The 250 m global land cover change product from the Moderate Resolution Imaging Spectroradiometer of NASA ’ s Earth Observing System

One of the products to be derived from data acquired by theModerate Resolution Imaging Spectroradiometer (MODIS) onboard the National Aeronautics and Space Administration (NASA) Earth Observing System (EOS) will identify locations where land cover changes attributable to human activities are occurring. The product aims to serve as an alarm where rapid land cover conversion can subsequently be analysed with higher resolution sensors such as Landsat 7. This paper describes the production procedure and change detection algorithms for the 250m MODIS land cover change product. Multiple change detection algorithms, including three spectral methods and two texture methods, are utilized to generate the product from the two 250m MODIS bands. The change detection methods are implemented with look-up tables (LUTs) which are initially generated using data from the Advanced Very High Resolution Radiometer (AVHRR) and Landsat Thematic Mapper (TM) until MODIS data become available. The results from the Ž ve methods are combined to improve conŽ dence in the product.We test the performanceof each method against several sets of simulated MODIS data derived from Landsat TM image pairs. The test data represent tropical deforestation, agricultural expansion and urbanization. The commission errors of the Ž ve methods and the combinationare approximately 10%, with reasonable omission errors less than 25%.