Evaluation of the MM5 Simulations with CASTNet Data for the 1997 Ozone Season over the Eastern United States

Jia-Yeong Ku¹, Winston Hao¹, Gopal Sistla¹, and Da-Lin Zhang²

Abstract

One of the critical components in the application of air quality simulation models (AQSM) is the need for accurate representation of meteorological fields which provide for the transport and dispersion of the pollutant species. Often the application of these AQSMs is based upon historical meteorological events, and in recent years there has been an increased application of prognostic meteorological models such as MM5 and RAMS to develop the required meteorological fields. While it is critical to evaluate these simulations prior to their use as inputs into the AQSMs, such an assessment is some what limited by the sparsity of measurements that are not utilized by the FDDA technique.

In this study, we report on the evaluation of the MM5 simulations performed at 12 kmgrid resolutionfor a 46-day period during the summer 1997 over the eastern United States. To this end we utilized meteorological data from the Clean Air Status and Trends Network (CASTNet) that were not part of the data set employed in the FDDA. The CASTNet consists of about 50 stations which are primarily situated in remote, rural regions of the eastern United States. The database considered in this study comprises of hourly temperature, wind speed and direction, and relative humidity. In general, MM5 performed very well in simulating temperature fields and less of an agreement in the case of wind fields. Also, the performance of MM5 isfound to behighly sensitive to the geographic locations of CASTNet sites. Reasons for agreement and or disagreement are discussed.

1. New State Department of Environmental Conservation, Albany, NY 12233

2. Department of Meteorology, University of Maryland, College Park, MD