A multitemporal algorithm of active fire detection based on Fengyun-3C VIRR

The wildfire is one of the most destructive and uncontrollable disasters and causes huge environmental, ecological, social effects. To better serve scientific research and practical fire management, an algorithm and corresponding validation work of active fire detection based on FengYun-3C/VIRR data, which is an optical sensor onboard the Chinese polar-orbiting meteorological sun-synchronous satellite, is hereby introduced. Different from the original algorithm which utilizes the single-time image, the core content of the algorithm consists of the changes reflected by the profiles of time series curves together with the observed data. By calculating the prediction MIR value and the stable MIR value of the aim research area, fire pixels can be detected. A two-stage validation is designed and carried out with different spatial resolution data. The rough comparison is done and the compared objects include MODIS Fire and Thermal Anomalies products (MOD14A1) based on fire detection algorithm Collection 6 and results generated from the previous algorithm. The detailed validation work is conducted with the support of Landsat series data even if the different imaging time may affect the actual validation results. OLI (Optical Land Imager) onboard Landsat-8 and ETM+ (Enhanced Thematic Mapper Plus) onboard Landsat 7 are utilized.