Multi-Pollutant Model - 7 January 2010 Version 4.7.1 of the CMAQ provides an updated version of a model that can simulate the atmospheric fate of mercury compounds and other Hazardous Air Pollutants (HAPs). Roselle et al. 2007 presented a prototype of this model and called its mechanism CB05TXHG_AE4_AQ. The mechanism included the same mercury compounds and HAPs in the CB05 based mechanisms for mercury compounds and HAPs in CMAQ version 4.6. Version 4.7 (and 4.7.1) upgrades this prototype based on the new science options available. The upgrade had to modify new options for the photochemical mechanism, aerosol physics, cloud chemistry and vertical diffusion. The result produced a new mechanism called CB05TXHG_AE5_AQ or the Multiple Pollutant mechanism and specific option settings needed to build and run this version of the CMAQ model. In the Multi-Pollutant model, phase, i.e., gas or aerosol, determines the chemical and physical processes that they undergo. Each HAPs is transported and deposited. Wet deposition is determined by precipitation rate and the Henry's Law Constant or scavenging rate of the aerosol mode. Aerosol mode also determines dry deposition velocity for aerosol phase pollutants. For the gas phase HAPS, dry deposition has a nonzero velocity if the EPI Suite program (USEPA, 2005) and the SPECTRUM Laboratory database (http://www.speclab.com/price.htm) indicate dry deposition as a fate determining process. Check the NR_DEPV.EXT and GC_DEPV.EXT files for which gas phase HAPs undergo dry deposition. In version 4.7 (and 4.7.1), five HAPs in the NR species have explicitly calculated velocities. Their values are calculated within simulations if the model is compiled the acm2_inline_txhg option and if the environment variable, CTM_ILDEPV, is set to T (true) or Y (yes). MCIP version 3.4 also provides the velocities in the METCRO2D files when the MCIP is run to calculate the maximum number of deposition velocities. The following paragraphs describe the Multiple Pollutant model, updates incorporated in version 4.7.1, and how to build the Multiple Pollutant model. GAS PHASE CHEMISTRY The CB05TXHG_AE5_AQ mechanism adapts the CB05CL_AE5_AQ mechanism by adding the reactions for mercury compounds and reactive tracers in the mercury and HAPS mechanisms available in CMAQ version 4.6. The new mechanism also adds a reaction between elemental mercury and monatomic chlorine that produces HGIIGAS (Donohoue et al. 2005). The nature of reaction changes the definition of the model species from divalent gaseous mercury because the new reaction should produce monovalent gaseous mercury. To uphold the point, the definition of HGIIGAS becomes reactive gaseous mercury representing sum of the two valence states. The monovalent state is assumed quickly converted into the divalent state but the model physics does not represent the process. Two methods compute the chemical transformation of gas phase HAPs (Table 1). The first is done within the standard numerical solver for ozone and radical chemistry such as the Euler Backward Iterative solver (Hertel et al., 1993). The chemical reactions are listed in the mech.def file. The method may affect the solution for ozone and radical concentrations if the pollutant has high enough concentrations. The second method estimates loss from chemical reactions based on the solution from ozone and radical chemistry and does not alter the ozone and radical concentrations. Luecken et al. (2006) describes both methods. The first method mentioned above treats two types of model species. Type one destroys and produces model species influencing ozone and radical concentrations. Formaldehyde, acetaldehyde, benzene and elemental mercury belong to type one. Type two does not alter ozone and radical concentrations and serves as tracers of emitted pollutants. Tracers for formaldehyde, acetaldehyde, and acrolein emissions allow determining photochemical production of the given pollutant. The type two method also applies to emissions tracers for toluene, alpha-pinene, beta-pinene and three xylene isomers (Table 2). CMAQv4.7.1 modifies the mechanism for gas phase chemistry in two ways. The first modification changes the photochemical yields of acrolein from reactions of 1,3- butadiene. Acrolein yields have been corrected to be consistent with laboratory experiments (Baker et al. 2005, Lui et al. 1999, and Tuazon et al. 1999). Yields for reactions with chlorine atoms were set to the same yield as reaction with OH In all cases, the yields have been lowered, and in the case of reaction with O3P, the yield was zero so the reaction has been deleted (Cvetanovic and Doyle 1960). The second modification corrects reaction between elemental mercury and chlorine. The correction changes the reaction type and rate constant to include the temperature and density dependence in Donohoue et al. (2005). It also changes the reaction's products to conserve chlorine based on assumption that the HGIIGAS is HgCl2. Check the mech_cb05hgtx_ae5_aq.def file for more information these two modifications. AEROSOL PHYSICS Adapting the new aerosol module resulted from adding aerosol species representing mercury and other toxic metals (Table 3). Based on the emissions of these pollutants, concentrations of the added species should be minor regarding the bulk composition of aerosols. Their concentrations then are not used to determine the rates of aerosol microphysics and deposition but they do coagulate and mode merge. The species representing particulate mercury differ from the other metallic aerosols species because photochemistry produces mass for particulate mercury. The gas species, HGIIAER, represents the produced mass. The mass goes directly into the accumulation mode and does not get divided between the fine modes unlike the CB05HG_AE4_AQ mechanism in CMAQ version 4.6. The change assumes that the surface area of the accumulation mode dominates condensation onto aerosol modes. CLOUD PHYSICS AND CHEMISTRY Adapting the cloud module added the in-cloud scavenging for metallic aerosol species and the cloud chemistry for mercury compounds. For the metallic species, the method follows the same approach as elemental carbon in the fine modes and unidentified material in the coarse mode. For mercury, in-cloud chemistry follows the method outlined in Bullock and Brehme (2002) and the mercury release notes for CMAQ version 4.6. Although the reactions involving mercury do not directly change other aqueous species, mercury chemistry can alter particulate sulfate predictions because the mercury chemistry requires using the gas phase HO2, HOCl and Cl2. These gas species affect pH and ion balance in cloud droplets based on Lin et al. (1998). Side effects increase wet deposition of each compound and possibly produce gaseous HOCL from clouds with low or no participation. See release notes on Hazardous Air Pollutants about the cloud chemistry for trivalent and hexavalent compounds. CMAQv4.6 used the aqueous phase reduction mechanism of Pehkonen and Lin (1998) and applied it to a lumped variable containing all divalent Hg species, Hg(II). The model under predicted warm season wet deposition and over predicted cool season wet deposition. Hg(II) + HO2 => Hg+ + products Hg+ + HO2 => Hg(aq) + products In CMAQv4.7 the reduction mechanism was limited to only the free divalent Hg2+ cation. This resulted in an over prediction of wet deposition. Hg(2+) + HO2 => Hg+ + products Hg+ + HO2 => Hg(aq) + products CMAQv4.7.1 corrects several problems in the aqueous chemistry routine uncovered in the CMAQv4.7 multipollutant model. In addition to the changes described in file "AQCHEM_NOTES.txt", the multipollutant version of aqchem included changes to the aqueous Hg reduction mechanism. Hg reduction was changed because the basis, Pehkonen and Lin (1998), is considered improbable. CMAQv4.7.1 uses the aqueous dicarboxylic acids (DCA) reduction mechanism in Si and Ariya (2008). All aqueous species containing divalent mercury are reduced assuming that AORGCJ represents DCA. The rate is scaled by the cosine of the solar zenith angle to include the observed dependence on the sunlight. The reaction applies to all aqueous phase divalent species but these species are treated more explicitly than in v4.6 to improve the aqueous phase mass balance. These changes resulted in improved warm season wet deposition estimates and an overprediction of cool season wet deposition. The general reduction mechanism can be written as: Hg(II) + DCA => Hg(aq) + products. VERTICAL DIFFUSION The vertical diffusion module calculates dry deposition velocities, biogenic emissions and plume rise. For the Multi-Pollutant model vertical diffusion differs in two changes from the standard option. Reading aerosol emissions adds routines for particulate mercury and other metals. Second, calculation of emissions includes a constant source of molecular chlorine over open oceans. The source is set "off" by default. The run script turns the chlorine source on by setting the environment variable, CTM_CL2_SEAEMIS, to true. This source mimics production of molecular chlorine observed in the marine boundary layer that may come from the heterogeneous chemistry of sea salt aerosols (Spicer et al. 1998). Knipping and Dadbub (2002 and 2003) have proposed a reaction mechanism for the production. We do not attempt to use the mechanism because its reaction efficiencies are not well defined and because the CB05CL_AE5_AQ mechanism does not include all the nitro- chlorine compounds needed. BUILDING AND RUNNING As mentioned above, building CMAQ with the Multi-Pollutant mechanism requires different build settings than the standard version of CMAQ. Table 4 shows the build settings needed to construct CCTM using this mechanism with its EBI solver. Settings not specified in Table 4 remain the same as the standard version. NOTE that the smvgear and ros3 options for the chem module also work for this mechanism. NOTE: You must use the I/O API version 3.1beta or newer to support the larger number of variables required by the Multipollutant version of CMAQ. To run the CMAQ with the Multi-Pollutant mechanism, the user needs emissions files containing rates listed in the GC_EMIS.EXT, NR_EMIS.EXT and AE_EMIS.EXT files. The files contain emissions that are not identical to the original CB05CL mechanisms. A user must complete SMOKE processing with correct ancillary files such as GSREF and GSPRO and the merged NEI/Toxics database. To obtain these items contact the CMAS Help desk at www.cmascenter.org. References Baker, J.; Arey, J.; Atkinson, R., Formation and reaction of hydroxycarbonyls from the reaction of OH radicals with 1,3-Butadiene and Isoprene. Environmental Science and Technology 2005, 39, 4091-4099. Bullock, O. R. and K. A. Brehme, 2002. Atmospheric mercury simulation using the CMAQ model: formulation description and analysis of wet deposition results. Atmospheric Environment, 36, 2135-2146.. Carlton, A.G., et al., CMAQ model performance enhanced when in-cloud SOA is included: comparisons of OC predictions with measurements. Environ. Sci. Technol., 2008. 42(23): p. 8798-8802. Cvetanovic, R. J.; Doyle, L. C., Reaction of oxygen atoms with Butadiene. Canadian Journal of Chemistry 1960, 38, 2187-2195. Donohoue, D.L., Bauer, D. and Hynes, A.J., 2005. Temperature and Pressure Dependent Rate Coefficients for the Reaction of Hg with Cl and the Reaction of Cl with Cl: A Pulsed Laser Photolysis-Pulsed Laser Induced Fluorescence Study Journal of Physical Chemistry A, 109, 7732-7741. Hertel, O., R. Berkowicz and J. C. Hov, 1993. Test of two numerical schemes for use in atmospheric transport-chemistry models. Atmospheric Environment, 27, 2591-2611. Knipping, E. M.and Dabdub, D. J., 2003. Impact of Chlorine Emissions from Sea- Salt Aerosol on Coastal Urban Ozone. Environmental Science and Technology 2003, 37, 275-284. Knipping, E. M.and Dabdub, D. J., 2002. Modeling Cl2 formation from aqueous NaCl particles: Evidence for interfacial reactions and importance of Cl2 decomposition in alkaline solution. Journal of Geophysical Research, 107 (D18), 4360-4390. Lin, C.-J. and Pehkonen, S.O., 1998. Oxidation of elemental mercury by aqueous chlorine: implications for tropospheric mercury chemistry. Journal of Geophysical Research, 103 (D21), 28,093-28,102. Liu, X.; Jeffries, H. E.; Sexton, K. G., Hydroxyl radical and ozone initiated photochemical reactions of 1,3-Butadiene. Atmospheric Environment 1999, 33, (3005-3022). Luecken, D. J., W. T. Hutzell and G. L. Gipson 2006. Development and analysis of air quality modeling simulations for hazardous air pollutants. Atmospheric Environment, 40, 5087-5096. Pehkonen, S.O.; Lin, C-J. Aqueous photochemistry of mercury with organic acids, J. Air Waste Manage. Assoc. 1998, 48, 144-150 Roselle, S.J., D.J. Luecken, W.T. Hutzell, O.R. Bullock, G. Sarwar, and K.L. Schere, 2007. Development of a multipollutant version of the community multiscale air quality (CMAQ) modeling system. Extended Absract for the 5th Annual CMAS Conference, Chapel Hill, NC. Sarwar, G., D. Luecken, G. Yarwood, G. Whitten, B. Carter, 2008. Impact of an updated Carbon Bond mechanism on air quality using the Community Multiscale Air Quality modeling system: preliminary assessment. Journal of Applied Meteorology and Climatology, 47, 3-14. Si, L.; Ariya, P.A. Recition of oxidized mercury species by dicarboxylic acids (C2-C4): kinetic and product studies, Environ. Sci. Tehcnol., 2008, 42, 5150- 5155 Spicer, C. W., Chapman, E. G., Finlayson-Pitts, B. J., Plastridge, R. A., Hubbe, J. M., Fast, J. D. and Berkowitz, C. M., 1998. Unexpectedly high concentrations of molecular chlorine in coastal air. Nature, 394, 353-356. Tuazon, E. C.; Alavarado, A.; Aschmann, S. M.; Atkinson, R.; Arey, J., Products of the gas-phase reactions of 1,3-Butadiene with OH and NO3 radicals. Environmental Science and Technology 1999, 33, 3586-3595. US Environmental Protection Agency, cited 2005. Estimations Programs Interface for Windows (EIPWIN), version 3.12. Available online at http://www.epa.gov/opptintr/exposure/pubs/episuitedl.htm Tanaka P.L., D.T. Allen, E.C. McDonald-Buller, S. Chang, Y. Kimura, G. Yarwood and J.D. Neece, 2003. Development of a chlorine mechanism for use in the carbon bond IV chemistry model. Journal of Geophysical Research, 108, 4145. Yarwood, G., S. Rao, M. Yocke, and G.Z. Whitten, 2005. Updates to the Carbon Bond Mechanism: CB05. Report to the U.S. Environmental Protection Agency, RT-04- 00675. Available online at http://www.camx.com/publ/pdfs/CB05_Final_Report_120805.pdf. Table 1 Gas Phase HAP Species in CB05TXHG_AE5_AQ ===================================================================== Species Name Compound CAS# In mech.def ===================================================================== FORM total FORMALDEHYDE 50-00-0 Yes ALD2 total ACETALDEHYDE 75-07-0 Yes BENZENE BENZENE 71-43-2 Yes ACROLEIN total ACROLEIN 107-02-8 Yes BUTADIENE13 1,3-BUTADIENE 106-99-0 Yes HG Elemental Mercury NA Yes HGIIGAS Reactive Gaseous Mercury NA Yes HGIIAER Particulate Mercury Precursor NA Yes FORM_PRIMARY FORMALDEHYDE emissions 50-00-0 Yes ALD2_PRIMARY ACETALDEHYDE emissions 75-07-0 Yes ACROLEIN_PRIMARY ACROLEIN emissions 107-02-8 Yes ACRYLONITRILE ACRYLONITRILE 107-13-1 No CARBONTET CARBON TETRACHLORIDE 56-23-5 No PROPDICHLORIDE PROPYLENE DICHLORIDE 78-87-5 No DICHLOROPROPENE 1,3-DICHLOROPROPENE 542-75-6 No CL4_ETHANE1122 1,1,2,2TETRACHLOROETHANE 79-34-5 No CHCL3 CHLOROFORM 67-66-3 No BR2_C2_12 1,2DIBROMOETHANE 106-93-4 No CL2_C2_12 1,2DICHLOROETHANE 107-06-2 No ETOX ETHYLENE OXIDE 75-21-8 No CL2_ME METHYLENE CHLORIDE 75-09-2 No CL4_ETHE PERCHLOROETHYLENE 127-18-4 No CL3_ETHE TRICHLOROETHYLENE 79-01-6 No CL_ETHE VINYL CHLORIDE 7501-4 No NAPHTHALENE NAPHTHALENE 91-20-3 No QUINOLINE QUINOLINE 91-22-5 No HYDRAZINE Hydrazine 302-01-2 No TOL_DIIS 2,4-Toluene Diisocyanate 584-84-9 No HEXAMETHY_DIIS Hexamethylene 1,6-Diisocyanate 822-06-0 No MAL_ANHYDRIDE Maleic Anhydride 108-31-6 No TRIETHYLAMINE Triethylamine 121-44-8 No DICHLOROBENZENE 1,4-Dichlorobenzene 106-46-7 No ===================================================================== Table 2. Additional Gas Phase Species in CB05TXHG_AE5_AQ ===================================================================== Species Name Compound CAS# In mech.def ===================================================================== TOLU Toluene Emissions 108-88-3 YES MXYL M-Xylene Emissions 108-38-3 YES OXYL O-Xylene Emissions 95-47-6 YES PXYL P-Xylene Emissions 106-42-3 YES APIN Alpha-Pinene Emissions 80-56-8 YES BPIN Beta-Pinene Emissions 127-91-3 YES ===================================================================== Table 3. Aerosol Phase HAP species in CB05TXHG_AE5_AQ (Note that species exist in each aerosol mode) ===================================================================== String in Aerosol Species Represents PHG Mercury Compounds BE Beryllium Compounds NI Nickel Compounds CR_III Chromium (III) Compounds CR_VI Chromium (VI) Compounds PB Lead Compounds MN Manganese Compounds CD Cadmium Compounds DIESEL Diesel Emissions ===================================================================== Table 4. Option setting needed in CCTM build scrip if using EBI solver. NOTE that unspecific options remain same as CCTM with aerosols. ===================================================================== #Select a HAP mechanism set Mechanism = cb05txhg_ae5_aq #VDIFF has two options set ModVdiff = ( module acm2_txhg $Revision; ) # OR the standard CCTM setting set ModVdiff = ( module acm2_inline_txhg $Revision; ) # Select correct EBI solver # NOTE THAT ros3 and smvgear options also work set ModChem = ( module ebi_cb05txhg_ae5 $Revision; ) #AERO option required set ModAero = ( module aero5_txhg $Revision; ) #cloud processing and aqueous chemistry setting set ModCloud = ( module cloud_acm2_ae5_txhg $Revision; ) =====================================================================