The impact of using different land cover data on wind-blown desert dust modeling results in the southwestern United States

• Published in: Atmospheric environment (1994) Vol. 41; no. 10; pp. 2214 - 2224
• Main Authors: Yin, Dazhong, Nickovic, Slobodan, Sprigg, William A.
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.03.2007; Elsevier Science
• Subjects: Air-borne particulate matter pollution; Analysis methods; Applied sciences; Atmospheric pollution; Desert dust modeling; Earth, ocean, space; Exact sciences and technology; External geophysics; Meteorology; Model performance; Particles and aerosols; PM2.5 concentrations; Pollution; United States; North America; America; USA, Southwest; USA; PM2.5 concentrations; Model performance; Air-borne particulate matter pollution; Desert dust modeling; Space remote sensing; Satellite observation; Forecast model; Land use; Coarse particle; Dust storm; Wind erosion; Arid region; Fine particle; Semi arid zone; Atmospheric pollution forecasting; Aerosols; Air pollution; Numerical simulation; Desert
• ISSN: 1352-2310; 1873-2844
• DOI: 10.1016/j.atmosenv.2006.10.061

Olson World Ecosystem (OWE) land cover data based on data sources of the 1970s and 1980s with a 10-min spatial resolution, and up-to-date Moderate Resolution Imaging Spectroradiometer (MODIS) land cover data with a 30-s resolution, were used, respectively, in modeling wind-blown desert dust in the southwest United States. The model using different land cover data sets preformed similarly in modeling meteorological field patterns, vertical profiles and surface wind and temperature, in comparisons against observations. The differences of wind and temperature at a specific time and location can be big. Compared against satellite and ground measurements, modeled dust spatial distributions using MODIS land cover data were considerably better than those using OWE land cover. Site against site comparisons of modeled and observed surface PM2.5 concentration time series showed that model performance improved significantly using MODIS land cover data. Modeled surface PM2.5 contour distributions using MODIS land cover data compared more favorably against observations. The performance statistics for modeled PM2.5 concentrations at 40 surface sites increased from 0.15 using OWE data, to 0.58 using MODIS data. This demonstrates that the survey updates and spatial resolution of land cover data are critical in correctly predicting dust events and dust concentrations. Using land cover data such as MODIS data from satellite remote sensing is promising in improving wind-blown dust modeling and forecasting.