The Effect of Ethene and Propene Emissions on Ozone Production in Houston during the TexAQS 2000 Field Campaign.

A coupled meteorological and chemical model, PEGASUS, was used to simulate ozone and ozone precursors in the vicinity of Houston between 28 August and 1 September during the 2000 TexAQS Field Campaign. A nested grid configuration was used with an outer grid encompassing the south-central U.S. with a grid spacing of 16 km and an inner grid encompassing the Houston metropolitan area with a 1.3 km grid spacing. Area, mobile, biogenic, and point source emission rates were based on a data set provided by the Texas Natural Resource Conservation Commission (TRNCC). The meteorological model employed data assimilation using observations from five radar wind profilers to limit forecast errors in the simulated wind fields around Houston. The simulated magnitude and distribution of ozone over downtown Houston was very similar to observations from the surface monitoring network. Observed and simulated ozone mixing ratios were as high as 140 ppb on 31 August. The model results were also in very good agreement with the spatial distribution of ozone, NOx, and NOy observed aloft by research aircraft. However, at surface stations and aircraft positions close to large point source emissions of VOCs, the simulated ozone mixing ratios were usually too low. For example, ozone mixing ratios at LaPorte were about 200 ppb on 30 and 31 August, but the predicted values at that location were between 110 and 120 ppb. Simulated VOCs, including propene and ethene, were also under-predicted. To examine whether an underestimation of the VOC point source emission rates may have contributed to the low ozone mixing ratios predicted by the model, two sensitivity simulations were performed in which the point source emission rates of ethene and propene were increased by a factor of 5 and 10. With the increased point source emission rates, ozone and VOC predictions were in better agreement with the surface and aircraft observations. Trajectories from the model indicated that the high ozone and VOC concentrations at LaPort on 30 August were from sources in the Houston ship channel; however, point source emissions close to Galveston were transported northward to LaPorte on 31 August. Because afternoon easterly flow associated with the bay-breeze was relatively weak between 28 August and 1 September, point source emissions did not significantly affect ozone concentrations in the downtown area.