A New Dust Index Based on Reflective and Thermal Infrared MODIS Bands to Detect the Spatial Extent of Sand and Dust Storm in Western Asia

Abstract

Moderate Resolution Imaging Spectroradiometer (MODIS) provides appropriate images for studying of Sand and Dust Storms (SDS). Commonly used MODIS based dust indices, can’t monitor SDS more accurately and properly. In this study, the proposed new dust index is based on reflective and thermal infrared MODIS bands, which consists of a combination of bands 3, 7, 31, and 32. This index is applied to identify two dust events in western Asia ocured July 15 and 16, 2008. The results of applying this index compared to well-known MODIS based dust indices, such as Brightness Temperature Difference (BTD) index between band 32 and band 31 (BTD32-31), and band 20 and 31 (BTD20-31), and Normalize Difference Dust Index (NDDI). The results of this study indicated that the new dust index captured the spatial extent of SDS with high accuracy. Validation of these indices with SDS extracted by MODIS true color image showed that the new dust index detected SDS extent with an overall accuracy more than 88%, which was 7%, 15% and 31% higher than the results derived from BTD20-31, NDDI and BTD32-31, respectively. Also, according to SDS detected by MODIS MCD19A2 Aerosol Optical Depth (AOD) product data, the proposed index identified SDS with an overall accuracy 82%, which was 5%, 8% and 31% higher than the results derived from NDDI, BTD20-31 and BTD32-31, respectively. Our results suggest that the new dust index can effectively capture large-scale SDS and seprate dusty pixels from dust-free areas in western Asia.

Keywords

• sand and dust storms (sds)
• dust spatial indices
• aerosol optical depth (aod)
• modis
• western asia