Use of MODIS data to monitor the oriental migratory locust plague

Abstract

Outbreaks of the oriental migratory locust Locusta migratoria manilensis are a serious problem in China. This study evaluated the potential of MODIS data to monitor locust outbreaks in a heavily damaged area in Nadagang State Farm, Hebei Province. Two cloud-free MODIS images of the study area, taken before and during peak damage, were compared to determine the geographical extent and severity of the 2002 locust plague. Ground survey data were used to evaluate locust damage indicated by the normalized difference vegetation index (NDVI) derived from the MODIS data. The results showed that the NDVI reliably distinguished between before and peak damage situations for each category of damage. NDVI values were calculated pixel by pixel and changes at the same sites were compared. Areas where NDVI decreased could be clearly mapped and classified into light, moderate, and heavy damage categories according to the decrease in their NDVI value. Compared to extensive ground surveys, locust damage could be identified with 88.8% accuracy from MODIS data obtained before and during peak.