A simple method for decomposing surface-based ozone observations into a background component and a locally produced component is applied to six years of ozone observations from the EPA non-attainment region around Houston, Texas. The climatology of ozone exceedances (per EPA guidelines) is presented both in full and subcategorized by the importance of each component.

Local meteorological parameters, including wind speed and direction, precipitation, temperature, and parameters specific to the sea-breeze circulation are compared with both the background and locally produced components via linear regression to determine relationships between these parameters and each component of the total ozone. Principal component analysis is utilized to introduce the background wind field across the south-central United States as a parameter for comparison and a statistical model of the sea-breeze circulation is utilized to introduce sea-breeze effects.

Background winds, in the form of selected high-order principal components, and aspects of the sea-breeze circulation such as timing and intensity are found to be quite useful in describing and predicting both background and locally produced ozone in southeast Texas. Significant relationships exist not only on days with high ozone but up to two days in advance of high ozone days as well. The utility of these results is discussed as both a means of statistically forecasting ozone and a method of determining how much of Houston's ozone problem might be mitigated through local controls.