase, and heterogeneous chemistry of young biomass-burning smoke plumes. STILT-ASP simulates over 600 gas-phase species using more then 1500 chemical reactions and uses the Volatility Basis Set approach to simulate the chemical evolution of organic aerosol. The source code and User's Guide for STILT-ASP, as well as a pre-compiled version for Windows that includes a Graphical User Interface (GUI), are available from AER.

Here we demonstrate the use of STILT-ASP to assess the impact of wildfires on urban ozone concentrations during two high ozone events: one in the Austin-Round Rock metropolitan area in September 2011 and one in the Houston metropolitan area in late August 2011. We will also show the results of initial comparisons of STILT-ASP with data from the 2013 NOAA Southeast Nexus (SENEX) campaign, as well as comparisons of STILT-ASP and CMAQ simulations at different receptor locations during this campaign.

Aerosol Assimilation Based on NCEPs GSI using Surface PM2.5 and Satellite AOD

Youhua Tang^1,2, Tianfeng Chai^1,2, Li Pan^1,2, Pius Lee¹, Mariusz Pagowski^4,5, Barry Baker^1,2, Daniel Tong^1,2,3, and Hyun-Cheol Kim^1,2

1.NOAA Air Resources Laboratory, 5830 University Research Court, College Park, MD 20740.2.Cooperative Institute for Climate and Satellites, University of Maryland, College Park, MD 20740.3.Center for Spatial Information Science and Systems, George Mason University, Fairfax, VA 22030.4. NOAA Earth System Research Laboratory, Boulder, CO.5. Cooperative Institute for Research in the atmosphere, Colorado State University, Fort Collins, CO

The existing National Air Quality Forecasting Capability (NAQFC) expanded to support PM2.5 prediction. PM2.5 is not a single aerosol species, but the combination of fine aerosol mass concentrations. Any error on estimating their sources and sinks could lead to underprediction or overprediction of PM2.5. In order to correct the model bias, we use data assimilation methods to adjust the model's initial condition. The existing NCEP's GSI (Gridpoint Statistical Interpolation) uses 3D-var method to adjust the initial condition. We applied this adjustment to the aerosol field generated from CMAQ (version 5.1) and tested it for summer 2011 over continental USA in 12km horizontal resolution. This assimilation was compared to the optimal interpolation (OI) method (Tang et al, 2015). Both assimilations can improve the model's PM2.5 prediction, though their effects could fade out eventually depending on locations and events. The GSI method usually yielded smoother adjustment over broader area, and the OI method can pick up some peak events over certain locations. We compared the Pros and Cons of these two methods. The issue of combining surface PM2.5 and MODIS AOD in the assimilation is also discussed.

Modeling Single Source Secondary Impacts with the Higher-Order Decoupled Direct Method of Sensitivity Analysis

Christopher Emery, Tasko Olevski, Ralph Morris

Ramboll Environ, Novato, CA

In 2015, EPA introduced the concept of Model Emission Rates for Precursors (MERPs) for secondary PM2.5 and ozone. New or modified single sources of NOx and SO2 (PM2.5) and/or NOx and VOC (ozone) with emissions below MERPs would also be assured to result in respective ambient concentration impacts below Significant Impact Levels (SILs). Sources below SILs would not need to perform any additional modeling to address PM2.5 and ozone impacts according to Prevention of Significant Deterioration (PSD) requirements. EPA has not proposed what the ozone and PM2.5 MERPs would be or how they will be calculated, but EPA has suggested approaches using photochemical grid models with Brute Force, Source Apportionment or sensitivity methods. Furthermore, published studies by EPA (Baker and Kelly, 2014; Baker et al., 2015) have provided some initial insights into modeled ozone and PM impacts from a series of large hypothetical point sources located throughout the central US.

In this project, we develop an approach using the CAMx photochemical model employing the "higher-order" Decoupled Direct Method (HDDM) of sensitivity analysis to estimate ozone impacts from various hypothetical sources located throughout different geographies of the US, the degree to which those impacts are non-linear, and whether those impacts are associated with NOx- or VOC-limited chemistry. CAMx is run on the EPA's 2011 modeling platform (12 km grid resolution) for the May-September ozone season, using "2017EH" emissions. A key product of this analysis is the development of sensitivity metrics for each source that report tons of NOx and VOC precursor per unit ozone (ppb) impact. Results will augment EPA's MERP modeling results performed to date with new information on NOx and VOC sensitivity and the conditions under which ozone impacts can be assumed to respond linearly (i.e., the extent to which existing zero-out or source apportionment results can be linearly scaled for different emission rates).

In-Line Coupling of the NMMB and CMAQ Models through NCEPs ESMF and NUOPC Framework

Barry D. Baker II, Pius Lee, Dusan Jovic, Li Pan, Youhua Tang, Daniel Tong

CMAQ is currently run operationally at NOAA in an offline mode where the meteorology is first run using the North American Mesoscale Model on the B-grid (NMMB). The output from the meteorology is then processed into CMAQ ready input using PREMAQ (Lee et al., Weather and Forecasting 2016). In this work we demonstrate CMAQ as a component in the National Unified Operational Prediction Capability (NUOPC) and Earth System Modeling Framework (ESMF). NUOPC allows CMAQ to be coupled with a variety of meteorological models such as NMMB and the soon-to-become National Weather Service (NWS) operation Next Generation Global Prediction System (NGGPS) without modification to the CMAQ code itself and minimal changes to the ESMF layer. Additionally, it gives the ability to couple with other types of models such as land surface and ocean models. ESMF allows each model to reside on their own grid. In this case NMMB is on the Arakawa B-grid and CMAQ is on the Arakawa C-grid. Currently, the data is only brought into CMAQ and so there are no direct or indirect effects being fed into the meteorology, however, in the future this feature could easily be added into the ESMF layer. A test case of a May 11, 2014 dust event is presented with a coupling timestep of five minutes.

Recent Improvements to SCICHEM and Comparison of SCICHEM Single-Source Impacts with Photochemical Grid Models

Prakash Karamchandani¹, Lynsey Parker¹, Biswanath Chowdhury², Ana Alvarez³, Greg Yarwood¹, Eladio Knipping⁴, Naresh Kumar⁵

¹Ramboll Environ, Novato, CA; ²Xator Corp., Princeton, NJ; ³Independent Consultant, Chantada, LU - 27500, Spain; ⁴EPRI, Washington, DC; ⁵EPRI, Palo Alto, CA

SCICHEM is a publicly available state-of-the-science reactive puff model for both short-range and long-range regulatory single-source applications. Puff dispersion in SCICHEM 3.0 is based on a non-steady-state Lagrangian puff model (SCIPUFF) while the chemistry modules are based on those used in the U.S. EPA Community Multiscale Air Quality (CMAQ) model. The model has been extensively tested with measurements from field studies of power plant plumes. Since the public release of SCICHEM 3.0 in July 2015, a number of revisions have been made to the model prior to the next public release in 2016, including improvements to the calculation of dry deposition velocities. We describe these improvements to SCICHEM and present results from applications of the model to calculate single-source impacts on ozone and PM_2.5 concentrations. The impacts are compared with those calculated using regional-scale photochemical grid models (PGMs), such as CMAQ and CAMx.

Construction of Multi-fan Wind Tunnel for Radionuclides Atmospheric Dispersion

Haimin Fan, Yuanwei Ma, Dezhong Wang

For research on the aerosol dispersion characteristics in the diabatic boundary layer, the wind tunnel experiment for simulating gradient temperature was introduced in this paper. The wind tunnel used is equipped with an array of small fans (8 columns 10 rows), which is whole controlled by a single computer. As the few wind tunnel have successfully modelled unstable atmospheric boundary layer because of the poor thermal conductivity of air. However, with the multi-fan wind tunnel, this work realized the vertical gradient temperature in the moderate wind speed, as well as the wind speed profile in the vertical distribution.

Different scale of eddy structures and their roles on pollutant dispersion in and over urban canopy layers
Yifan Fan1, Julian Hunt1,2,3 , Shi Yin1 and Yuguo Li1,*
1Department of Mechanical Engineering, The University of Hong Kong, Hong Kong, China
2Department of Earth Sciences, University College London, United Kingdom
3Malaysian Commonwealth Studies Centre, Cambridge, United Kingdom

Different sizes of eddy structures co-exist in and over an urban canopy layer. The turbulent plumes are important for heat transfer and pollutant dispersion where people live. To study the plumes in and over urban canopy layers, basic fluid mechanics of different category of plumes need to be understood. A study on the complex turbulent plumes in urban-like stably stratified turbulent boundary layers is prerequisite for understanding what happens in a real urban area.
Field measurements were carried out to explore the natural convective flow along a 60 meter tall building wall (it is referred to as wall flows), which helps to understand the flow at the building scale. The wall flow was found to be governed by the plume function and described by Gaussian profile across the boundary layer. Water tanks are used to simulate the city scale and regional scale flow under calm (no synoptic wind) and stable stratification conditions. The typical urban area CBL considered here has a well-defined boundary with the external rural area; the shape of the boundary may be circular, or square, or split into sub-elements. Plumes exist in the atmosphere and dominate the turbulent flow structures over the built-up urban area when the background wind speed is low compared to the convective velocity. Different sizes of plumes from city-scale, sub-city scale, and street scale to laminar plumes co-exist in and over urban canopy layers and the city shape affects the mean flow field, which have significant influences on air temperature and pollutant concentration in the urban canopy layer. This study may also indicate how urban convection affects patterns of precipitation, which can be extremely intense and inhomogeneous in these areas.
More research is needed to quantify these results and discussions between environmental scientists and planners are required, to understand how to plan cities that make best use of the physical and natural factors determining the local meteorology, and that can reduce environmental pollution.

Characterization of Traffic Emissions Exposure Metrics in the Dorm Room Inhalation to Vehicle Emissions (DRIVE) Study

Jennifer Moutinho, Donghai Liang, Rachel Golan, Chandresh Ladva, Roby Greenwald, Rodney Weber, Stefanie Sarnat, Dean Jones, Jeremy Sarnat, Armistead Russell

Detailed measurements and fine scale dispersion modeling was conducted to develop more accurate integrated or biologically-relevant metrics to assess exposure to potentially high pollutant levels of primary traffic emissions. A 13-week intensive sampling campaign was conducted at six ambient and two indoor monitoring sites surrounding a congested highway segment in Atlanta with the study area focusing on the Georgia Institute of Technology campus. Fifty-four college students living in dorms near (20 m) or far (1.4 km) from the highway were recruited for personal exposure monitor sampling and weekly biomonitoring. Traffic-related contaminant indicators selected to capture the heterogeneity of primary traffic emissions were measured at each site, including particle mass and number, elemental and organic carbon, nitrogen oxides, and carbon monoxide. RLINE was used to develop spatial concentration fields at a 250m resolution over the Atlanta area and a 25 m resolution over the area of primary exposures. Initial RLINE results were biased, due either to errors in the emissions or the model. Analysis suggests that both may be important, depending upon species. Both the measurement observations and dispersion modeling results show that the highway has a substantial impact on primary traffic pollutant (particularly elemental carbon and carbon monoxide) concentrations and capture the prominent spatial gradients across the Georgia Tech campus, though the gradients were highly species dependent. Results were further used to develop an overall indicator of exposure to traffic related emissions for use in health assessments. In addition to quantifying a multipollutant traffic exposure indicator, metabolic response was evaluated by finding elevated levels of specific metabolites in plasma samples. These results were used to identify which exposure metrics are most predictive of biologically-relevant responses to primary traffic exposures that could be used for large panel-based epidemiologic studies.

Fine-Scale WRF/Chem Simulations over the Western U.S. for the Assessment of Future Technology-Driven Air Quality

Michael Pirhalla*¹, Patrick Campbell¹, Yang Zhang¹, Fang Yan^{2, 3, 4}, Zifeng Lu^{2, 3}, David Streets^{2, 3}

¹Department of Marine, Earth, and Atmospheric Sciences, NCSU, Raleigh, NC 27695, USA

²Computation Institute, University of Chicago, Chicago, IL 60637, USA

³Energy Systems Division, Argonne National Laboratory, Argonne, IL 60439, USA

⁴Currently at the California Air Resources Board, Sacramento, CA 95814, USA

 

Email: *mpirhal@ncsu.edu

The interconnection between anthropogenic emissions and air quality changes is an important area of focus for future climate change research. Anthropogenic emissions are often one of the most uncertain input parameters for global and regional air quality modeling, as emission inventories may under- or overestimate pollutant emissions. It is already a challenge to predict the responses of Earth's climate and air quality to future emission changes, and large uncertainties in projected future emissions pose additional difficulties. In this study, the online-coupled Weather Research and Forecasting Model with Chemistry (WRF/Chem) is applied for the years of 2005 and 2050 over an area in the western U.S. covering California and a large portion of Nevada at a fine-scale resolution of 4-km, which downscales regional simulations at a 12-km resolution through a one-way nesting approach. The main objectives of this work are to evaluate the model's capability in reproducing observations and projecting changes in future air quality at 4-km, and to identify likely causes for large model biases between model simulations and observations for future model improvement. Fine-scale modeling with high-resolution emissions and topography inputs are chosen because it may more accurately represent certain chemical processes and their feedbacks to meteorology compared to coarse grid modeling. Anthropogenic emissions are processed through the Sparse Matrix Operator Kernal Emissions (SMOKE) model using the 2005 U.S. EPA National Emissions Inventories for the 2005 baseline simulation. For the future 2050 simulations, a technology-driver model (TDM) is used to generate dynamic emission growth factors, relative to the baseline year of 2005, based on projected socioeconomic and technologically-driven changes under the Intergovernmental Panel on Climate Change (IPCC) A1B and B2 scenarios. Model performance for 2005 is being evaluated against a variety of surface and satellite-based meteorological and air quality observations to assess the credibility of the 2005 baseline simulations. Model biases in simulated O₃ or PM_2.5 will be identified to suggest areas of future model improvement. Future changes in key gas and aerosol species (e.g., NO_x, CO, O₃, and PM_2.5) by 2050 will also be assessed. Compared to results at 12- and 36-km, our preliminary results show that the finer resolution outputs can better represent the pollution interactions with topography, local effects of emission sources, and certain chemical processes, such as O₃ and PM formation. The high spatial detail of these simulations illustrates the benefit of fine-scale model applications, which is of critical importance in the development of future anthropogenic emission-control strategies.

Use of CMAQ for the 2011 National Air Toxics Assessment (NATA)

Madeleine Strum, Sharon Phillips, James Thurman, Rich Scheffe, Alison Eyth, Ted Palma, Mark Morris, Rich Cook

EPA released the 2011 National Air Toxics Assessment results in December 2015. NATA provides census tract-level cancer and noncancer risks from inhalation of air toxics. Risk results were based on modeled ambient concentrations using CMAQ and AERMOD models. A hybrid approach, combining the results from CMAQ and AERMOD annual concentrations was used for 40 of the 180 pollutants modeled, including the pollutant with the highest national cancer risk - formaldehyde. Source attribution was done primarily through AERMOD, however CMAQ zero out runs were used determine the fires and biogenic contribution to the primary emissions. CMAQ also allowed the attribution to primary and secondary contribution to formaldehyde, acetaldehyde and acrolein. A map application that displays risks, ambient concentrations emissions and ambient data was developed to facilitate the key uses of NATA, which are to prioritize sources, pollutants and areas for further study. The 2011 NATA methods and results will be discussed.

Modeling prescribed fire impacts on local to regional air quality and potential climate effects
Luxi Zhou, Kirk Baker, Sergey Napelenok

Biomass burning, including wildfires and prescribed burns, are of increasing concern due to the potential impacts on ambient air quality. The direct and indirect radiative forcings associated the particulate matter from biomass burning are also raising questions regarding the potential link between wildfires and climate change. A photochemical grid model CMAQ is used to conduct 4 and 2 km grid resolution simulations of prescribed burning experiments in southeast Washington state and western Idaho state in summer 2013. The ground and airborne measurements from the field experiment are used to evaluate the model performance in capturing surface and aloft impacts from the burning events. Source apportionment and source apportionment analysis of downwind ozone and particulate matter is conducted to quantify the impacts of experiments on ambient air quality. With the latest improvements in aerosol and cloud schemes in the model, this study also investigates the potential perturbation to surface energy balance as well as cloud formation.

Effect of global emissions on photochemical air quality in the Lower Fraser Valley Canada

Christian Reuten, Zahra Hosseini, Golnoosh Bizhani, Jeff Lundgren

Studies have shown that emissions transported over continental scales can have significant impact on local air quality. The scale of this influence is often too large to practically include explicitly within a typical 36km parent CMAQ domain. Using global scale chemical transport air quality models to provide more realistic non-uniform and transient lateral boundary conditions for regional models has been shown to improve model predictions. In this work, global air quality model GEOS-Chem was ran at 0.5 x 0.666 resolution for North America and the results are used as initial and boundary conditions for CMAQ to obtain concentrations in the Lower Fraser Valley region. CMAQ results are compared with the results of a CMAQ simulation for the same region using global average values (RWDI, 2015). GEOS-Chem and CMAQ results are compared with observations at a few representative monitoring stations.

Highlights from the Third Phase of the Air Quality Model Evaluation International Initiative (AQMEII3)

Christian Hogrefe¹, Stefano Galmarini², Efisio Solazzo², Peng Liu¹, George Pouliot¹, Rohit Mathur¹, Shawn Roselle¹, and AQMEII3 Modeling Groups

 

¹Computational Exposure Division, National Exposure Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, NC 27711, USA

²European Commission Joint Research Centre, Ispra, Italy³

We present highlights of the results obtained in the third phase of the Air Quality Model Evaluation International Initiative (AQMEII3). Activities in AQMEII3 were focused on evaluating the performance of global, hemispheric and regional modeling systems over Europe and North America and on comparing source / receptor relationships simulated through perturbation experiments reflecting emission changes in upwind continents. The work performed under AQMEII3 was coordinated with ongoing work under the Task Force on Hemispheric Transport of Air Pollution (TF-HTAP) framework. The list of highlights to be presented includes the following:

• A multi-model analysis of tracers of lower tropospheric, free tropospheric, and stratospheric ozone concentrations specified at the lateral boundaries of regional-scale simulations over North America
•  An analysis of four annual CMAQ simulations over North America using boundary conditions derived from four different global/hemispheric models
• A multi-model analysis of the effects of perturbed emissions and boundary conditions simulated by regional models participating in AQMEII3
• A comparison of the impacts of perturbed emissions in East Asia on air quality over North America simulated by global and linked global/regional models.

Relative impact of the projected emissions from industry and transportation on regional air quality in Ontario

L. Huang, A. Chtcherbakov, J. Alvarado-Melgar, Y. Hall

In the past decade, there has been a noticeable decreasing trend in overall anthropogenic emissions in Ontario, Canada. However, the degree of reduction has varied among different species and emission source types. That has resulted in changes in relative contributions of different emission source types to ambient air concentrations for given species spatially.

 This study investigates the relative impact on the ambient air quality in Ontario due to point and transportation sources. The base year for the analysis was selected as 2025, due to the availability of projected emission inventories by US EPA and Environment Canada.  Analysis of the spatial distribution in projected for 2025 Ontario emissions for SO2, NOx, and PM2.5 was performed for each by source type and showed considerable differences among the aforementioned emitting species.

 The relative impact analysis was then carried out based on a series of CMAQ runs performed with the base year (i.e. 2025) emission and several emission scenarios. The emission scenarios were prepared using the base year emission and assuming zero emission rates in Ontario for 1) all anthropogenic emissions, 2) on-road source emissions, 3) off-road source emissions and 4) point source reported to the National Pollutant Release Inventory (NPRI), individually. Results of the gridded and the population-weighted annual concentrations through all scenarios will be presented, as well as the estimated contributions by source type and the associated spatial variability across Ontario in 2025.

Decadal Application of WRF/Chem under Current and Future Climate/Emission Scenarios: Part II. Impact of Projected Climate and Emission Changes on Future Air Quality over the U.S.

Chinmay Jena, Yang Zhang, Kai Wang, and Patrick Campbell

Department of Marine, Earth, and Atmospheric Sciences, NCSU, Raleigh, NC 27695, USA

Fang Yan*, Zifeng Lu, and David Streets

Computation Institute, University of Chicago, Chicago, IL 60637, USA

Energy Systems Division, Argonne National Laboratory, Argonne, IL 60439, USA

*Currently at the California Air Resources Board, Sacramento, CA 95814

Rapid increases in modern society activities, economics, and population are becoming complex and vulnerable to changes in weather, climate, and air quality. Emission projections are critical elements to the understanding of future climate impacts on regional air quality. In the past, emission projections have often been estimated by combining fuel consumption with an averaged emission factor that represents the whole emitting sector while neglecting relationship between socioeconomic factors and projected technology changes. Following a comprehensive evaluation of the regional Weather research and forecasting model with Chemistry (WRF/Chem v 3.7), in this Part II work, we use the emissions projected by the newly-developed Technology Driver Model (TDM) under two IPCC scenarios (i.e., A1B and B2) and WRF/Chem v 3.7 to investigate the impact of both climate change and emission projections on the future regional air quality over the continental U.S. (CONUS) during present (2001 - 2010) and future (2046 - 2055) decades. The TDM takes into account the impacts of socio-economic factors, technological change, and federal and regional environmental policies and overcomes to some extent the aforementioned limitations in projecting emissions. Dynamical downscaling method is applied to link the global Community Earth System Model/Community Atmosphere Model (CESM/CAM5) with WRF-Chem. Results show that the temperature at 2-m (up to 3oC), planetary boundary layer height, wind speed at 10-m, and precipitation will increase for the future decade under both TDM A1B and B2 scenarios. Under the TDM/A1B scenario, the projected future surface ozone (O3) levels will increase across the CONUS region, ranging from 2 to 7 ppbv due mainly to increased levels of greenhouse gases, temperature, and biogenic emissions. Under the TDM/B2 scenario, the large reductions of O3 precursor emissions (e.g., nitrogen oxides (NOx), carbon monoxide, and volatile organic compounds (VOC)) lead to the overall decrease in surface O3 concentrations compared to the current decade, but surface O3 concentrations increase at urban centers due to the dis-benefit of NOx emission reduction over areas with VOC-limited O3 chemistry and weakened NO titration to O3 formation. The reduced primary anthropogenic emissions and the increase in precipitation lead to decreases in the surface PM2.5 concentrations under both scenarios. The results of this study provide important information about the impacts of projected climate change and emissions on air quality and will be useful for policy makers to implement integrated strategies to control anthropogenic emissions and mitigate adverse climate change. 

Prediction of harmful water quality parameters combining weather, air quality and ecosystem models with in-situ measurements

Catherine Nowakowski¹, Marina Astitha¹, Valerie Garcia², Penny Vlahos³, Ellen Cooter², Chunling Tang², Brian Hinckley²

¹ Civil and Environmental Engineering, University of Connecticut, Storrs, CT, USA

² National Exposure Research Laboratory, Office of Research and Development, US Environmental Protection Agency, Research Triangle Park, NC 27711, USA

³Marine Sciences, University of Connecticut, Groton, CT, USA

The ability to predict water quality in lakes is important since lakes are sources of water for agriculture, drinking, and recreational uses. Lakes are also home to a dynamic ecosystem of lacustrine wetlands and deep waters. They are sensitive to pH changes and are dependent on dissolved oxygen and nutrient levels. Even small changes in these variables can have drastic impacts on a lake's biota, physiochemical state, and hydrology. To date, numerical prediction models do not dynamically describe the entirety of air-water-soil interactions. Nevertheless, there is abundance of data from observations and weather, air quality, agroecosystem and hydrological models. In this study we demonstrate how modeled and observed variables can be used to identify algal blooms using chlorophyll-a concentrations as proxies. The area of focus is Lake Erie because of its history of excessive algal blooms and the abundance of available data for the period 2002-2012. We are using weather variables from the WRF model, hydrological variables from the VIC model and variables from the Community Multiscale Air Quality Bidirectional (CMAQ-Bidi) modeling system which includes agricultural practices. Both VIC and CMAQ-Bidi were run with the same WRF meteorology, land use and emissions files to retain consistency among the models. We built various regression models and examine the relationships between factors such as meteorology, land use, hydrology, applied and deposited nutrients (nitrogen and phosphorus) and chlorophyll-a concentrations taken by the Lake Erie Forage Task Group. Through a multivariate analysis we determine the significance of these predictors (simulated variables) on chlorophyll-a concentrations, an indicator of productivity, within Lake Erie.

Using a simple operational global aerosol model to provide dynamic chemical boundary condition for dust to the operational NAQFC

Youhua Tang^1,2 , Li Pan^1,2, Sarah Lu^3,4, Pius Lee¹ , Daniel Tong^1,2,5, Jun Wang³, Hyun-Cheol Kim^1,2, and Jeff McQueen³

^1.NOAA Air Resources Laboratory, College Park, MD 20740, 2. Cooperative Institute for Climate and Satellites, University of Maryland, College Park, MD 20740 3. NOAA NCEP, College Park, MD 20740. 4. University at Albany, State University of New York, Albany, NY 12222 5. Center for Spatial Information Science and Systems, George Mason University, Fairfax, VA 22030

The NOAA Environmental Modeling System (NEMS) Global Forecasting System (GFS) Aerosol Component (NGAC) has been operationally providing 120 h dust aerosol forecast initialized daily at 00 UTC since summer 2015 at 0.5^o x 0.5^o resolution. The operational National Air Quality Forecasting Capability (NAQFC) over the contiguous U.S. (CONUS) has been using the NGAC predicted dust fields with 5 sectional bins to derive dynamic lateral boundary conditions for dust aerosols. Dust intrusions into the CONUS domain have often been inferred from ground- and space-based observations through the western boundary from sources in Asia and the southern and eastern boundaries from sources in and around the Sahara Desert, in spring and summer, respectively.  Although NGAC accounts for hydrophilic and hydrophobic as well as sea salt particles, we only use the dust species as inert tracers. The NGAC captured dust sources mitigate in part the perennial under-estimation of surface fine particulate (PM_2.5) predicted by the NAQFC in summer. We will present regional verification statistics over areas subjected to dust intrusion in terms of both seasonal characteristics as well as spatial distribution of elevated surface concentration of PM_2.5 for sensitivity case studies when compared to forecasts that had not accounted for these outside dust sources. 

 

Prediction of fine particulate matter (PM2.5) by the National Air Quality Forecast Capability
Ivanka Stajner (1), Jeff McQueen(2), Pius Lee(3), Jianping Huang (2,4), Li Pan (3,5), Ho-Chun Huang (2,4)Daniel Tong (3,5), Ariel Stein(3), Phil Dickerson(6), Sikchya Upadhayay (1,7)

(1) NOAA NWS/OSTI
(2) NOAA NWS/NCEP
(3) NOAA ARL
(4) IMSG
(5) CICS, University of Maryland
(6) EPA
(7) Syneren Technologies

NOAA's National Air Quality Forecast Capability (NAQFC) provides operational air quality predictions for ozone and wildfire smoke over the United States (U.S.) and predictions of airborne dust over the contiguous 48 states at http://airquality.weather.gov. Predictions of fine particulate matter (PM2.5) became publicly available in February 2016. Ozone and PM2.5 predictions are produced using a system operationally linking Community Model for Air Quality (CMAQ) with meteorological inputs from the North American Mesoscale Forecast System (NAM). Smoke and dust predictions are produced by NOAA s HYSPLIT model.
Recent updates to NAQFC predictions have focused on public release of fine particulate matter PM2.5 predictions from the model and bias-corrected PM2.5 prediction using an analog ensemble algorithm. Some of previous seasonal biases in PM2.5 prediction were reduced by suppression of soil emissions in wintertime and the addition of intermittent wildfire smoke and dust emissions. Nevertheless, seasonal biases and biases in the diurnal cycle of PM2.5 remained substantial, so a new bias correction procedure based on an analog ensemble approach was introduced. This upgrade also included lateral boundary conditions from NOAA's global dust predictions and an increased number of model layers.
Current NAQFC efforts are focused on updating CMAQ to version 5.0.2 and improving PM2.5 predictions. Testing includes wildfire smoke emissions from a newer version of USFS BlueSky system and a new configuration of NAQFC CMAQ-system to re-run previous 24 hours, during which wildfires were observed from satellites, to better represent wildfire emissions prior to beginning 48 hour predictions. Anthropogenic emissions updates are focused on updating NOx emissions using recent observed trends. Bias correction for PM2.5 is being refined to better capture day-to-day variability in PM2.5. Overview of recent updates and evaluation of predictions will be presented.

Air quality real-time forecast before and during the G-20 Summit 2016 in Hangzhou with the WRF-CMAQ and WRF/Chem systems: Evaluation and Emission Reduction Effects
Shaocai Yu 1,2, Pengfei Li 1,2, Liqiang Wang 1,2, Weiping Liu 1,2, Yang Zhang 3, David Wong 4, Kiran Alapaty 5, Jon Pleim 4 and Rohit Mathur 4
1 Research Center for Air Pollution and Health,2 Key Laboratory of Environmental Remediation and Ecological Health, Ministry of Education, College of Environmental and Resource Sciences, Zhejiang University, Hangzhou, Zhejiang 310058, P.R. China.
3 Department of Marine, Earth, and Atmospheric Sciences, North Carolina State University, Raleigh, NC 27695, USA
4 Computational Exposure Division, National Exposure Research Laboratory, U.S. Environmental Protection Agency (EPA), RTP, NC 27711, USA.
5 Systems Exposure Division, National Exposure Research Laboratory, U.S. EPA, RTP, NC 27711, USA

The 2016 G-20 Hangzhou summit, the eleventh annual meeting of the G-20 heads of government, will be held during September 3-5, 2016 in Hangzhou, China. For a successful summit, it is important to ensure good air quality. To achieve this goal, governments of Hangzhou and its surrounding provinces will enforce a series of emission reductions, such as a forced closure of major highly-polluting industries and also limiting car and construction emissions in the cities and surroundings during the 2016 G-20 Hangzhou summit. Air quality forecast systems consisting of the two-way coupled WRF-CMAQ and online-coupled WRF-Chem have been applied to forecast air quality in Hangzhou regularly. This study will present the results of real-time forecasts of air quality over eastern China using 12-km grid spacing and for Hangzhou area using 4-km grid spacing with these two modeling systems using emission inventories for base and 2016 G-20 scenarios before and during the 2016 G-20 Hangzhou summit. Evaluations of models' performance for both cases for PM2.5, PM10, O3, SO2, NO2, CO, air quality index (AQI), and aerosol optical depth (AOD) are carried out by comparing them with observations obtained from satellites, such as MODIS, and surface monitoring networks. The effects of the emission reduction efforts on expected air quality improvements during the2016 G-20 Hangzhou summit will be studied in depth.

CMAQ simulations for Ozone over Region of Great Vitoria (Brazil): influence of boundary conditions

Rizzieri Pedruzzi, Taciana T. de A. Albuquerque, Igor Baptista Araujo, Barron Henderson, Nikolle Aravanis, Erick Sperandio, Neyval C. Reis jr., Davidson M. Moreira, Jane Meri Santos, Milena Machado Melo.

The Metropolitan Region of Grande Vitoria (RMGV) is located in Espirito Santo State in Brazil's southeast. There are many emissions source of air pollutant that has a negative impact on local air quality, like industrial sites, mobile emissions, biogenic emissions and we have the impact that come from others metropolitan areas, like Sao Paulo, Belo Horizonte and Rio de Janeiro that are big urban center with a great amount of pollutant emissions located in Brazil's southeast. Due to impact on air quality that these emissions can cause, CMAQ simulations were performed over RMGV to evaluate the PM10 and O3 with different boundary conditions to understand influence on these pollutants, which come from edges and also were performed the process analysis to understand which process had greater impact on final concentrations. Were simulated four scenarios for August 2010, the first scenario (M1) using fixed, time-independent boundary conditions with zero concentration (zero) for all pollutants; a second scenario (M2) with fixed, time-independent concentration values, with average values from monitoring stations from RMGV and from Aracruz's stations on north and Anchieta's stations on south; the third scenario (M3) used boundary conditions varying with time from a previous simulation with CMAQ over a larger area, centered on RMGV; and finally, the fourth scenario (M4) using boundary conditions varying with time from simulations of global model GEOS-Chem. All scenarios used the same meteorology conditions and pollutant emissions, meteorological conditions was generated by the model WRF version 3.6.1 and pollutant emissions inventory are from the official emissions inventory of RMGV which were inputted in SMOKE 3.5.1 with a domains 61 x 79km centered on coordinates -20,25°S, -40,28° with a resolution of 1 km. The air quality simulations were made with the same SMOKE's domains, on CMAQ 5.0.1 using the CB05 and Aero6 and still analyzer CMAQ processes (PROCAN). The results were compared with the measured data in monitoring stations from RMGV.

Recent updates to the CMAQ model evaluation tools and the new AMET version 1.3

K. Wyat Appel, K.M. Foley, C. Hogrefe, R. Gilliam and S. Roselle

In the spring of 2016, the evaluation tools distributed with the CMAQ model code were updated and new tools were added to the existing set of tools. Observation data files, compatible with the AMET software, were also made available on the CMAS website for the first time with the updated tools. In addition, in the fall of 2016 a new version of the AMET software will be released, which includes the elimination of the PERL based scripts, updated functionality, various bug fixes, and efforts to streamline the installation and general overall function of the software. This presentation will discuss the updated and new CMAQ evaluation tools available, as well as discuss the major changes in AMET 1.3.

Modeled PM2.5 and O3 contribution from lateral boundary inflow and wildfires

Kirk R. Baker

Lateral boundary inflow and two specific fires from 2011 are tracked for local to regional scale contribution to ozone (O₃) and fine particulate matter (PM_2.5) using a regulatory modeling system. The modeling system was applied to track the contribution from lateral boundaries, a wildfire (Wallow), and prescribed fire (Flint Hills) using both source sensitivity and source apportionment approaches. The model estimated contributions to primary and secondary pollutants are comparable using source sensitivity (brute-force zero out) and source apportionment approaches. The modeling system tends to overestimate hourly surface O₃ at routine rural monitors in close proximity to the fires when the model predicts elevated fire impacts on O₃ and Hazard Mapping System (HMS) data indicates possible fire impact. A sensitivity simulation in which solar radiation and photolysis rates were more aggressively attenuated by aerosol in the plume reduced model O₃ but does not eliminate this bias. A comparison of model predicted daily average speciated PM_2.5 at surface rural routine network sites when the model predicts fire impacts from either of these fires shows a tendency toward overestimation of PM_2.5 organic aerosol in close proximity to these fires. The spatial pattern of monthly average lateral boundary inflow contribution to O3 and PM2.5 varies by season and by lateral face. 

VERDI Visualization of Geospatial Datasets
Jo Ellen Brandmeyer and Liz Adams
UNC Institute for the Environment

We will describe the process that was used to verify the geolocation and visualization of model and observational data and areal interpolation calculations within VERDI v1.6. Test datasets were created by extracting variables from larger model output datasets. Model output provenance and the characteristics needed to specify their geolocations were investigated and verified, including the datum, projection and spheroid. We will present the workflow used to verify that data displayed in VERDI match how they appear in a GIS such as QGIS and ESRI ArcGIS, NCL, python and other tools that visualize geospatial netCDF data: http://www.unidata.ucar.edu/software/netcdf/software.html

We will describe the I/O API and netCDF test datasets that were produced by SMOKE, WRF, MCIP, CMAQ, CAMx and MPAS as well as observational datasets. We will discuss the typical projections (e.g., Lambert Conformal, Polar Stereographic, and Mercator) written by these types of models. We will share tips and workarounds for loading and visualizing datasets in VERDI, GIS, and netCDF data visualization tools.

EXPLORING PARALLEL PROCESSING OPPORTUNITIES IN AERMOD

George Delic, HiPERiSM Consulting, LLC, P.O. Box 569, Chapel Hill, NC 27514

HiPERiSM Consulting, LLC, has investigated opportunities for thread parallelism in the AERMOD regulatory model developed by the U.S. EPA Office of Air Quality Planning and Standards (OAQPS). This is within the scope of HiPERiSM's mission to develop (or enhance) software and improve performance on current and future computers for legacy Air Quality Models. This report uncovers opportunities for AERMOD parallel execution on either host CPU or Intel Phi Many Integrated Core (MIC) architectures. Parallel scaling is evident from a simple investigation. However, a full thread parallel implementation would require substantial computer code restructuring with a view to streamlining for a tighter coupling between coding practice and computer processor capability. In the case of AERMOD, several obstacles need to be overcome. The first of these is a call tree of 3 to 4 levels deep in parallelizable loops. The second is that nearly all variables are global, and those in parallel regions would need to be privatized when written to memory, The third is that the target loops contain I/O which is best hoisted out of any thread parallel region to avoid synchronization costs. The purpose of this presentation is to provide quantitative evidence of the measured wall clock time to demonstrate the potential performance improvements. Results with the serial version of AERMOD-HPC are presented on Intel processors with both host CPU and MIC 64-bit Linux operating systems for an input data set with 1001 sources and 916 receptors. The loop structure in AERMOD has 27 source loops and 38 receptor loops (typically nested within the source loops). A compute intensive receptor loop is one target for thread parallelization. To test potential effectiveness of such a strategy, the input data stream was systematically divided into smaller subsets of receptors. These subsets were submitted as multiple concurrent executions in a task-farming fashion to all available cores on either a multicore host or many core MIC processors. In both cases wall-clock time was drastically reduced by adding more cores. Therefore, the best scope for performance gains is on MIC processors. With the availability of such processors the multiple levels of parallelism latent within AERMOD provide ample opportunity for parallel performance scaling.

Continuous, Near Real-Time Application and Evaluation of WRF-CMAQ

Brian Eder, Rob Gilliam, George Pouliot

Historically, the EPA's Computational Exposure Division (formerly the Atmospheric Modeling and Analysis Division) has evaluated retrospective, often annual length, simulations of WRF-CMAQ, summarizing the performance using monthly or seasonal statistical summaries. Although informative, such an approach often masks finer scale temporal (i.e., diurnal to weekly) and spatial (mesoscale to synoptic) variability that greatly impacts the simulation of the atmosphere and hence air quality. In order to maintain WRF-CMAQ's state-of-the-science status, as well as its ability to address emerging Agency needs, it is crucial that innovative evaluation approaches are developed and utilized that will allow for more rapid testing and hence more efficient evolution of the modeling system's science. Accordingly, the Division began running WRF-CMAQ continuously and in near real-time (CMAQ-NRT) in 2014, which has allowed for immediate and ongoing analysis, thereby facilitating model evaluation (both performance and diagnostic) of PM2.5 (mass only) and O3 concentration. Observations obtained from EPA's Air Quality System (AQS) are used in the evaluation incorporating roughly 450 PM2.5 mass and 900 O3 monitors.  Examples of recent results will be presented, using a variety of statistical and visualization tools.

Lateral Boundary Contributions to Ozone Differ using Inert or Reactive Tracers

Uarporn Nopmongcol, Zhen Liu, Till Stoeckenius, Greg Yarwood  

The third phase of the Air Quality Model Evaluation International Initiative (AQMEII-3) organized by the European Commission's Joint Research Centre (JRC) focuses on applying regional scale atmospheric models jointly with global models to examine the contribution of inter-continental transport to regional air quality and its uncertainties. Results from AQMEII-3 are expected to inform policy regarding international emission control programs and approaches to dealing with the impacts of emissions on human health, ecosystems and climate change. In AQMEII-3, modelers were asked to quantify contributions of international O₃ transport by adding inert O₃ tracers to their photochemical model simulations. Inert tracers are not destroyed by chemical reactions and therefore will tend to over-state O₃ transport, e.g., the influence of Asian O₃ on North America, or North American O₃ on Europe. We conducted CAMx simulations over North America using the AQMEII-3 datasets and added both inert and reactive tracers to track O₃ introduced by the CAMx boundary conditions (BCs) from different lateral boundary and altitude segments. This study quantifies how the choice of methodology, i.e. Inert or reactive tracers, influences boundary contributions to daily maximum 8-hour ozone concentrations in AQMEII-3. 

Dynamic evaluation of CMAQ wet deposition estimates: Observed vs modeled trends from 2002-2012
Kristen M. Foley, Jesse O. Bash, Donna Schwede, Joseph Pinto
US EPA

Recent updates in the CMAQ system have led to improved seasonal and annual total estimates of wet deposition compared to previous model versions. However errors in modeled precipitation and in emission inputs continue to lead to bias and error in the simulation of wet deposition. We present an approach to bias-correct CMAQ model output over the contiguous United States (CONUS) using observation-based gridded precipitation data generated by the Parameter-elevation Regressions on Independent Slopes Model (PRISM) and wet deposition observations from the National Atmospheric Deposition Program/National Trends Network (NADP/NTN). A cross validation analysis of model output for 2002 through 2012, shows that the adjusted annual total wet deposition model values for nitrate, ammonium and sulfate have less bias and are more highly correlated with observed wet deposition values than the base model output without adjustment. A trend analysis shows that temporal trends in the observed wet deposition totals during the eleven year time series are captured well by the adjusted model predictions across the entire CONUS. Spatial maps of the model predicted trends quantify the steady decrease in wet deposition of nitrate and sulfate due to decreasing emissions in the eastern half of the US, particularly in northeastern states where the average decrease was 30% for nitrate and 50% for sulfate. In the West, wet deposition of these species has remained relatively constant. Trends in wet deposition of ammonium are more spatially and temporally heterogeneous, with some increasing trends in the Great Plains and flat or slightly decreasing trends in the South.

Evaluation of PM2.5 concentration in Yunlin County in Taiwan

Chia-Hwa Hsu and Fang-Yi Cheng

Department of Atmospheric Sciences, National Central University, Taiwan

Yunlin County is located in central-southern portion of western Taiwan. The local industrial emissions (Maliao industry), vehicle exhausts, and burnings of agriculture wastes all contribute to the poorer air quality in Yunlin. Besides, the pollutant from nearby power plants, Taichung metropolitan area, and Changhua industrial park also contribute to the PM_2.5 concentration in Yunlin County. From 2014 to 2015, based on the surface air monitoring stations, the averaged PM_2.5 concentration in Yunlin County is ranked as the top highest stations in Taiwan. The objective of this study is to evaluate the CMAQ predicted PM_2.5 concentration in Taiwan.

        The preliminary evaluation of the CMAQ simulation result shows an underestimation of the sulfate and organic carbon in Yunlin County of Taiwan. The model error could be due to the meteorological uncertainty, underestimation of emission inventory and the insufficiency of the chemical and aerosol mechanism. Several sensitivity tests have been performed to figure out the possible reason for the underestimation including (1) update of the emission inventory; (2) comparison of CMAQ version between v4.7.1 and v5.1; (3) comparison of CMAQ simulations using different biogenic emission inputs (MEGAN and BEIS). The detail of this study will be presented in the workshop.

Data Fusion of Air Quality Model Simulations and Ground-based Observations: Application over North Carolina, USA

Ran Huang, Xinxin Zhai, Cesunica E. Ivey, Mariel D. Friberg, Xuefei Hu, Yang Liu, James A. Mulholland, Armistead G. Russell 

Data fusion method is to blend ground-based observations and simulated data from the Community Multiscale Air Quality (CMAQ) modeling system. Spatial kriging is used to provide one estimate of pollutant concentrations, while CMAQ fields, adjusted to spatially and temporally better align with observations, provide a second set of fields.  The two fields are then merged based on correlation analysis.  This leads to fields that capture the spatial and temporal information provided by the air quality model, as well as the finer temporal scale (i.e. daily) variations provided by the observations, and also decreases model biases. Here, the approach is applied for the time period from 2006 to 2008 over North Carolina (USA) for 24hr averages of PM_2.5 total mass, five major particulate species (OC, EC, SO₄, NO₃, and NH₄), and three gaseous pollutants (CO, NO_x, NO₂) that are related to emissions from mobile sources. The correlations between CMAQ and observations and data fusion and observation improve between 23% to 263% for all the species, the biases are reduced to almost zero at monitor locations, and the errors (as measured by RMSE) decrease from 33% to 78%. The results show an improvement in estimation of spatial and temporal variability, which is important in health studies. Comparison of PM_2.5 total mass concentration results with results from a satellite-retrieved aerosol optical depth (AOD) method shows good agreement between the data fusion fields and AOD-derived fields, with slightly less overall error in the data fusion results. This indicates that for the areas where monitoring data and air quality model results are not available, the AOD method is a good surrogate for generating spatial-temporal fields of air quality data.  Integrated Mobile Source Indicator method has been applied to illustrate that the data fusion fields can be used to show mobile source impacts. The data fusion application in NC shows reasonable estimation of spatial and temporal variation in air pollution, and demonstrates that the method can be used for health and planning studies.

Impact of GOES Enhanced WRF Fields on Air Quality Model Performance
Maudood N. Khan, Andrew White, Arastoo P. Biazar, and Dr. Richard T. McNider
University of Alabama in Huntsville. Huntsville, AL.

A Comprehensive Performance Evaluation of WRF/Chem version 3.7.1 over the Contiguous United States for 2008-2012

Mike Madden, Khairunnisa Yahya, Kai Wang, and Yang Zhang

Department of Marine, Earth, and Atmospheric Sciences,

North Carolina State University, Raleigh, NC

Major pollutants such as ozone (O₃) and particulate matter (PM) can cause a variety of environmental pollution issues and human health effects. Their contributions to those pollution issues and health effects have not yet been fully understood and accurately quantified.  For example, while PM exposure itself has been shown to cause numerous short- and long-term health impacts, such as hypertension, reduced blood flow, thinning of the vascular walls, and increased respiratory infections and mortality, more research is being devoted toward ailments in relation to the specific compositions of the inhaled particulates (e.g., sulfate, nitrate, ammonium, organic carbon, elemental carbon).  Analyses of PM composition, O₃, and their precursors offer clues upon their emission sources, which aids regulators to control emissions and improve air quality. To accurately perform source apportionment and simulate the health impacts of O₃ and PM, it is important to evaluate and improve the abilities of air quality models, such as the Weather Research and Forecasting Model with Chemistry (WRF/Chem) in reproducing observed PM, PM composition, O₃, and meteorological variables.

In this work, WRF/Chem version 3.7.1 is applied for a 5-year period (2008 to 2012) over the contiguous United States at a 36-km horizontal resolution to evaluate its ability in simulating major air pollutants such as O₃, PM, and PM composition and identify causes of model biases for model improvement.  Statistical evaluations using observations from multiple surface and satellite networks and multiple skill scores are conducted on various time scales (e.g., seasonal, annual, 5-year period).  Major meteorological variables such as temperature and relative humidity at 2-m, wind speed and direction at 10-m, and precipitation and air pollutants such as O₃, PM, and major PM species are evaluated.  The evaluation will also include the temporal variation of chemical species, the concentration of PM and its species, and percentage contributions of PM species to total PM concentrations at a number of representative sites with various characteristics of geography (e.g., urban, rural, coastal, mountain locations), emissions (e.g., anthropogenic vs. biogenic), and chemistry (e.g., NO_x- vs. VOC-limited O₃ and PM formation)).  Preliminary annual performance analyses for 2008-2010 show that WRF/Chem simulates meteorological variables and the concentrations of PM_2.5 well in comparison to observations and previous studies.  However, the concentrations of PM₁₀ and important PM components such as sulfate, ammonium, and organic species are underpredicted, which signifies potential roadways for model improvement.

Evaluation of a line source dispersion model, RLINE, using multi-year hourly pollution measurements in Detroit, MI.

Chad Milando, Stuart Batterman 

Recent analyses have shown that in some Midwest cities in the US, the contribution of roadway pollution is increasing relative to other sources.  By extension, the health impacts attributed to roadway pollution may therefore also be increasing. Thus, it is relevant for air pollution modeling efforts to evaluate new models and features that can enhance our current understanding of near-roadway pollutant dispersion.  The Research Line-source (RLINE) model, a recently developed air pollution dispersion model designed specifically to model near-roadway pollution, includes novel features not present in other line-source models: improved downwind and upwind dispersion algorithms, and algorithms to account for dispersion of pollutants from depressed roadways and around road-side barriers.  However, several of these features have not been evaluated in the peer-reviewed literature, and of the few studies that have evaluated the performance of RLINE, there have been limitations in the number of monitoring sites, the pollutants measured and the duration of studies.  To address these issues, we evaluate RLINE using continuous multi-year hourly measurements of CO, NO, NO₂, NO_X and PM_2.5 at several monitoring sites in Detroit, MI.  The sites include near-roadway sites and sites aimed at estimating exposure of urban populations.  The evaluation uses multi-year traffic volumes (Annual and Commercial Average Daily Traffic), emission factors (from MOVES 2014a), meteorological data (using hourly National Weather Service data and the AERMET meteorological processor), and point source emission data (from the National Emission Inventory and Michigan Air Emission Reporting System).  RLINE predictions are compared to monitored data as well as modeled results from other air pollution dispersion models (e.g., the California LINE source dispersion model, CALINE4).  A diagnostic evaluation includes spatial and temporal performance overall and during rare events that could contribute disproportionately to the overall public health burden of roadway pollution, e.g., hours with low wind speeds, hours with high measured pollution levels, and hours where the near-road monitoring stations are upwind of nearby sources.  We also investigate the influence of using various meteorological input files.  The results can be used to improve future versions of RLINE, and can help guide public health interpretations of roadway dispersion modeling efforts.  

Strong Influence of Deposition and Vertical Mixing on Secondary Organic Aerosol Concentrations in CMAQ and CAMx
Qian Shu, Barron H. Henderson

Department of Environmental Engineering Sciences, Engineering School of Sustainable Infrastructure and Environment, University of Florida, Gainesville, FL, 32611-6450, USA.

Atmospheric organic aerosols (OA), especially secondary organic aerosols (SOA) make up a majority fraction of atmospheric fine particle matters (PM), which involve global climate and human health. In order to fully understand these effects of PM, optimal computational modeling is required associated with experimental research on OA. Recently two commonly applied regulatory Air Quality Models (CMAQ: the Community Multiscale Air Quality; CAMx: the Comprehensive Air quality Model with extensions) predict very different OA concentrations, even when using identical emissions and SOA formation mechanisms over the eastern U.S. (Koo et al., 2014). These differences exist despite shared emissions and formation mechanisms revealing important uncertainties in SOA fate. This work attempts to reproduce SOA concentration biases existed in previous work and find out the underlying reason to cause the biases.

We firstly develop four case studies to reduce the simulation burden by minimizing the temporal and spatial extent needed to produce SOA concentration biases existed in previous work. Preliminary results demonstrated that the minimum simulation duration can be 1 day (Aug. 2nd) for 12km nested-grid domain covering the eastern U.S. We therefore select the case study and perform a detailed diagnostic evaluation to quantify modeled processes affecting both the SVOC (g) and aerosol (a) phases to identify causes of concentration differences between the two models. We find that differences in deposition of semi-volatile organic compounds (SVOC) and SOA contribute to significant OA biases in CMAQ relative to ambient data. Different deposition between the two models also causes similar biases for inert compounds like elemental carbon (EC). Daytime low-biases in EC were also observed in CMAQ as expected, but were offset by nighttime high-biases. Nighttime high-biases were a result of overly shallow mixing in CMAQ leading to a higher fraction of EC total atmospheric mass in the first layer. OA concentrations are also affected by vertical mixing.

Our anticipated results will help to characterize model processes in the context of SOA. We will particularly focus on deposition process that contributes to appreciable OA bias differences between CMAQ and CAMx. By isolating the process and mechanism by which modeled OA changes, we can improve the possibility for laboratory models to succeed.

Modeled Source Contributions to CO and NOy Concentrations during the DISCOVER-AQ Baltimore Field Campaign

Heather Simon1, Kirk R. Baker1, Jim Kelly1, Brian Timin1, R. Chris Owen1, Luke Valin2, Kristen Foley2, Pat Dolwick1, Norm Possiel1, James H. Crawford3
1US EPA, Office of Air Quality Planning and Standards, RTP, NC
2US EPA, National Exposure Research Laboratory, RTP, NC
3NASA Langley Research Center

Ratios of pollutants (NOx, NOy, PM2.5, BC etc) to CO are often analyzed in field data to provide insight into atmospheric sources and sinks.  Here we take a closer look at CO:NOy ratios measured during the 2011 DISCOVER-AQ field campaign in Baltimore.  The ratio of total NOy to CO varies by day as well as spatially on any given day indicating a variety of potential controlling factors. We then use results from the CMAQ photochemical model instrumented with source apportionment capabilities (ISAM) to further evaluate source contributions to both of these pollutants during the field campaign and to understand how varying source contributions may impact the ratio of these pollutants in the ambient atmosphere.  We use a combination of the modeled source apportionment results and relationships of observed ratios with other measured parameters to investigate the drivers for variability in the CO:NOy ratio. 

In-depth examination of emissions inventories to support EPA evaluation of modeled ambient nitrogen oxides (NOx and NOy).

Claudia Toro, Megan Beardsley, Kristen Foley, Alison Eyth, Kirk Baker, Sharon Phillips, Heather Simon

This work presents a detailed investigation of gridded hourly NOx emissions inventories used in air quality models such as CMAQ to provide insight into potential causes driving discrepancies between modeled and observed NOx and NOy ambient mixing ratios. In this context, we identified specific grid cells with periods of high NOx or NOy bias and extracted SMOKE emissions output by mobile source type and emission process. We then explored temporal and spatial allocation of activity inputs for onroad and nonroad mobile sources and other NOx relevant sources. Source apportionment analyses were also performed to further understand the contribution of specific source groups. The outcome of this study will shed light on the role of emissions inventories in the bias in modeled NOx and NOy.

Constraining Biogenic Secondary Organic Aerosol (BSOA) production in CMAQ during the SOAS Campaign

Momei Qin, PETROS VASILAKOS, Christopher Boyd, Nga Lee Ng, Armistead G. Russell, Athanasios Nenes

It has been recently found that biogenic emissions and anthropogenic pollution interact in order to produce Biogenic secondary organic aerosol (BSOA). In this work we model the interactions between anthropogenic emissions and biogenic precursors using the Community Multi-scale Air Quality (CMAQ) model, for the time period of the Southeast United States during the Southern Oxidant Aerosol Study (SOAS) campaign, which took place in the summer of 2013 in the Southeast United States. We used a CMAQ version incorporating extended isoprene chemistry  (Pye et al. 2013), which is further tuned by utilizing the rich SOAS dataset, in order to by adjust the model parameters controlling the production of BSOA, such as rate constants, partitioning coefficients and deposition (Vasilakos et al. 2016). In our model version, we update the SOA yields for monoterpenes+NO3 reactions using on recently published experimental results (Boyd et al., 2015), in addition to an updated representation of terpene-derived BSOA.  Consequently, monoterpene-derived BSOA is increased by ~60% during nighttime, while being highly correlated with less-oxidized oxygenated OA (LO-OOA), despite model concentrations that were biased low when compared to the observations. By implementing the multigenerational oxidation scheme, the simulated OA was similar to the observations, suggesting that production sources are still not captured by the current model. We also conducted sensitivity tests in order to gauge the response of BSOA to future reductions in NOx and SOx. Our model results suggest a strong correlation between isoprene SOA and sulfate, while LO-OOA production is mediated by the nitrate radical (NO3) during nighttime, consistent with the correlations derived from the SOAS data.

Developing and Evaluating a Multi-Pollutant, Risk-Based Air Quality Management Strategy for the Upstate South Carolina Region

Andy Hollis, Tommy Flynn, Maeve Mason, Kimber Scavo, Neal Fann, Julia Gamas, Ali Kamal, Mark Morris, Ted Palma

This poster describes the results of a collaborative effort between USEPA and SCDHEC to develop and analyze a multi-pollutant risk based air quality management strategy. A suite of test emission controls were evaluated using CoST, CMAQ, BenMAP and the 2011 NATA. While reductions in criteria pollutant concentrations were small to modest, the project shows that even minor improvements in air quality can lead to significant health benefits. The project also provided valuable information used to guide future state air program initiatives.

Source apportionment for sulfate aerosols over East Asia: Case study on the year of 2005
Syuichi Itahashi, Hiroshi Hayami, Keiya Yumimoto, Itsushi Uno

Policy making to reduce anthropogenic emissions are urgently needed to improve air quality. The source apportionment of sulfate aerosol on the year of 2005, the the anthropogenic emissions from China peaked, are evaluated in this study.

Prototype air-water environmental system with linkage between meteorology/ hydrology/ air quality model system and watershed acidification model

Chunling Tang, Jason A. Lynch, and Robin L. Dennis

To help manage excess nitrogen in the environment includes streams and avoid unintended consequences, it is important to better understand how land-use, water use, climate and emission changes may modulate the system's exposures to pollutants and influencemanagementtargets. The biogeochemical processing of nitrogen and associated pollutants is driven by meteorological and hydrological processes in conjunction with pollutant loading. There are feedbacks between meteorology and hydrology that will be affected by land-use change and climate change. Changes in meteorology will affect pollutant deposition. It is important to account for those feedbacks and produce internally consistent simulations of meteorology, hydrology, and pollutant loading to drive the (watershed/water quality) biogeochemical models. In this study, the ecological response to emission reductions in streams in the Potomac watershed were modeled using the linked one comprehensive/environment model, namely the linkage of Community Multiscale Air Quality (CAMQ) model, Weather Research & Forecasting (WRF) model, Variable Infiltration Capacity (VIC) model and Model of Acidification of Groundwater In Catchment (MAGIC) model from 2002 to 2010.The simulated results (such as NO3, SO4, and SBC) fit well to the observed values. The linkage provides a generally accurate, well-tested tool for evaluating sensitivities to varying meteorology and environmental changes on acidification and other biogeochemical processes, with capability to comprehensively explore strategic policy and management design.

Current and Future Mobile Source Contributions to Air Quality

Kirk Baker, Ken Davidson, Sharon Phillips, Margaret Zawacki

A CAMX source apportionment analysis was done for two years, 2011 and 2025, to look at contributions to ambient PM2.5 and ozone from 17 mobile source sectors (e.g. light duty gasoline vehicles, heavy duty diesel trucks, ocean going vessels, lawn and garden equipment). A 12 km grid cell was used across the contiguous U.S. The CAMx source apportionment approach for ozone (OSAT) includes contributions from NOX and VOC to ozone, and the source apportionment approach for PM (PSAT) includes contributions from NOx to PM2.5 nitrate ion, SO2 to PM2.5 sulfate ion, NH3 to PM2.5 ammonium ion, primary EC, primary OC, and other primary PM2.5. The results provide information regarding which mobile source sectors have the largest impact on ambient PM2.5 and ozone concentrations and how those contributions from various mobile source sectors change over time.

Evaluating ammonia (NH3) predictions in the NOAA National Air Quality Forecast Capability (NAQFC) using ground-based and satellite-based measurements on a national scale

William Battye, Casey Bray, Viney Aneja, Daniel Tong, Pius Lee, and Youhua Tang

Ground level and satellite-based measurements are used to test the performance of the National Air Quality Forecast Capability (NAQFC) for predictions of gaseous ammonia. We also evaluate the potential for using satellite-based measurements for improving emissions inputs to the NAQFC CMAQ model.

Influence of the Bermuda High on interannual variability of summertime ozone in the Houston-Galveston-Brazoria region

Yuxuan Wang, Beixi Jia, Sing-Chun Wang, Mark Estes, Lu Shen, Yuanyu Xie

The Bermuda High quasi-permanent pressure system is the key large-scale circulation pattern influencing summertime weather over the eastern and southern US. Here we
developed a multiple linear regression model to characterize the effect of the Bermuda High on year-to-year changes of monthly-mean maximum daily 8-hour average (MDA8) ozone in the Houston-Galveston-Brazoria (HGB) metropolitan region during June, July and August. The Bermuda High (BH) indicators include the longitude of the BH western edge (BH-Lon), and the BH intensity index (BHI) defined as the pressure gradient along its western edge. Both BH-Lon and BHI were found to be significant predictors (p < 0.05) of the interannual (1990-2015) variability of the HGB-mean ozone throughout June-July-August, while local-scale meridional wind speed is selected as an additional predictor for
August only. Local-scale temperature and zonal wind speed are not identified as important factors for any summer month. The BH-Lon is the most important factor, which alone
explains 38%-48% of such variability. The location and strength of the Bermuda High appears to control whether or not low-ozone maritime air from the Gulf of Mexico can enter southeastern Texas and affect air quality. This mechanism also applies to other coastal urban regions along the Gulf Coast (e.g. New Orleans, LA; Mobile, AL; and Pensacola, FL), suggesting that the BH circulation pattern can affect surface ozone variability through a large portion of the Gulf Coast.

Quantification of emission sources apportionment to the concentration of PM2.5 in Temuco, Chile, using receptor model.

Ernesto Pino-Cortes, Luis Diaz-Robles, Giselle Soto, Francisco Cereceda-Balic, Alberto Vergara-Fernandez

Air pollution by particulate matter in the South of Chile is observed during winter season especially during low temperature periods. Temuco, located in Region of Araucania is one of the cities where that situation takes place, with a lot of frequency of overcoming of national standard limit. In that city, measurements of particulate matter (PM10) have been done since 1997, changing to continuous monitoring since 2000. The zone was declared as saturated by PM10 because of the high levels of that pollutant obtained during that period. By other side, an air decontamination plan for PM2.5 was elaborated by authorities and was approved recently in 2015. Many studies were developed since 2000, including air emission inventories in 2005. According to that, 90 % of contribution to PM2.5 can be related to residential wood combustion. The second source was agricultural burning, which only 2 % of contribution. Those results were questioned by some considerations and high uncertainty. The continuous monitoring showed a high contribution of these sources especially during winter, but the real input information was not available for other seasons. In this study the composition of particulate matter was characterized in relation to it's sources. The samples were obtained in the monitoring station called Las Encinas, located in Temuco's City. Fifty samples were collected, from August 18th to October 6th, during winter and spring time in 2008. Also, 51 samples were collected between January 27th and March 21st in 2009, during summer season in Chile. Software UNMIX was used to identify and quantify of the main sources of PM10 and PM2.5 during monitoring period, by receptor model and multivariable analysis. IMPROVE samplers and Teflon filters of 25 mm were used for sampling. The elemental composition was determined by spectrometry XRF in Crocker Nuclear Laboratory at UCDavis. A high correlation between potassium and calcium was observed during winter-spring time. A significant relation between potassium and rubidium was also observed. On the other side, a distinguished correlation between chlorine and sodium was observed during summer. The results predict that the vegetal material combustion or wood burning plus suspended particulate were the principal sources of PM2.5 in winter-spring in 2008. Otherwise, 3 sources for that pollutant were found in summer of 2009: the two mentioned before and salt materials represented by Na and Cl results. Finally, the conclusions obtained were confirmed with simulation of wind direction and speed in Temuco s city, using WRF model meteorology. The marine source mentioned before was validated by this way. Also, this study showed the air emission inventories in Temuco, Chile, have to be compared with receptor model studies on this area.

Estimating Daily Ambient PM2.5 Concentrations in Texas Using High Resolution Satellite Product

Xueying Zhang¹, Yiyi Chu², Kai Zhang¹

1. Department of Epidemiology, Human Genetics and Environmental Sciences, 2. Department of Biostatistics, School of Public Health, The University of Texas Health Science Center at Houston, TX 77030 USA.

The purpose of this study is to estimate ground-level concentrations of Particulate Matter with aerodynamic diameter less than 2.5 m (PM2.5) using satellite-retrieved Aerosol Optical Depth (AOD), a proxy of ground-level PM2.5 concentrations. We used the AOD values generated from the Multi-angle Implementation of Atmospheric Correction (MAIAC) algorithm based on the Moderate Resolution Imaging Spectroradiometer (MODIS) satellites data. The MAIAC algorithm can provide high spatial and temporal resolution AOD data to estimate daily PM2.5 concentrations at 1km geographic scale. Air quality in some areas of Texas is of great concern while the spatial coverage of existing ground monitors is limited. However, the use of AOD to predict PM2.5 concentrations in Texas are not well studied. Therefore, we developed mixed-effect models to estimate daily PM2.5 concentrations using MAIAC AOD in Texas for the study period 2008 -to 2013. The average R2 between fitted and monitored PM2.5 of individual study year ranged from 0.74 to 0.79, and the models showed higher R2 at the monitors located where is close to the East and West borders of Texas than those at the monitors located in the Central Texas. Our results indicate that with the adjustment of the highly varied meteorological conditions and land use types, the MAIAC AOD can be applied to estimate the ground-level PM2.5 concentration with high prediction precision.

The poster will include the study background, the data sources, the model fitting and validation, and the study limitations as well as conclusions.

Impact of Meteorology on Dispersion Model Performance

Fatema Parvez, Kristina Wagstrom

In the United States more than 19% of the total population lives near high traffic roadways. Near road emissions of different pollutants adversely affect both human health and the environment. Vehicular emissions are one of the primary sources of air pollution in cities and lead to elevated morbidity and mortality rates in individuals living near roadways. Currently available near-roadway dispersion models use meteorological conditions from nearby weather stations thereby limiting their implementation in areas lacking available station data. In order to overcome this limitation, processed based meteorological data is used as a substitute. This raises the question of how model results vary when using processed based meteorological data instead of station data.

In this study we employ a Gaussian line dispersion model, R-LINE, to simulate near road concentrations in three cities in Connecticut using both station and Weather Research and Forecasting (WRF) model data. R-LINE simulates the dispersion from line sources by numerically integrating point source emissions along multiple road configurations. We evaluate the seasonal and diurnal variability of roadway dispersion for both meteorological datasets and compare the impact on model estimates. Our result suggests that the variation of model performance is more prominent during night time when atmosphere is at stable condition. As meteorological parameters are the key factors for estimating pollutant concentrations when chemistry is not considered, we also explore R-LINE's sensitivity to different meteorological parameters in both stable and unstable atmospheric conditions for multiple source locations. We find that monin obukhov length, surface roughness and wind speed are the major meteorological parameters that impact roadway dispersion at any atmospheric conditions. In general R-LINE's performance is comparatively less sensitive to meteorology during unstable conditions resulted from predominant impact of convection on dispersion mechanism.

Sensitivity of Simulated Severe PM2.5 Pollution to WRF-CMAQ Model Configurations

Hikari Shimadera and Akira Kondo

This study focused on the impact of WRF-CMAQ configurations to simulated PM_2.5 fields during heavy pollution. The severe PM_2.5 pollution episode in China in January 2013 was simulated by WRF-CMAQ with the East Asian domain (120 100 horizontal grid cells with a resolution of 45 km and 30 vertical layers from the surface to 100 hPa).

The Baseline simulation (Base) was conducted by using WRF v3.4 (analysis data: NCEP FNL and JMA GPV-MSM; PBL scheme/land surface model: ACM2/PX), MOZART-4 for boundary concentrations, various emission datasets (INTEX-B, EAGrid2010-JAPAN, MEGAN v2.04, etc.), and CMAQ v5.0.2 (gas-phase chemistry/aerosol module: CB05/AERO6). The sensitivity of simulated PM_2.5 fields to the model configurations was analyzed with the following cases (change from Base): FNL (NCEP FNL only), ERA (EAMWF ERA-Interim only), ERA+MSM (EAMWF ERA-Interim and JMA GPV-MSM), YSU (YSU/Noah), MYJ (MYJ/Noah), OBSGRID (snow depth update using NCEP FNL), Online1 (online coupled WRF-CMAQ without feedbacks), Online2 (online coupled WRF-CMAQ with feedbacks), Z1mod (modified first layer height from 55 to 40 m), REAS2 (REAS v2 instead of INTEX-B), W37 (WRF v3.7), C51 (CMAQ v5.1), W37-C51 (WRF v3.7 and CMAQ v5.1), SP07 (SAPRC07tc/AERO6), and C51-SP07 (CMAQ v5.1 with SAPRC07tc/AERO6).

In Beijing, while the highest PM_2.5 concentration observed on January 12 was underestimated in all the simulation cases, there were substantial differences between the simulation cases. The maximum simulated PM_2.5 concentration on the day was the highest in the Online2 case (higher by 51% than Base), followed by the YSU case (higher by 33% than Base), and the lowest in the OBSGRID case (lower by 59% than Base).

Improving Cloud Prediction in WRF Through the use of GOES Satellite Observations for SIP Modeling

Andrew T. White, A. P. Biazar, R. T. McNider, K. Doty, and B. Dornblaser

The correct development of clouds in space and time within numerical meteorological models is essential for producing an accurate representation of the physical atmosphere for input into air quality models. This is due to the fact that clouds directly modulate the radiation budget over their area of influence. Therefore, errors in cloud cover inhibit the model's ability to accurately predict variables such as temperature and radiation, and lead to a misrepresentation of vertical mixing, inaccurate development of the boundary layer, photolysis rate, biogenic emission estimates, etc. Reducing the errors in these fields is highly advantageous for improving the meteorological inputs into air quality models, especially for SIP modeling studies. Through assimilation of Geostationary Operational Environmental Satellite (GOES) derived cloud fields within the Weather Research and Forecasting (WRF) model, cloud placement in time and space within the model can be improved. The assimilation method consists of determining where the model over- or under-predicts cloud coverage by comparing the model predicted cloud field to the GOES derived cloud fields. Once these locations are known, a technique was developed to introduce vertical velocity into the model to create or dissipate clouds to correct the over and under prediction of clouds present in the model. This technique has been tested on numerous grid resolutions, 36-km, 12-km, 4-km, and numerous grid configurations over the CONUS domain. The technique was developed and validated on an August 2006 study and was applied and evaluated on an August-September 2013 modeling study. The results indicate that the assimilation technique significantly improves the agreement between the model predicted and GOES derived cloud fields. The daily average percentage increase in the cloud agreement surpassed or approached ten percent in nearly all conducted modeling simulations. Results from the use of the method will be presented.

Evaluation and Comparison of Fourteen Air Pollution Field Development Methods Regarding their Application in Exposure Assessment

Haofei Yu, Armistead Russell, Jim Mullholland, Talat Odman, Yongtao Hu and Howard Chang

To investigate the adverse health consequences of air pollution, many methods have been applied in the past to develop spatial and temporal fields of pollutant concentrations for exposure and epidemiological studies. Outcomes of these studies rely heavily on the fields developed and could differ among methods. To better inform users on the selection of appropriate methods, and to identify potential issues regarding the application of different methods, we evaluated and compared pollution concentrations fields developed from a common dataset using 14 different methods. The methods selected include using the central monitor (CM), site averaging (SA), four spatial interpolation methods (inverse distance weighting, tessellation, kriging and land use regression), two air quality models ( RLINE dispersion model and CMAQ chemical transport model), four air quality model-based data fusion techniques, satellite aerosol optical depth (AOD), and AOD downscaling. The study domain is the Atlanta Metro Area in Georgia, USA and the focus pollutants are CO, NO2, SO2, O3, PM2.5 and three of its constituents: elemental carbon, organic carbon and sulfate. The CM and SA methods provide no spatial variability, and are temporally and chemically-limited. How monitors are weighted should be chosen carefully to avoid significant over- or under-estimated exposures. Concentration fields from interpolation methods are similar to each other though the spatial patterns of some pollutants (such as SO2) are not appropriately captured. Raw outputs from RLINE and CMAQ models reasonably captured spatial variations of pollutant concentrations, but temporal variation errors and biases undermined their usefulness in exposure estimations. Based upon air quality model outputs, data fusion techniques corrected for these errors and biases and significantly improved method performances, both spatially and temporally. The satellite AOD methods perform similarly as data fusion techniques, although subjected to non-retrieval issues. The AOD downscaling methods, on the other hand, substantially improved the spatial resolution of concentration fields. Results suggest that air-quality model based data fusion approaches can provide the spatial, temporal and chemical detail desired for health studies, though some issues (e.g., resolution and potential biases) remain to be more comprehensively addressed.

High resolution OMI satellite retrievals of tropospheric NO2 in the eastern United States

Daniel L. Goldberg, David G. Streets, Zifeng Lu, Lok N. Lamsal, Christopher P. Loughner

Satellite measurements provide greater spatial coverage than any other observing platform. However, their use in air quality policy community remains tepid due to their coarse spatial and temporal resolution. Improving the spatial resolution of satellite data products may spur their use in the policy community. This work presents a new high resolution NO₂ dataset derived from the standard NASA Ozone Monitoring Instrument (OMI) NO₂ v2.1 product for the eastern United States. The standard product uses NO₂ vertical profile shape factors from a 2.5 x 2 resolution NASA Global Model Initiative (GMI) model simulation to calculate air mass factors, a critical value used to determine tropospheric NO₂ vertical columns. While GMI can provide global coverage and is extremely useful in an operational setting due to its quick runtime, the shape factors generated on a 2.5 x 2 grid are not representative of regions with large spatial heterogeneities, such as near major urban areas and large point sources. To better estimate vertical profile shape factors, we use regional air quality simulations (<12 km) to recalculate tropospheric air mass factors and tropospheric NO₂ columns during the summer of 2011. Results show that retrievals using these new air mass factors capture the fine-scale gradients near urban areas and large point sources. Although the current work is focused on the eastern United States, the methodology developed in this work can be applied to other world regions to produce high-quality region-specific NO₂ satellite retrievals.

AQMEII3: the EU and NA regional scale program of the Hemispheric Trasport of Air Pollution Task Force

Stefano Galmarini1,*, Christian Hogrefe2, Efisio Solazzo1 and the AQMEII3 Modeling-Community

The proposed paper builds on the work presented last year at the 14th CMAS meeting and it is applied to the work performed in the context of the AQMEII-HTAP collaboration. The analysis is conducted within the framework of the third phase of AQMEII (Air Quality Model Evaluation International Initiative) and encompasses the gauging of model performance through measurement-to-model comparison, error decomposition and time series analysis of the models biases for several fields (ozone, CO, SO2, NO, NO2, PM10, PM2.5, wind speed, and temperature). Through the comparison of several regional-scale chemistry transport modelling systems applied to simulate meteorology and air quality over two continental areas, this study aims at i) apportioning the error to the responsible processes through time-scale analysis, and ii) help detecting causes of models error, and iii) identify the processes and scales most urgently requiring dedicated investigations. The operational metrics (magnitude of the error, sign of the bias, associativity) provide an overall sense of model strengths and deficiencies, while the apportioning of the error into its constituent parts (bias, variance and covariance) can help assess the nature and quality of the error. Each of the error components is analysed independently and apportioned to specific processes based on the corresponding timescale (long scale, synoptic, diurnal, and intra-day) using the error apportionment technique devised in the previous phases of AQMEII. The application of the error apportionment method to the AQMEII Phase 3 simulations provides several key insights, which will be presented.

Utilization of Geostationary Satellite Observations for Air Quality Modeling During 2013 Discover-AQ Texas Campaign

Arastoo Pour Biazar¹, Maudood Khan, Andrew White¹, Rui Zhang², Dan Cohan², Bright Dornblaser³, Richard T. McNider¹

1. University of Alabama in Huntsville
2. Rice University
3. Texas Commission on Environmental Quality (TCEQ)

By altering photochemical reaction rates, as well as modulating the emissions of biogenic volatile organic compounds (BVOCs), clouds significantly impact atmospheric photochemistry. Additionally, clouds affect temperature, boundary-layer development, lead to deep vertical mixing of pollutants and precursors, and induce aqueous phase chemistry and recycling of particulate matter. Therefore, model cloud errors can considerably impair air quality simulations. Unfortunately, numerical meteorological models have difficulty in creating clouds in the right place and time compared to observed clouds. This is especially the case when synoptic-scale forcing is weak, as often is the case during air pollution episodes.

In the current study we utilize Geostationary Operational Environmental Satellite (GOES) observations to improve the realization of clouds and their impact on air quality simulations. GOES provide an excellent observational platform for monitoring clouds. For this study, a new GOES derived product, photosynthetically active radiation (PAR), was generated and used for estimating BVOC emissions. Satellite-based PAR estimates rely on the technique used to derive insolation from satellite visible brightness measurements. PAR retrievals were evaluated against surface observations and exhibited a good agreement. Moreover, GOES observed clouds were assimilated within Weather Research and Forecasting (WRF) model to improve simulated clouds. Cloud assimilation technique dynamically support cloud formation/dissipation within WRF based on GOES observations. A series of WRF/CMAQ simulations for the summer of 2013 that coincides with the Deriving Information on Surface Conditions from COlumn and VERtically Air Quality campaign over Texas were performed to quantify the impact of satellite data. While direct use of satellite-derived PAR improves biogenic emission estimates, correcting cloud fields in WRF would be preferable as it recreates a more accurate physical environment for CMAQ simulations. In this presentation, first satellite-derived PAR will be introduced followed by the results from the CMAQ simulations. In particular, the impact of direct use of satellite data in BVOC estimates will be compared with the results from cloud assimilation in WRF.

Multi-model Comparison of Lateral Boundary Contributions to Ozone Concentrations over the United States

Peng Liu, Christian Hogrefe, Johannes Bieser, Ulas Im, Rohit Mathur, Uarporn Nopmongcol, Shawn Roselle, Tanya Spero

As the National Ambient Air Quality Standards (NAAQS) for ozone become more stringent, there has been growing attention on characterizing the contributions and the uncertainties in ozone from outside the US to the ozone concentrations within the US. The third phase of the Air Quality Model Evaluation International Initiative (AQMEII3) provides an opportunity to investigate this issue through the combined efforts of multiple research groups in the US and Europe. The model results cover a range of representations of chemical and physical processes, vertical and horizontal resolutions, and meteorological fields to drive the regional chemical transport models (CTMs), all of which are important components of model uncertainty (Solazzo and Galmarini, 2016). In AQMEII3, all groups were asked to track the contribution of ozone from lateral boundary through the use of chemically inert tracers. Though the inert tracer method tends to overestimate the impact of ozone boundary conditions compared with other methods such as chemically reactive tracers and source apportionment (Baker et al., 2015), the method takes the least effort to implement in different models, and is thus useful in highlighting and understanding the process-level differences amongst the models.

In this study, results from four models were included, which are named as US3 (CMAQ driven by WRF, US EPA), US1 (CAMx driven by WRF, Ramboll Environ, US), DE1 (CMAQ driven by CCLM), and DK1 (DEHM). First, the model performance for surface ozone predictions was compared with observations, grouped by season, station type (rural or urban sites), and elevation. Then, at each site, the distribution of daily maximum 8-hour ozone, and the corresponding concentrations of the inert tracers was calculated and compared across different models. Furthermore, as required by AQMEII3, the impacts of lateral boundary ozone from lower troposphere, free troposphere and stratosphere were tracked separately by the inert tracers during the simulation. Hence, in this study, there boundary tracers were used to investigate how boundary ozone at different altitudes may affect the surface ozone concentrations.

Source Influences on Ambient Ozone Precursor Concentrations in the Colorado Front Range

Shannon L. Capps1, Gokulram Paranjothi1, and Jana B. Milford1

1 Department of Mechanical Engineering, University of Colorado Boulder, Boulder, CO

Increased oil and gas development, particularly through horizontal fracturing, in the Denver-Julesburg Basin in Colorado over the last decade has been identified as a potential source of emissions of air pollutants. As one effort to evaluate its impact, ambient concentrations of volatile organic compounds (VOCs) that serve as a precursor to ozone formation have been measured in areas expected to be influenced primarily by oil and gas development or by urban activity. To evaluate the source influences on these ambient concentrations, we use the EPA''s Positive Matrix Factorization tool with data from the Ozone Precursor Study conducted by the Colorado Department for Public Health and Environment during 2013 and 2014, which provides early morning measurements of an extensive suite of ozone-precursor VOCs from a site in Platteville, CO, and another in Denver, CO. Additional data collected during more brief measurement campaigns in the last five years is used to augment the findings from the primary data set. The factors most closely aligned with an oil and gas signal will be compared with the VOC emissions estimates for the same sector currently being used in a 2011 CAMx simulation, which is nested to 4-km x 4-km horizontal resolution over the Rocky Mountain region.

Preliminary Results of the Model Intercomparison Study in the Asia (MICS-Asia) Phase III
Kan Huang¹, Joshua S. Fu¹, Zifa Wang², Jun-ichi Kurokawa³, Baozhu Ge^2
1Department of Civil and Environmental Engineering, University of Tennessee, Knoxville, TN, USA
²Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China
³Asian Center for Air Pollution Research, Niigata, Japan

The MICS-Asia Phase III project is a continuing activity after the Phase I and Phase II. This is a cooperative multi-model effort by brining together modelers from China, Japan, Korean, Southeast Asia and the United States. The MICS-Asia III is an independent project supported by Institute of Atmospheric Physics (IAP) and Asian Center for Air Pollution (ACAP) through Joint International Center on Air Quality Modeling Studies (JICAM) between IAP and ACAP. The model inter-comparison activities in Phase III have been conducted to be carried out using common meteorological fields, emission data, boundary conditions, etc. in order to allow the discussion of the cause of disagreement among participating models rather than just showing the variability of model output and uncertainties. The participating models including, CMAQ (different versions), WRF-Chem, NAQPMS, NHM-Chem, GEOS-Chen, etc. It is aimed to evaluate strengths and weaknesses of current air quality models for air quality prediction, and provide techniques to reduce uncertainty and improve performance in Asia.

It is expected that the simulation capabilities of air pollution in East Asia will be improved in the future via this task, especially the ability to simulate heavy pollution. This activity not only provides scientific assessment of current status of air quality and regional climate in East Asia but also can provide sound recommendation for the efficient abatement of air pollutant emissions. The MICS-Asia is also invited to join the Hemispheric Transport of Air Pollution Phase II regional modeling activities.

Recent Performance of the NOAA Air Quality Prediction System using CMAQ and the Impact of Driving Meteorology

Jeff McQueen, Jianping Huang, Ho-Chun Huang, Perry Shafran, Geoff DiMego, Amanda Sleinkofer

Pius Lee, Li Pan, Daniel Tong

Ivanka Stajner, Sikchya Upadhayay

Operational air quality predictions for the United States (U. S.) are provided by National Air Quality Forecasting Capability (NAQFC). NAQFC provides nationwide operational predictions of ozone and particulate matter. Predictions are produced twice per day (at 06 and 12 UTC cycles) at 12 km resolution and 1 hour time intervals through 48 hours and distributed at http://airquality.weather.gov. The NOAA National Centers for Environmental Prediction (NCEP) operational North American Mesoscale (NAM) 12 km weather predictions is used to drive the Community Multiscale Air Quality (CMAQ) model. Higher resolution operational NWP models have recently been introduced as part of the NCEP model suite. These include the NAM CONUS Nest (4 km horizontal resolution) and the Fire Weather nest (1.33 km for smaller domains) run four times per day through 60 hours and the High Resolution Rapid Refresh (HYYR, 3 km) run hourly out to 15 hours. This paper will evaluate the impact of higher resolution models for capturing boundary layer and orographic flows important for driving air quality prediction. The NWP models will be evaluated against standard and mesonet fields averaged for various regions during the summer 2015. Surface ozone and PM observations are used to identify cases where large NAQFC prediction errors occurred. The Updated Real-time Mesoscale Analysis (URMA) will also be used for evaluations. URMA adjusts a background meteorological field to the nearby observations using the two dimensional version of the NCEP variational Global Statistical Interpolation (GSI) assimilation system. Such diagnostic downscaled meteorological analyses and predictions that account for complex terrain and coastal effects are often used to drive air pollution models on local scales. Large errors in ozone prediction are often found in California during the summer months while using the operational 12 km NAM - CMAQ system. The presentation overviews the NCEP mesoscale analysis and prediction systems (e.g.: NAM and its high resolution nests, HYYR, RTMA). An evaluation of meteorological fields important to air quality modeling (eg: near surface winds, temperatures, moisture and boundary layer heights, cloud cover) will be reported on for the NAM, NAM nests and downscaled predictions especially for the Los Angeles basin, and Sacramento Valleys. The meteorological performance will be presented along with errors in the operational NAM-CMAQ system to evaluate the correlation between ozone error and the ability of the mesoscale model to resolve local flows influenced by the complexity of surrounding topography.

Evaluation of the Community Multiscale Air Quality (CMAQ) modeling system version 5.2

K. Wyat Appel, Sergey Napelenok, Christian Hogrefe, George Pouliot, Kristen M. Foley, Jesse O. Bash, William Hutzel, Deborah Luecken, Golam Sarwar, Donna Schwede, David C. Wong, Shawn J. Roselle, Jonathan E. Pleim, and Rohit Mathur

The Community Multiscale Air Quality (CMAQ) model is a state-of-the-science air quality model that simulates the emission, transport and fate of numerous air pollutants, including ozone and particulate matter. The Computational Exposure Division (CED) of the U.S. Environmental Protection Agency develops the CMAQ model and periodically releases new versions of the model that include bug fixes and various other improvements to the modeling system. In the fall of 2016, CMAQ version 5.2 will be released. This new version of CMAQ will contain important bug fixes to several issues that were identified in CMAQv5.1 (the current public release version of the CMAQ model), and additionally include updates to other portions of the code. Some specific model updates include a new implementation of the wind-blown dust calculation in CMAQv5.2 which fixes several bugs that were identified in the current implementation of wind-blown dust in CMAQv5.1. Several other major updates to the model include an update to the calculation of aerosols; implementation of full halogen chemistry (CMAQv5.1 contains a partial implementation of halogen chemistry), which is particularly important for hemispheric applications of the CMAQ model, as halogen chemistry is need to accurately simulation the destruction of ozone over the ocean; the new Carbon Bond 6 (CB6) chemical mechanism; and various other updates to the modeling system.

Impacts of WRF lightning assimilation on offline CMAQ simulations

Nicholas Heath, Jonathan Pleim, Robert Gilliam, Daiwen Kang, Matthew Woody, Kristen Foley, Wyat Appel

Deep convective clouds vertically redistribute trace gases and aerosols and also provide a source for scavenging, aqueous phase chemistry, and wet deposition, making them important to air quality. Regional air quality simulations are typically driven by meteorological models that use relatively coarse resolution and require a convective parameterization. Unfortunately, convective parameterizations generally do not represent the timing and location of deep convection accurately. Additionally, positive rainfall biases commonly exist during summer months due to overactive parameterizations. These shortcomings adversely impact air quality simulations. In this study, lightning assimilation was applied in the Kain-Fritsch (KF) cumulus parameterization to improve the simulation of deep convection in the Weather Research and Forecasting (WRF) model. The method has a straightforward approach: Force KF deep convection where lightning is observed and, optionally, suppress deep convection where lightning is absent. WRF simulations with and without lightning assimilation were made for July 2011. Major improvements were seen in WRF 6-h precipitation accumulations when lightning assimilation was used. For example, when compared to Stage-IV observations, the monthly averaged spatial correlations more than doubled from 0.22 to 0.47 and the mean absolute error was reduced from 0.83 to 0.57 mm. The two WRF simulations were then used to drive offline CMAQ simulations and here we present the impacts of the lightning assimilation on surface ozone and particulate matter concentrations evaluated against routine monitoring networks.

Alternative Future Realities - Considerations for Modeling

Moderator: Tom Moore (WESTAR-WRAP)

Panelists: Michael Barna (National Park Service - Air Resources Division), Chris Emery (Ramboll-Environ), Dan Loughlin (U.S. EPA), Tanya Spero (U.S. EPA)

One application of regional-scale air quality models is for projecting medium (10-20 years) to long-range (> 20 years) air quality. The projections are powerful applications of the models that support decisions about emissions control programs and are used to study the influence of global change on regional air quality. While conventional regional modeling projections only consider the impacts of changes in emissions, the other model inputs (e.g. meteorology, boundary conditions, land use) should also be projected to reflect future conditions. This year's CMAS Community Forum will discuss whether the time is right to improve air quality modeling of future years by projecting variation in model input parameters in addition to emissions.

Topics that we would like to consider during this discussion include:

• Which input parameters (e.g. meteorology, emissions, boundary conditions, landuse) can be feasibly and reliably projected given current understanding and technical resources?
• What tools need to be developed to facilitate bringing other types of projected data into operational regional modeling studies?
• To what extent are integrated modeling systems available to account for the couplings in future landuse, meteorology, emissions, and air quality?
• If future climate and emissions patterns are not likely to track historical trends, what are alternative ways to project variation?
• Given that multiple future year scenarios are possible, should we consider alternatives to deterministic models for projecting medium to long-range air quality?

The forum will begin with a panel discussion on these and related topics followed by the community forum.

Slides

• Intro
• Barna's Slides
• Emery's Slides
• Loughlin's Slides
• Spero's Slides

NOX emissions, isoprene oxidation pathways, vertical mixing, and implications for surface ozone in the Southeast United States

Katherine R. Travis, Daniel J. Jacob, Jenny A. Fisher, Patrick S. Kim, Eloise A. Marais, Lei Zhu, Karen Yu, Christopher C. Miller, Robert M. Yantosca, Melissa P. Sulprizio, Anne M. Thompson, Paul O. Wennberg, John D. Crounse, Jason M. St. Clair, Ronald C. Cohen, Joshua L. Laughner, Jack E. Dibb, Samuel R. Hall, Kirk Ullmann, Glenn M. Wolfe, Illana B. Pollack, Jeff Peischl, Jonathan A. Neuman, and Xianliang Zhou

Ozone pollution in the Southeast US involves complex chemistry driven by emissions of anthropogenic nitrogen oxide radicals and biogenic isoprene. Model estimates of surface ozone concentrations tend to be biased high in the region. We use observations from the SEAC4RS aircraft campaign in 2013, interpreted with the GEOS-Chem chemical transport model, to better understand the factors controlling surface ozone in this region. We find that EPA's National Emission Inventory for NOX is too high nationally by 50%. This is demonstrated by SEAC4RS observations of NOX and its oxidation products, by surface network observations of nitrate wet deposition fluxes, and by OMI satellite observations of tropospheric NO2 columns. Upper tropospheric NO2 from lightning makes a large contribution to the satellite observations that must be accounted for when using these data to estimate surface NOX emissions. GEOS-Chem with reduced NOX emissions provides an unbiased simulation of ozone observations from the aircraft and from ozonesondes, and reproduces the observed ozone production efficiency in the boundary layer as derived from a regression of ozone and NOx oxidation products. The model is still biased high relative to observed surface ozone.

Equatorward Redistribution of Emissions Dominates the Tropospheric Ozone Change, 1980-2010

Yuqiang Zhang, Owen R. Cooper, Audrey Gaudel, J. Jason West

Since 1980, global emissions of ozone precursors have shifted toward the equator as less industrialized nations have increased their emissions and industrialized mid-latitude nations have reduced emissions. Modeling studies have also shown that the global tropospheric ozone burden is much more sensitive to changes in emissions in the tropics than in the mid-latitudes. However, previous research has not previously quantified the importance of the historical change in emission location on the tropospheric ozone burden. Here we use global model simulations to separate changes in the global tropospheric ozone burden from 1980 to 2010 into components due to changes in: i.) the global spatial distribution of anthropogenic emissions, ii.) the total magnitude of global emissions, and iii.) the global methane concentration. Using the CAM-chem global atmospheric model, we show that the change in the spatial distribution of emissions is responsible for over half of the net ozone burden change from 1980 to 2010 - more than the effects of the global emission magnitude and the methane change combined. This greater sensitivity to tropical emissions is caused by the strong convection, fast photochemistry, and highly NOx-sensitive conditions in this region, despite the shorter lifetime of ozone. We present evidence suggesting that emissions increases from Southeast Asia, East Asia, and South Asia have been most important for this global ozone increase. Observations from global ozonesondes, satellites, and in-service aircraft support a rapid growth in ozone columns over Southeast and South Asia. The spatial distribution of emissions, particularly emissions from the tropics, dominate global ozone. Previous analyses may have put too much emphasis on emission magnitude and too little on emission location. A continued equatorward shifting of emissions may cause global ozone to continue to increase even if global ozone precursor emissions decrease.

Ongoing EPA efforts to evaluate modeled NOy budgets

Heather Simon1, Darrell Sonntag2, Megan Beardsley2, Kristen Foley3, Chris Owen1, Claudia Toro2, Norm Possiel1, Alison Eyth1
1 US EPA, Office of Air Quality Planning and Standards
2 US EPA, Office of Transportation and Air Quality
3 US EPA, National Exposure Research Laboratory

Some recent studies have suggested that modeled concentration and deposition fields of NOy in the U.S. are biased high. This presentation describes a cross-EPA effort to further explore this issue and diagnose the potential causes for any such discrepancies in modeled and measured NOy budgets. Here we provide an overview of ongoing efforts in EPA's Office of Air Quality Planning and Standards (OAQPS), Office of Transportation and Air Quality (OTAQ), and Office of Research and Development (ORD) related to this issue. Projects include analysis of recent near road measurements by EPA in Las Vegas and Detroit, review of recent tunnel and near-road measurement data reported in the literature, an in-depth diagnostic model evaluation with measurements from field studies and monitoring networks using recent versions of EPA's modeling inventory and CMAQ photochemical model, and an examination of uncertainties and impacts of spatial and temporal allocation of mobile source activity and emissions.

Estimating age-segregated per-vehicle health benefits for the Canadian fleet

Angele Genereux, Amanda Pappin, Amir Hakami, Shunliu Zhao (Carleton University); Matt D. Turner, Shannon L. Capps, and Daven K. Henze (University of Colorado); Peter B. Percell (University of Houston); Jaroslav Resler (ICS Prague); Jesse O. Bash, Sergey L. Napelenok, and Kathleen Fahey (US EPA), Rob W. Pinder, Armistead G. Russell, and Athanasios Nenes (Georgia Tech); Jaemeen Baek, Greg R. Carmichael, and Charlie O. Stanier (University of Iowa); Adrian Sandu (Virginia Tech); Tianfeng Chai (University of Maryland); Daewon Byun (NOAA)

The transportation sector has been identified as the largest source of NO_X emissions in Canada (e.g. in 2014, the transportation sector was responsible for 42% of total NO_X emissions). In this study, we examine the impact of NO_X emissions from transportation on Canadian public health, and segregate these impacts based on the vehicle age. To do this, mobile emissions modeling in SMOKE is used to generate light-duty gasoline (LDG) and light-duty diesel (LDD) emission factors for NO_X. Through the combination of these emission factors with location-specific, adjoint-based benefit-per-ton estimates for NO_X emissions (Pappin et al. 2016), we calculate the Canadian health benefit of removing one average vehicle from the road (the per-vehicle health benefit [PVHB]). While our preliminary results focus on gas-phase emissions, we are working to extend our analysis to account for PM_2.5 emissions as well. Using our PVHB results we also examine the health impact of manipulation of NO_X control devices in Canada. We characterize the distribution of PVHBs, on a per-year basis, by grouping the vehicles into five age bins within the light-duty vehicle fleet, including 0-2 years, 3-5 years, 6-9 years, 10-13 years, and 14+ years. For example, for a LDG vehicle in Ottawa, PVHBs are found to range from $9,200/vehicle-year for a 14-year old vehicle to $220/vehicle-year for a new model. The wide range of PVHBs by vehicle age, accentuates the importance of removing older, more damaging vehicles from the fleet.

Top-Down Constraints on Emissions of NH3, NOx, and SO2 during the 2013 NOAA SENEX Campaign

M. J. Alvarado, C. R. Lonsdale, E. Winijkul, C. M. Brodowski, K. E. Cady-Pereira, D. K. Henze, S. Capps

Accurate modeling of the formation of ozone (O₃) and fine particulate matter (PM_2.5) requires accurate estimates of the emissions of precursor species such as ammonia (NH₃), nitrogen oxides (NO_x = NO+NO₂) and sulfur dioxide (SO₂). Here we present an evaluation of the 2011 NEI emissions of NH₃, NO_x, and SO₂ using CMAQv5.0.2 and data from the 2013 NOAA Southeast Nexus (SENEX) field campaign. Model results are compared to surface and aircraft measurements during each campaign, as well as satellite NH₃ observations from the NOAA Cross-track Infrared Sounder (CrIS) and satellite observations of NO₂ and SO₂ from the NASA Ozone Monitoring Instrument (OMI). We discuss the lessons learned in using CrIS NH₃ observations in the southeast US, where CMAQ predicts most of the gas-phase NH₃ is very close to the surface. We discuss the use of two methods -- a mass balance approach and an approach using the CMAQ adjoint -- to optimize these emissions and evaluate the improvement in model performance for gas-phase NH₃, NO_x, and SO₂, as well as for the formation of O₃ and PM_2.5.

Evaluation of rainfall Intensity-Duration-Frequency (IDF) curves developed from dynamically downscaled regional WRF simulations

Chuen Meei Gan¹ and Tanya Spero²

(1) CSRA LLC , 79 T. W. Alexander Drive, RTP, NC 27709

(2) NERL, U.S. EPA, 109 T. W. Alexander Dr., RTP, NC 27711 USA

The total rainfall collected in a specified period at a location is highly variable from one year to another. This variability depends on the type of climate and the duration of the considered period. Overall, it can be stated that the drier the climate, the stronger the variability of the rainfall in time. The same holds for the length of the period: the shorter the period, the higher the annual variability of rainfall in that period. Due to the strong variability of rainfall in time, the planning, design, and management of water resources projects are not based on long-term averaged of rainfall records but on particular rainfall intensities that can be expected for a specific probability or return period. These rainfall intensities can only be obtained by a thorough analysis of long time series of historic rainfall data. In particular, estimates of extreme rainfall intensities are required for the design of drainage or sewer systems. For this reason, the annual maximum series, which is the set of the maximum values observed during a period (e.g. minutes, hours or day) in each year, are analyzed. In this study, a 23 year (1988 - 2010) dynamically downscaled 36-km Weather Research and Forecasting (WRF) model simulation is used. Initially, 20 sites located in the south east US region are selected which include differing types of geography. The Gumbel distribution, which is commonly used in rainfall analysis and always gives satisfactory results, is applied to develop Intensity-Duration-Frequency (IDF) curves at each site. The accuracy of the results will be assessed by comparison against IDF curves created using observation data.

Dynamic analysis: assessing CMAQs ability to capture air quality trends over a time period of changing emissions

Cong Liu, Lucas RF Henneman, Yongtao Hu, James A Mullholland, and Armistead G Russel

The time period 2001-2012 saw dramatic decreases in anthropogenic emissions and subsequent improvements in air quality in the United States. We use multiple year-long model runs (2001-02 and 2011-12) to assess CMAQ's ability to capture both absolute concentrations and observed monthly and long-term trends in the Eastern United States. Evaluations of the model runs using multiple metrics shows the results are within accepted ranges. A unique aspect of our application and evaluation is the assessment of model sensitivities derived from DDM and how they have evolved over a decade of greatly reducing emissions, as well as the use of CMAQ applied over periods a decade apart, to assess Ozone Production Efficiencies (OPEs).Results show that in the Southeast, CMAQ-simulated PM2.5 is biased low by roughly 30%, and under-predicts the change in observed ambient concentrations between these years by a similar amount. Detailed evaluation of PM2.5 species shows that different species contribute more to the bias at various times of the year. For example, organic carbon (OC) and nitrate aerosols tend to be biased high in the winter, and OC and sulfate aerosols are biased low in the summer. Overall, simulated PM2.5 is biased high in the winter and low in the summer. Ozone changes between 2001 and 2011 show an increase in the winter time and a decrease in the summer. The increased ozone in the winter is due to reduced NO titration, while the decreased summertime ozone is due to reductions in NOx and VOC emissions. An analysis of impacts of meteorological and emission changes is performed by running CMAQ using switched meteorological fields between 2001 and 2011. Results show that the emissions effect has a greater mean than the meteorological effect, but meteorology impacts dominate above the 95th percentile of ozone concentrations. Finally, we investigate sensitivities of ozone and PM2.5 to mobile and point sources using CMAQ's DDM extension, and assess their change over time. DDM sensitivities to those derived using empirical methods show coherence, but some important differences, such as the concentration where the sensitivity to NOx emissions changes sign.

Using Extreme Events to Compare USGS and NLCD Land Use Data Sets in WRF for Dynamical Downscaling

Stephany M. Taylor, Tanya L. Spero, and Megan S. Mallard

This study compares the United States Geological Survey (USGS) land use data and the National Land Cover Data (NLCD) in the Weather Research and Forecasting (WRF) model for dynamical downscaling. Extreme climate events during 1988-1990 were selected as case studies to determine how these land use data sets react to different atmospheric scenarios. The three-year, 36-km WRF simulations of the contiguous United States using these two land use data sets were compared to Climate Prediction Center (CPC) precipitation data and North American Regional Reanalysis (NAYY) data to evaluate model performance. This presentation will show results from six regional case studies involving drought, tropical storms, hard freezes, flooding, and severe weather outbreaks.

Two Decades of WRF/CMAQ simulations over the continental United States: New approaches for performing dynamic model evaluation and determining confidence limits for ozone exceedances

Marina Astitha¹*, Huiying Luo¹, S. Trivikrama Rao¹, Christian Hogrefe², Rohit Mathur², and Naresh Kumar^3
1 Civil and Environmental Engineering, University of Connecticut, Storrs, CT, USA
² U.S. Environmental Protection Agency, Research Triangle Park, NC, USA
³ EPRI, Palo Alto, CA, USA

Confidence in the application of models for forecasting and regulatory assessments is furthered by conducting four types of model evaluation: operational, dynamic, diagnostic, and probabilistic (Dennis et al. 2010). Operational model evaluation alone does not reveal the confidence limits that can be associated with modeled air quality concentrations. This paper presents novel approaches for performing dynamic model evaluation and for evaluating the confidence limits of ozone exceedances using the WRF/CMAQ model simulations over the continental United States for the period from 1990 to 2010 (Gan et al., 2015). The methodology presented here entails spectral decomposition of ozone time series using the KZ filter (Rao and Zurbenko, 1994; Rao et al., 1997) to assess the variations in the strengths of the synoptic (i.e., weather-induced variation) and baseline (i.e., long-term variation attributable to emissions, policy, and trends) forcings embedded in the modeled and observed concentrations. The dynamic evaluation approach conducted here involves the use of anomalies, where the deviations from the 21-year mean in the observed and modeled 4th highest, 95th, 90th and 85th percentile of daily maximum 8-hr ozone concentrations are compared for all monitoring stations in the continental US. Also, an alternative method is presented, where the future year observations are estimated based on the changes in the concentrations predicted by the model applied to the current year observations. The results reveal that CMAQ reproduces the spatio-temporal features seen in observations (synoptic forcing, baseline, ozone exceedances, 4th highest and the correlation among them), with some features being more strongly correlated than others. These new methods show the root mean square error is nearly constant and small (0-5ppb) at all monitoring sites in the continental U.S., indicating that CMAQ is highly skillful in predicting the changes in concentrations. In a similar manner, the proposed method can provide confidence limits for ozone exceedances for a given emission reduction scenario. We present and discuss these new approaches to identify the strengths of the model in representing the changes in simulated O3 air quality over the 21-year period. These approaches show how regional-scale air quality models can be used with greater confidence for policy analysis.

Using Response Surface Modeling (RSM) for the Task Force on Hemispheric Transport of Air Pollution (HTAP)
Joshua S. Fu, Xinyi Dong, Jiani Tan, Kan Huang
Department of Civil and Environmental Engineering, The University of Tennessee, Knoxville, Tennessee, 37996, USA

The Task Force on Hemispheric Transport of Air Pollution (TF HTAP) is an international scientific cooperative effort to improve the understanding of the intercontinental transport of air pollution across the Northern Hemisphere. The assessment report for HTAP Phase1 was published in 2010, including 5 separate documents summarizing the research findings about ozone and particulate matter, mercury, persistent organic pollutants, and answers to policy relevant questions with respect the intercontinental transport of air pollutants. The HTAP Phase 2 was started in 2012 with the objective to continue the investigation of transport of air pollutants, assessment of potential emission mitigation options, evaluation of the anthropogenic impacts on regional and global air quality, public health, ecosystems, and near-term climate change.

This study is within the HTAP Phase2 efforts. We use the Response Surface Modeling (RSM) method with the WRF/CMAQ modeling system to investigate the source-receptor relationships between 7 sub-regions as shown in Figure 1. These sub-regions are defined in the modeling work based on their contributions to the global anthropogenic emission inventory. Conventionally, emission perturbation is applied to specify the regional emission changes under each experimental simulation, and then Brute Force (BF) method is applied to quantify the source-receptor relationships between different sub-regions. This method is usually limited by the number of experimental simulations due to intensive efforts for designing and preparing inputs, and consequently limit the assessment of the relationship as well. While in this study, the RSM technology enable us to conduct a batch of simulations with the support of the most-developed super computer. A total of 310 perturbation scenarios are generated based on Lartin-Hypercube sampling method and applied to CMAQ simulations. The response surface derived from these simulations thus enable us to investigate the air pollutants responses at any given emission perturbation, which help to develop the source-receptor relationships between the sub-regions. The intercontinental transport of air pollutants and their precursors will be analyzed based on the RSM result.

Decadal Application of WRF/Chem under Current and Future Climate/Emission Scenarios: Part I. Comprehensive Evaluation and Intercomparison with Results under the RCP 8.5 Scenario

Kai Wang, Yang Zhang, Patrick Campbell, and Khairunnisa Yahya
Department of Marine, Earth, and Atmospheric Sciences, North Carolina State University, Raleigh, North Carolina, 27695

Regional air quality is strongly affected by the changes in climate and pollutant emissions. Accurate predictions of responses of regional air quality to such changes can help policy makers in developing effective control strategies to reduce air pollutants and improve air quality in the future. The capability of models has been greatly enhanced in representing the complex interactions among population, socio-economic development, technological change, and federal and regional environmental policies owing to more advanced chemical treatments and complex chemistry-meteorology feedback processes included in regional climate/air quality models, the development of dynamical downscaling technique, and more sophisticated tools/approaches (e.g., the Technology Driver Model) in emission projections.

Taking advantage of recent progress in model improvements of aerosol formation and aerosol-cloud interactions, downscaling technology, and emission projection approaches, the state-of-science regional Weather Research and Forecasting model with Chemistry (WRF/Chem) v3.7 is applied over the continental U.S. (CONUS) for decadal simulations under current and future climate/emission scenarios. The results are presented in two parts. In this Part I study, current decadal (2001-2010) simulations of WRF/Chem over the CONUS are performed using the emissions based on the U.S. National Emission Inventories (NEI) and initial and boundary conditions derived from an advanced earth system model (i.e., the Community Earth System Model). The results are evaluated against available surface and satellite observations to assess the model's capability in replicating the current atmosphere. The results are also compared with those from previous simulations using the Representative Concentration Pathways (RCP8.5) emissions to examine the potential improvement in model performance resulted from the use of different emission inventories (i.e., NEI vs. RCP emissions) and cumulus cloud parameterizations (i.e., the Multi-Scale Kain-Fritsch (MSKF) scheme vs. the Grell 3D scheme). The preliminary results show that WRF/Chem performs well for meteorological variables such as temperature and relative humidity at 2-m, wind speed at 10-m, precipitation, and radiation variables such as incoming shortwave radiation, outgoing longwave radiation, shortwave and longwave cloud forcing, as well as many chemical variables such as ozone (O3) and fine particulate matter. Large model biases exist for some cloud variables such as cloud droplet number concentration, cloud optical thickness, and cloud water path due to limitations of cloud dynamics/thermodynamics treatments. Model performance for precipitation, cloud variables, and O3 has been greatly improved compared to the RCP8.5 simulation due mainly to the use of the U.S. NEI and the more advanced MSKF scheme for cumulus clouds. The evaluation of decadal simulations shows an overall satisfactory model performance of WRF/Chem in representing both current climate and air quality, thus building the confidence for its application for future decadal climate projections.

Sensitivity of WRF Regional Climate Simulations to Choice of Land Use Dataset

Megan S. Mallard, Tanya L. Spero, Stephany M. Taylor

The goal of this study is to assess the sensitivity of regional climate simulations run with the Weather Research and Forecasting (WRF) model to the choice of datasets representing land use and land cover (LULC). Within a regional climate modeling application, an accurate representation of air-surface interactions is critical in order to properly simulate key variables, such as near-surface temperatures, humidity, and precipitation. It can be anticipated that these fields would be sensitive to the source of LULC data as several variables that play an important role in the surface energy and radiative budgets are set based on the LULC type in WRF. These include but are not limited to surface albedo, emissivity, and green vegetation fraction. In the present study, WRF is run with the Noah land surface model (LSM) for a set of 3-year (1988-1990) historical downscaling simulations performed with 108- and 36-km grid spacing in a nested domain set-up. WRF is driven with two LULC datasets: the 24-category U.S. Geological Survey (USGS) dataset and the 40-category 2006 National Land Cover Dataset (NLCD). A small modification to the source code of WRF version 3.6.1 within the Noah LSM is needed in order to prevent NLCD shrub and grasslands from being incorrectly treated as urban areas, and results from this modified simulation will be compared with results from the released code. However, the analysis will focus on the contrast between the USGS- and NLCD-driven runs by comparing both monthly mean temperature and precipitation as well as daily extremes.

UDINEE: EVALUATION OF URBAN DISPERSION MODELS AGAIN JU2003 DATA, AN INTERNATIONAL INITIATIVE

Miguel A. Hern ndez-Ceballos, Stefano Galmarini, Steven Hanna, Thomas Mazzola, Joseph Chang, Roberto Bianconi, Roberto Bellasio

Since the mid-1980s, the European Commission Joint Research Center (EC-JRC) has carried out a series of studies to compare and evaluate atmospheric dispersion models (ADMs) for specific source scenarios. This kind of evaluations has contributed to assess the real capacity of these systems to respond to emergency under many aspects such as: timeliness of the prediction, accuracy of the prediction& Considering the features of the urban environment, these would be likely to be major targets of a Radiological Dispersive Device (RDD) event, as high numbers of people and important infrastructural elements could be affected. In this environment, the meteorological and concentrations fields are very inhomogeneous and vary rapidly with the time, being a challenge its accurate temporal and spatial simulation with the current modelling capabilities. In this context, the EC-JRC with the support of the Defense Threat Reduction Agency (DTRA) launched last December 2014 the "Urban Dispersion International Evaluation Exercise" (UDINEE) project, with the purpose to create a framework to evaluate the atmospheric dispersion models' capabilities to simulate RDD events in an urban environment. Currently, 10 institutions from Europe, U.S. and Canada, are participating in the project, simulating the transport and dispersion of the set of puff releases carried out (by popping a balloon containing SF6 tracer) during the Joint Urban 2003 (JU2003) field experiment in Oklahoma City. The detailed data collected during this campaign are available and, currently, these are being used to evaluate the modelling results. The project is described in the present work.

Recent Updates to the Canadian Operational Regional Air Quality Deterministic Prediction System

Mike Moran¹, Sylvie Gravel², Verica Savic-Jovcic¹, Alexandru Lupu¹, Radenko Pavlovic³, Alain Robichaud², Paul Makar¹, and Qiong Zheng¹

¹Air Quality Research Division, Environment and Climate Change Canada, Toronto, Ontario, Canada

²Air Quality Research Division, Environment and Climate Change Canada, Montreal, Quebec, Canada
³Air Quality Modeling Applications Section, Environment and Climate Change Canada, Montreal, Quebec, Canada

The Regional Air Quality Deterministic Prediction System (RAQDPS) is Environment and Climate Change Canada's operational regional air quality forecast system. It is run twice daily to produce publicly-available 48-hour forecasts of hourly concentration fields of O3, PM2.5, NO2, and other pollutants. At the heart of the RAQDPS is a limited-area version of the GEM-MACH model, an on-line meteorologychemistry model, configured for a North America grid with 10-km horizontal grid spacing and 80 vertical levels. Since 2015 two major upgrades have been made to the RAQDPS. First, in June 2015 a new set of input emissions files was implemented; the new files had been built from newer Canadian and U.S. emission inventories using an improved emissions processing methodology. And second, in April 2016 the GEM-MACH model itself, which had been based on version 3 of the GEM numerical weather prediction model and version 1 of the MACH air quality module, was upgraded to a next-generation version based on version 4 of GEM and version 2 of MACH. New GEMv4 features included a new vertical coordinate, a new vertical discretization, and a mass-conserving semi-Lagrangian advection scheme. New MACHv2 features included an updated gas-phase dry deposition scheme, mass-conserving tracer advection, new chemical lateral boundary conditions, new vertical diffusion scheme, and the correction of two significant errors affecting surface emissions and gas-phase dry deposition. This presentation will provide an overview of the above updates and will show the impact of some of them. The improved forecast performance that resulted from these updates will also be described.

Investigating Causes of CMAQ Under Predictions of Sea Salt Aerosol in the San Francisco Bay Area
Su-Tzai Soong, Cuong Tran, David Fairley, Yiqin Jia, Saffet Tanrikulu
Bay Area Air Quality Management District, San Francisco, CA

Christopher Emery, Andrew Wentland, Bonyoung Koo
Ramboll Environ, Novato, CA

The Bay Area Air Quality Management District (BAAQMD) is studying the sources, chemical evolution, transport and fate of particulate matter (PM) and ozone throughout the San Francisco Bay Area (SFBA) and Central California using the Community Multi-scale Air Quality (CMAQ) model. The entire year of 2012 is simulated with CMAQ. The region presents several challenges related to complex geography, distinct seasonal air flow patterns, and a multitude of natural and anthropogenic emission sources. Model-simulated PM mass from natural ocean-borne emissions (sea salt including sodium, chloride, and sulfate) are found to be significantly under predicted during the late spring and early summer seasons when strong on-shore winds dramatically increase the generation and transport of these compounds directly into the SFBA. Measured 24-hour average sodium chloride can reach up to 10 ug/m³ on some days, while CMAQ exhibits little variation from 0.5 ug/m³. Such large under predictions lead to poor model performance for total PM mass budgets, and further affect certain secondary PM components such as nitrate. We will characterize these performance issues and investigate potential refinements, including model performance for winds that drive sea salt generation and transport, alternative formulation for sea salt emissions rates, and assumptions regarding the "surf zone". Central California improvements will be presented along with general recommendations for improving CMAQ's sea salt emissions module.

Quantifying the contribution and analyzing the chemical reactions of long-range transport and local pollutants for PM2.5 in Taiwan under winter monsoon

Ming-Tung Chuang^a* , Hui-Chun Hsu^b, Chung-Te Lee^b, Neng-Huei Lin^c, Joshua S. Fu^d, Wei-Syun Huang^c , Yun-Ru Lu^a,

^aGraduate Institute of Energy Engineering, National Central University, Chung-Li, 32001, Taiwan

^bGraduate Institute of Environmental Engineering, National Central University, Chung-Li, 32001, Taiwan

^cDepartment of Atmospheric Sciences, National Central University, Chung-Li, 32001 ,Taiwan

^dDepartment of Civil and Environmental Engineering, University of Tennessee, Knoxville, TN, USA

The regional transport from one region to another is always am important issue of research. The contribution of long-range transport (LRT) and local pollution (LP) is therefore a necessary answer when air quality control is implemented in one place. During the winter season, the air quality in Taiwan was deteriorated by the high PM_2.5 air masses originating from Asian continent. Therefore, the arising question is the quantitative contribution of LRT and LP.

In the present study, we applied 10 years (2006 to 2015) observations from several monitoring sites located at the northern tip of Taiwan. The present results, without influence of dilution, deposition, first estimated the rough contributions of LRT and LP at northern, central, and southern Taiwan, respectively. Then we used the 3 months (2010/1-2010/3) simulations results from CMAQ, considering factors of diffusion and deposition and chemical reactions, to modify the values of contributions. Finally the quantification of LRT and LP were estimated when there was prevailing northeast wind. Further, we tried to find out the chemical reactions of long-range and local pollutants, e.g., the HNO₃ evaporated from nitrate in the LRT air masses would react with local NH₃ and form fresh nitrate.

Interactive model performance evaluation tools

Doug Boyer and Weining Zhao

Most photochemical model performance evaluations are static and many programs that produce the graphical output are difficult to configure/adjust.The TCEQ has developed unique, interactive, and intuitive web-based performance evaluation tools to help assess the model's ability to replicate observations. Interactive statistics, time series, bar charts, Google Map overlays, and other graphics will be shown.

Ozone Source Apportionment Modeling to Support Policy Initiatives in the Eastern United States

Kenneth Craig, Garnet Erdakos, Lynn Baringer, Stephen Reid
Sonoma Technology, Inc., 1450 N. McDowell Blvd., Suite 200, Petaluma, CA 94954

STI conducted ozone source apportionment modeling with the Comprehensive Air Quality Model with Extensions (CAMx) to support various policy initiatives (e.g., source culpability assessment) with a focus on the eastern United States. Modeling was conducted for the 2011 ozone season for the continental United States at 12-km spatial resolution, based on emissions from EPA's 2011v6.1 modeling platform. Modeling configurations were based on those developed by EPA in recent ozone transport modeling assessments. A comprehensive source tagging strategy with over 150 tags was implemented to support various stakeholder objectives. A total of 52 power plants were tagged individually, while several dozen additional power plants were tagged across 61 geographic groups. Emissions from low-level source sectors, including non-road, on-road, and biogenics, were also tagged in 15 geographic regions. To accommodate the large number of tags that were modeled, processing was divided into three separate simulations. Daily contributions to the modeled peak 8-hr average ozone concentrations at monitoring locations throughout the eastern United States were calculated for each source tag and compiled into a database to support data mining and analysis. Example case study analyses and insights from this modeling work will be shared in the presentation.

A THREAD PARALLEL SPARSE CHEMISTRY SOLVER FOR CMAQ 5.1

George Delic, HiPERiSM Consulting, LLC, P.O. Box 569, Chapel Hill, NC 27514

For CMAQ 5.1 this presentation reports on integration of the new Chemistry Transport Model (CTM) sparse solver (FSparse) [1] as a replacement of the JSparse sparse solver method based on the work of Jacobson and Turco [2]. This has been implemented in the Rosenbrock and SMV Gear algorithms. Both algorithms are then well suited for an OpenMP thread-parallel implementation. Results for FSparse have been previously reported [1] for the U.S. EPA 4.7 version of CMAQ. However, since that time the arrival of Intel's Phi many integrated core (MIC) architecture has made the implementation of a hybrid MPI and OpenMP version of CMAQ more opportune for performance reasons. Especially after the successful results for the SMV Gear algorithm reported last year [3]. Results for thread scaling will be presented with both Rosenbrock and SMV Gear algorithms in CMAQ5.1 on both host CPU and MIC processors. Issues related to porting for either processor will be discussed in relation to performance opportunities inherent in the Intel compiler and hardware. However, along the way some peculiar interactions between this compiler and CMAQ 5.1 were found. While some bugs have been corrected in the standard distribution of CMAQ 5.1, these issues are suspected to arise from the new memory model in CMAQ 5.1 that differs significantly from 4.7.1. Also, a detailed analysis of numerical performance in the CTM is presented with investigation of the residual error in progress of time step iterations. The chemistry time step calculation in the Rosenbrock or SMV Gear algorithms uses the RMS error estimate in each iteration but this allows larger errors in the infinity norm. If FSparse is found to be more accurate then relaxation of the time step error criterion could result in even greater reduction of wall clock time.
[1] G. Delic, Annual CMAS conference, 2012 and 2013.
[2] M. Jacobson and R.P. Turco (1994), Atmos. Environ. 28, 273-284
[3] T. Clune, M.R. Damon, and G. Delic., Annual CMAS conference, 2015.

Examining Changes to Extreme Temperatures and Precipitation Across the U.S. Through 2100

Tanya L. Spero, Megan S. Mallard, Stephany M. Taylor, and Christopher G. Nolte

Extreme weather events can have myriad devastating societal impacts, including effects on human health, the environment, agriculture, land use, and the economy. This presentation will show the evolution of projected changes to extreme weather events under two climate change scenarios. Simulations from the Community Earth System Model (CESM) are dynamically downscaled using the Weather Research and Forecasting (WRF) model. Two of the Representative Concentration Pathways (RCPs) are chosen: RCP 4.5, which is a modest warming pathway, and RCP 8.5, a more extreme warming pathway. The climate change following each of the RCPs is relative to recent historical climate simulated by CESM and downscaled by WRF. The analysis will focus on differences in projected extremes to temperature and precipitation across the U.S. through 2100.

Can machine learning features identify fitness of meteorology simulations for application to air quality

Robert Nedbor-Gross, Graduate Research Assistant, University of Florida, Gainesville, FL, USA.

Barron H. Henderson, Assistant Professor University of Florida, Gainesville, FL, USA.

Jorge E. Pachon. Associate Professor Universidad de La Salle, Bogota, Colombia

Standard meteorological model performance evaluation (MPE) can be insufficient in determining "fitness" for application to air quality modeling. Typical MPE compares predictions of temperature, wind, and humidity to community-based thresholds. Conceptually, these thresholds are used to measure the model's capability to represent mesoscale features that cause variability in pollution from emissions. Thus, a method that instead examines features through machine learning could provide a better estimate of fitness. This work compares measures of fitness from standard MPE analysis to those derived from machine learning. Both threshold MPE and feature analysis are then evaluated as predictors of "acceptable" air quality model performance. We expect feature-based meteorological fitness to provide a more accurate measure of performance.

Meteorology and air pollution simulations for Bogota, Colombia provide an ideal case study. This case is particularly interesting because the complex local topography presents challenges for the Weather Research and Forecasting (WRF) model. A k-means cluster analysis identified 4 dominant meteorological features associated with wind speed and direction. The model predictions are able to pass several MPE thresholds, but as expected show poor performance for wind direction error. By comparison, the model is more likely to reproduce the cluster analysis features. The four observationally-derived features have clear relationships with particulate matter concentrations, which suggests that reproducing the features will indicate better air quality model performance.

Meteorological model performance is fundamentally important to air quality modelling, but threshold MPE analysis may insufficiently describe fitness. We demonstrate the relationship between meteorological MPE, thresholds, and feature prediction. Then, we'll discuss further the relationship between the evaluation techniques and model fidelity. Feature-based analysis may be particularly useful for high resolution modelling (1km or less). In these cases, small-scale variability could be spatially offset and cause poor performance with standard threshold analysis even when the general pattern is well reproduced.

Modeling green infrastructure land use changes on future air quality in Kansas City

Yuqiang Zhang, Jesse Bash, Shawn Roselle, Alice Gilliland, Angie Shatas, Robin DeYoung, and Jamie Piziali

US EPA (Office of Research and Development, Office of Air and Radiation, Office of Water)

Green infrastructure can be a cost-effective approach for reducing stormwater runoff and improving water quality as a result, but it could also bring co-benefits for air quality: less impervious surfaces and more vegetation can decrease the urban heat island effect, and also result in more removal of air pollutants via dry deposition with increased vegetative surfaces. Lower surface temperatures can also decrease ozone formation through the increases of NOx titration; however, lower surface temperatures also decrease the height of the boundary layer resulting in more concentrated pollutants within the same volume of air, especially for primary emitted pollutants (e.g. NOx, CO, primary particulate matter). To better understand how green infrastructure impacts air quality, the impact on all of these processes must be considered collectively.

In this study, we use a comprehensive coupled meteorology-air quality model (WRF-CMAQ) to simulate the influence of planned green infrastructure land use changes in Kansas City (KC) on regional meteorology and air quality. Current and future land use data was provided by the Mid-America Regional Council for 2012 and 2040 (projected land use due to green infrastructure implementation). These land use datasets were incorporated into the WRF-CMAQ modeling system allowing the modeling system to propagate the changes in vegetation and impervious surface coverage on meteorology and air quality. The WRF-CMAQ model was run for the conterminous US using a 12km by 12km horizontal grid spacing and a nested finer scale simulations using 4km by 4km. We found that the average 2-meter temperatures (T2) during summer (June, July and August) are projected to slightly decrease over the downtown of KC and slightly increase over the newly developed regions surrounding the urban core. The planetary boundary layer (PBL) height changes are consistent with the T2 changes: they decreased over the downtown somewhat, and increased over the newly developed areas. We also saw relatively small decreases in O₃ in the downtown area for the mean of all hours as well as for the maximum 8 hour average (MDA8), corresponding with the changes in T2 and PBL height. NO_x increases during the night in the future case, which is likely caused by the decrease of PBL height and consistent with the modeled changes in T2. However, we also found relatively small PM_2.5 increases over KC, especially over the downtown areas, with the largest contribution from components of organic carbon (OC), elementary carbon (EC), non-anion dust (SOIL), and unspeciated PM. More diagnostic analysis is needed to further investigate how these land use changes affect different processes (such as the dry deposition) and future work is needed to investigate the impact of temperature reduction on energy demand and anthropogenic emissions.