Validation of and Future Plans for GOES ABI Fire Detection

Abstract:

The operational Fire Detection and Characterization Algorithm (FDCA) has been undergoing validation and refinement since the launch of GOES-16. Based on the Wildfire Automated Biomass Burning Algorithm (WFABBA), the FDCA is a single frame contextual algorithm based on concepts developed for previous generations of geostationary and polar orbiting satellites. As it has evolved it has integrated new approaches that leverage some of the advantages of the Advanced Baseline Imager (ABI), but is a member of a class of algorithms with fundamentally limited sensitivity for early detection. Despite this, large improvements in the true positive and false positive rates of detection have been achieved in recent years. The fire characteristics produced by the algorithm, fire radiative power (FRP) and instantaneous fire temperature and fire size, are more difficult to properly validate due to complicating variables including a lack of a sizeable ground truth dataset and the difficulty of comparing results between different platforms due to the compounding impacts of sensor difference, viewing geometry, and terrain, among others. Validation results for FRP will be presented. An assessment of the ability of the FDCA to perform early detection will be presented. User expectations for geostationary fire detection and the directions in which that directs algorithm development will be discussed.