Using multi-temporal Landsat ETM+ data to monitor the plague of oriental migratory locust

This paper aims to evaluate the effectiveness of using Landsat ETM+ data to identify the extent and severity of locust damage. Two cloud-free Landsat ETM+ images of the study area, taken before and after peak locust plague, were compared to determine the extent and severity of the 2002 locust plague according to the decrease of the Normalized Difference Vegetation Index (NDVI) derived from the two images. The results showed that the locust plague could be classified into heavy, moderate and light damage degrees based on the NDVI value decrease calculated by each pixel, which further evaluated its accuracy by extensive ground survey data. Locust plague can be identified with 98% and 92% accuracy for determining geographic extent and severity respectively using Landsat ETM+ data.