Adapting CMAQ Deposition Fields for Critical Loads Analyses
Robin L. Dennis* and Kristen Foley
U.S. Environmental Protection Agency
Atmospheric Modeling and Analysis Division
National Exposure Research Laboratory
MD: E243-02
Research Triangle Park, NC 27711
A complete deposition budget, wet plus dry, is needed for critical loads analyses. It is desirable for the budget to be as accurate as possible. An air quality model can provide dry deposition estimates that have the advantage of including all species depositing and having complete spatial coverage, although not resolving well complex terrain. However, the error in the modeled precipitation amounts used by an air quality model is deemed unacceptably high. While there are errors in the emissions inventories driving the air quality models, monitoring sites are sparse and interpolation of these data can miss significant emissions that exist. Thus, it is would be an improvement if the wet deposition from the air quality model could be made more accurate and these could be combined with the model’s dry deposition to support critical loads analyses. To meet this objective we developed an approach for post processing CMAQ predictions of wet deposition to reduce the model errors. The approach involves correcting the CMAQ annual precipitation amounts with PRISM (Parameter-elevations Regressions on Independent Slopes Model) annual data, which has national coverage, is based on 24-hour coop station data and accounts for orographic effects. The precipitation correction is followed by a bias correction for NO3 and NH4 deposition based on NADP data. Precipitation-corrected SO4 does not need a bias correction, suggesting that the model can do well when the emissions are well known. The bias correction is developed separately for the intermountain west together with the plains and for the eastern US. No bias is developed for the west coast because NADP data are believed to miss important orographic effects there. We present the approach applied to 2002 CMAQ simulations and show that wet deposition error is significantly reduced for SO4 and NO3 and show the residual error across space for all species. We then show the results of a sensitivity analysis to demonstrate that changing wet deposition has an acceptably small effect on dry deposition, allowing us to add the dry deposition to the adjusted wet deposition for support of critical loads analyses. We have supplied critical loads scientists the adjusted CMAQ data, with national US coverage at 36-km and eastern half coverage at 12-km resolution. We conclude by presenting some hypotheses regarding the identified biases and briefly discuss future directions.
* Tel: 919-541-2870, Email: