||Nested Grid Mesoscale Atmospheric Chemistry Model.
Pleim, J. E. ;
Chang, J. S. ;
Zhang., K. ;
||State Univ. of New York at Albany. Atmospheric Sciences Research Center.;Environmental Protection Agency, Research Triangle Park, NC. Atmospheric Research and Exposure Assessment Lab.
Acid rain ;
Mathematical models ;
Atmospheric chemistry ;
Mesoscale phenomena ;
United States ;
Regional Acid Deposition Model ;
||Some EPA libraries have a fiche copy filed under the call number shown.
A nested grid version of the Regional Acid Deposition Model (RADM) has been developed. The horizontal grid interval size of the nested model is 3 times smaller than that of RADM (80/3 km = 26.7 km). Therefore the nested model is better able to simulate mesoscale atmospheric processes while maintaining consistency with larger-scale features. The nested model uses dynamic boundary conditions along inflow boundaries which are interpolated from coarse grid model (RADM) results. Since the nesting is one-way, the outflow boundary conditions are specified to minimize numerical reflections. Three-day wet deposition amounts of sulfate and nitrate simulated by the nested grid version of RADM as well as the coarse grid version were compared to measurements made by the regional scale network during the Oxidizing and Scavenging Characteristics of April Rains (OSCAR) IV experiment. Both models were able to predict wet depositions of sulfate to within a factor of 2 for over 80% of the sampling sites. Nitrate predictions were slightly worse with just over 70% of the pairs within a factor of two for both models. Although the nested grid model showed no improvement over the coarse grid model in terms of point by point comparisons, it did demonstrate the ability to predict a more realistic range and spatial variability of wet deposition amounts. Plots of predicted deposition fields show that the nested model predicted values similar to the observations in their vicinity more often than did the coarse grid model. Nested model simulations of wet deposition were also compared to measurements made by a high density sampling network over a 110 x 110 km area of northeastern Indiana. Most of the spatial features of the observed distribution were well simulated by the model demonstrating the nested model's ability to resolve processes on scales which are subgrid to the coarse grid model. The predictions of wet depositions by both models were in better agreement with observations than were
predictions of precipitation. Since the OSCAR IV experiment was characterized by very efficient washout, the rate of wet deposition was often primarily controlled by the rate of supply of airborne material to the precipitating system rather than aqueous oxidation and precipitation processes. Therefore, errors in precipitation rate had little impact on wet deposition predictions. (Copyright (c) 1991 by the American Geophysical Union.)