Verification of the WRF model during a high ozone event over Houston, TX

Abstract

High ozone values were observed in Houston, TX during August 25 - September 1, 2000. A comparison of WRF data with observations and MM5 data was conducted to determine the WRF model's performance in simulating the meteorological conditions that affected ozone. The most substantial errors in the WRF simulation were a cold bias during the observed high temperatures in the afternoon, and the timing of the sea breeze cycle. It was less accurate than the MM5 model for these aspects. The WRF model was more accurate at night, predicting the height of the boundary layer, cloud cover, the nocturnal low-level jet, and profiler winds in the lower levels. Multiple simulations of the WRF model were also run to test the sensitivity of the model to different parameters. The model was run with a different radiation scheme, different values for the diffusion schemes, different values for the levels of acoustic damping, and a different order for the time-splitting scheme. Most of the sensitivity runs only slightly altered the meteorological characteristics. The model run that changed the radiation to the Goddard scheme from the Dudhia scheme exhibited the largest changes overall.