Modeling Effects of Land Use Land Cover Changes on Meteorology and Air Quality in Houston, Texas, over the Two Decades

Recently, Texas Forest Service (TFS) generated a high-resolution land cover and land use (LULC) dataset from LANDSAT multi-spectral pictures taken in September 2000 for the purpose of managing the urban forest in Houston-Galveston Area (HGA). The dataset was compared to the 1992 National Land Cover Data (NLCD) to provide a reasonable LULC change scenario for 2010. To resolve problems inherent with the LULC data such as the misinterpretation of land use types due to the time of the LANDSAT pictures, the land cover data were revised based on the land use data to better simulate the meteorological responses to land-surface interactions. However, for the meteorological simulations, we backwardly projected the TFS 2000 LULC data to have a comparable set of LULC data for 1992, 2000, and 2010 because the NLCD 1992 data utilized a very different set of LULC categories, which is not compatible with the MM5 mesoscale model. We utilized these three sets of LULC data to study the overall impacts of the LULC change on air temperature, biogenic emissions, and air quality. First, we re-simulated meteorological conditions using a modified MM5 with comprehensive NOAH land surface model for the TexAQS 2000 air quality episode (August 22nd-September 1st) to study effects of LULC change on the meteorology. The newly simulated canopy and air temperatures were used to estimate biogenic emissions and to perform air quality simulations.

The results showed that the estimated biogenic isoprene emissions decreased due to reduction of vegetation and forest areas from 1992 to 2010. The air quality simulation results using these emissions and meteorological inputs realized combined effects of biogenic emissions and meteorological changes caused by different LULC data depending on the simulation conditions. In general, simulations predicted ozone concentrations to decrease with the reduced isoprene emissions when temperature increases were not much discernable. However, during the second half of the episode when air temperatures were high and local effects dominated weather, the increases in surface temperature due to the deforestation between 1992 and 2000 were much more apparent. For days from August 30th to September 1st, ozone concentrations increased over and downwind areas of the HGA eight counties in spite of the reduced biogenic emissions for 2000 compared with 1990. The continued decrease in isoprene emissions and little changes in surface temperature resulted in lower ozone concentrations in 2010 than in 2000.

The present study reveals that urban deforestation and subsequent increase of the urban heat island phenomenon result in higher ozone concentrations due to the temperature effects on the air quality. Conversely, if new trees and vegetations with minimal isoprene emissions are planted, the ozone concentrations over and downwind urban areas will be reduced due to the decrease in the urban heat island effects. Moreover, it is expected that the benefits become more effective over the regions with little influence of large direct anthropogenic VOC emission sources, such as the rural and suburban areas downwind of urban and industrial sources.