Creating emissions with SMOKE that are ready for input to an AQM must always include merging the hourly emissions created during temporal processing with the gridding matrices and the speciation matrices. In addition, for point sources for CMAQ, creating the model-ready emissions must also include merging with the layer fractions (see Section 2.14.1, “Computing layer fractions for CMAQ”), and for UAM-based models it must include creating the ASCII elevated file (see Section 2.14.2, “Creating an elevated-source file for UAM and CAM_X”). The Smkmerge program performs these processing steps using vector-matrix multiplication to combine the matrices and layer fractions with the hourly emissions vectors from the Temporal program.

If the overall SMOKE processing setup includes running the nonroad mobile category (or other source categories) as separate runs, then Smkmerge cannot be used to combine all source categories into a single output file. Instead, the Mrggrid program would combine the model-ready files from the individual source categories; for example, from separate SMOKE (including Smkmerge) runs for stationary area/nonpoint, nonroad mobile, windblown dust, wildfire, on-road mobile, and point sources. There is no limit to the number of model-ready files that Mrggrid can combine into a single model-ready file, and the input files can be 2-D or 3-D.

Smkmerge can be run for any or all SMOKE source categories, but it can use only one of each SMOKE inventory type (area, biogenic, mobile, and point source) per run. You can run it to create model-ready files for only one SMOKE source category (area, biogenic, mobile, or point), or you can run it to create both the individual and combined model-ready files. The following list indicates the modes in which Smkmerge can be run:

When creating model-ready emissions for any of the anthropogenic source categories, you may choose to apply one or more control matrices to the emissions to create controlled model-ready emissions. For each source category (area, mobile, or point), you can apply one multiplicative control matrix and one reactivity control matrix per run per source category. Smkmerge is the only way you can apply the reactivity control matrix to the inventory, while the multiplicative control matrix can be applied by either Smkmerge or the Grwinven program.

Many processing steps in SMOKE are independent of one another; for example, chemical speciation and temporal allocation can change without affecting one another. This independence means that when one step changes, another step does not need to be rerun in many cases. However, because Smkmerge combines the data from all of these processing steps to create the model-ready emissions, if one of the earlier steps changes, then the merging step must also be rerun. This includes rerunning Smkmerge to generate model-ready files, and if the Mrggrid program was used, also rerunning that to merge data from multiple source categories together.

Smkmerge also has the ability to input hourly emissions by day of the week and reuse days that are the same. For example, it can input separate hourly emissions files for Monday, a weekday, Saturday, and Sunday, and use these four files to generate model-ready emissions for every day in an entire month. This is accomplished using the MRG_BYDAY SMOKE option, described further in Section 4.15, “Smkmerge”. Special treatment can also be given to holidays in this case, since users generally wish to model holidays differently than other days.