Characterizing Annual Meteorological Modeling Performance for Visibility Improvement Strategy Modeling in the Southeastern US

The Visibility Improvement State and Tribal Association of the Southeast (VISTAS) is responsible for technical analyses and planning activities associated with the management of visibility and other regional air quality issues in the southeastern U.S. VISTAS analyses will support the states in their responsibility to develop, adopt, and implement their individual state air quality implementation plans for regional haze.

VISTAS has designed an extensive “one-atmosphere” modeling exercise to evaluate emissions contributions to fine particulate matter pollution (PM2.5) and visibility in the southeastern U.S. that consists of 3 main components: 1) meteorological modeling with MM5, 2) emissions modeling with SMOKE, and 3) chemical transport modeling with CMAQ. The meteorological modeling component is the primary focus of this paper.

In order to consider numerous areas in the southeastern U.S. simultaneously, a representative year is simulated. VISTAS chose to model the calendar year of 2002 using a 36-km grid for the national domain and a 12-km grid for the eastern U.S. Prior to the annual simulation, a series of sensitivity tests were conducted to identify the optimal configuration for the southeastern U.S. This testing established the model configuration for the annual simulation.

Analyses of the annual (2002) MM5 simulation are presented, including an examination of model performance characteristics at 36-km versus 12-km grid spacing. A comprehensive model analysis is performed, including the development of a unique set of metrics that allow evaluators to efficiently gauge the strengths and weaknesses of this substantial data set.

Model performance for the 12-month simulation in the southeastern U.S. is deemed adequate for visibility modeling. By and large, performance traits evident in the sensitivity modeling are seen in the annual simulation as well. Some of those characteristics include a dampened diurnal temperature cycle, a wintertime cold bias, and overestimates of summer precipitation in the southeastern U.S.