NOAA and EPA have created an Air Quality Forecast (AQF) system. This AQF system links an adaptation of the EPA’s Community Multiscale Air Quality Model with the 12 kilometer ETA model running operationally at NOAA’s National Center for Environmental Predication (NCEP). One of the major components of the linkage between these two models is the new pre-processor to CMAQ (PREMAQ). PREMAQ combines the functionality of the Meteorology-Chemistry Interface Processor (MCIP) and portions of the Sparse Matrix Operator Kernel Emissions (SMOKE) processing system. This paper explains the emissions processing required for the AQF system which is a major component of PREMAQ.

Ideally, real-time emissions data would be used in the real-time AQF system. However, the collection and transmission of such data in real-time is not yet feasible. The EPA creates a national emissions inventory (NEI) on a 3-year basis that is not available until several years after the year of interest. For the 2004 ozone season, the most recently available NEI is for 2001, based largely on the 1999 with some growth and year specific 2001 information included. The 2001 NEI was used as the basis for developing the inventory used in the AQF system. We will discuss modifications and updates made to the 2001 NEI for the 2004 AQF system.

For the 2004 AQF system, we divided the emissions processing into two components: calculations that can be done a priori (independent of meteorological fields) and calculations that are done in real-time (dependent on meteorological fields). Of the four emission inventory components, area sources, point sources, mobile sources and biogenic sources, only area source calculations are made completely a priori and therefore these were done outside of the AQF system and simply “merged” in real-time. Point Source and Mobile source emission processing are partially dependent on meteorological fields but also have components that are independent of meteorological fields. Emissions from point source were assumed to have a pre-determined temporal variability (similar to area sources) but a plume rise algorithm (dependent on such fields as wind components and temperature) requires real-time meteorological fields. Therefore for point sources only the plume rise component together with the “merging” of other information needed for point sources was included in PREMAQ.

For Mobile source emissions which are a function of temperature, vehicle activity, and vehicle fleet information, there is a highly detailed emissions model for on-road sources (MOBILE6). However, it is too expensive and inefficient to use in the AQF system in real time. An efficient method for estimating mobile source emissions using the SMOKE/MOBILE 6 implementation was devised for the AQF system.

For Biogenic emissions, the Biogenic Emissions Inventory System (BEIS) was directly integrated into PREMAQ because it is very fast and is highly dependent on meteorological conditions.

Although this work was reviewed by EPA and approved for publication, it may not necessarily reflect official Agency Policy.