Illinois State Water Survey - CAQIMS - Components, University of Illinois at Urbana-Champaign

A Basis for Achieving Economic, Societal and Environmental Goals in Illinois
Xin-Zhong Liang

Emission Inventory Model (EIM)
One of the crucial datasets necessary for realistic modeling is a detailed description of the fluxes of the gases and particulates that significantly affect air quality. The regional air quality model requires emission flux values that are appropriately speciated, temporalized and gridded. We are using the computer program SMOKE (Coats and M.R. Houyoux 1996; Houyoux et al. 2000) to produce those values for the modeling domain.

The “raw” datasets are inventories that consist of the best available emissions estimates for important pollutant species. In the case of the USEPA 1996 National Emissions Inventory (NEI96) and 1999 National Emissions Inventory (NEI99) that we are currently using, those species are: oxides of nitrogen (NOx), volatile organic compounds (VOC), carbon monoxide (CO), sulfur dioxide (SO2), elemental carbon (EC), organic carbon (OC), ammonia (NH3), and size-classified particulates (PM-2.5 and PM-10). Most of the NEI96 and NEI99 emissions data are given as a tons-per-day averaged value for the May 1st through September 30th “ozone season.” While most of those species can be used directly, the NOx fluxes must be divided into NO2 and NO, and the VOC fluxes must be appropriately allocated among the ten chemical species that are used in the air quality model. Temporalization consists of dividing the daily total flux for each emissions source into 24 one-hour portions that vary according to typical patterns for industrial, transportation and other activity classes. Finally, the gridding process partitions the combined hourly speciated emissions into a 3-dimensional array of cells that covers the modeling domain using the Lambert conformal conic projection system. At this stage, the emissions are ready for input into the regional air quality model.

Emissions Processed emissions for NOx and VOC, for one hour of one day of a 90 day simulation for the summer of 1995, are shown in the figures. Only the emissions in the surface layer are shown, although the model computes the vertical distribution of emissions, which is especially important for elevated point sources. NOx is primarily emitted by combustion processes so there are significant elevated NOx emissions from power plants. While automobile NOx emissions have been greatly reduced by catalytic converters, there are numerous ground level NOx sources especially from diesel engines. Like NOx ground level sources, there are significant VOC sources from combustion engines, so high emission contours of both emission types trace out major cities. VOC emissions also derive from natural biogenic sources predominantly in forested areas. The VOC emission contours clearly identify what has been termed the "Ozark volcano" from forests in that area.

 
CAQIMS Components
 
CAQIMS Home