Testing The Impact Of Interactively Coupling AMeteorological Model To An Air Quality Model.

Abstract

A coupled air quality / meteorological model was modified to test the accuracy of using the transport code in the meteorological model versus the transport code in the air quality model to move gases. Full model simulations were performed with both transport codes, and predictions were compared to data from the Southern California Air Quality Study period of August 26 - 28, 1987. Both transport codes resulted in similar overall prediction accuracy even though the meteorological model transport code used a 6 s time step and the air quality model transport code used a 300 s time step. A third test was performed in which the air quality model transport code time step was reduced to 60 s. Again, the results showed little change. Improving the accuracy of the species transport code in a coupled air quality / meteorological model does not significantly improve model accuracy simply because errors in other model processes overwhelm errors in the transport code. For example, errors in initial temperature fields, initial soil moisture fields, emissions, and other model components appear to have a much larger impact on prediction accuracy than do errors in the transport code