Detection of land cover changes using MODIS 250 m data

Abstract

The Vegetative Cover Conversion (VCC) product is designed to serve as a global alarm for land cover change caused by anthropogenic activities and extreme natural events. MODIS 250 m surface reflectance data availability was limited both spatially and temporally in the first year after launch due to processing system constraints. To address this situation, the VCC algorithms were applied to available MODIS 250 m Level 1B radiance data to test the VCC change detection algorithms presented in this paper. Five data sets of MODIS Level 1B 250 m data were collected for the year 2000, representing: (1) Idaho–Montana wildfires; (2) the Cerro Grande prescribed fire in New Mexico; (3) flood in Cambodia; (4) Thailand–Laos flood retreat; and (5) deforestation in southern Brazil. Decision trees are developed for each of the VCC change detection methods for each of these six cases. These decision trees are to be used for updating the look-up tables required by the VCC production code. For these change detection cases, the VCC change detection methods worked reasonably well. In the Idaho–Montana wildfire case, a fire perimeter polygon data set compiled by the USDA Forest Service was used to validate the output of the VCC change detection methods. Although the VCC output identified only 32% of the burned pixels within the ground observed Idaho–Montana fire perimeter polygons, the detection accuracy of the VCC output did reach 99% when the VCC product is considered as an alarm system identifying the occurrence of the change in an area. For other cases, the detection accuracy in per-pixel terms of the VCC output ranges from 55% to 90% against reference change bitmaps that were created by image interpretation. Look-up tables created with AVHRR and Landsat Thematic Mapper data require modifications for the MODIS data due to differences in radiometric response between MODIS and the heritage instruments. The applications presented in this paper also evaluate the relative performance of each of the five change detection methods used as VCC algorithms. Conclusions reached in this paper will be used for future refinement of the VCC product.