IN41D-03:
Asynchronous Processing of a Constellation of Geostationary and Polar-Orbiting Satellites for Fire Detection and Smoke Estimation
Thursday, 18 December 2014: 8:30 AM
Edward J Hyer1, David A Peterson2, Cynthia A Curtis3, Christopher C. Schmidt4, Jay Hoffman4 and Elaine M Prins5, (1)US Naval Research Laboratory, Marine Meteorology, Monterey, CA, United States, (2)National Research Council, Monterey, CA, United States, (3)Naval Research Lab, Monterey, CA, United States, (4)University of Wisconsin-Madiso, Madison, WI, United States, (5)University of Wisconsin-Madiso, CIMSS Consultant, Madison, WI, United States
Abstract:
The Fire Locating and Monitoring of Burning Emissions (FLAMBE) system converts satellite observations of thermally anomalous pixels into spatially and temporally continuous estimates of smoke release from open biomass burning. This system currently processes data from a constellation of 5 geostationary and 2 polar-orbiting sensors. Additional sensors, including NPP VIIRS and the imager on the Korea COMS-1 geostationary satellite, will soon be added. This constellation experiences schedule changes and outages of various durations, making the set of available scenes for fire detection highly variable on an hourly and daily basis. Adding to the complexity, the latency of the satellite data is variable between and within sensors. FLAMBE shares with many fire detection systems the goal of detecting as many fires as possible as early as possible, but the FLAMBE system must also produce a consistent estimate of smoke production with minimal artifacts from the changing constellation. To achieve this, NRL has developed a system of asynchronous processing and cross-calibration that permits satellite data to be used as it arrives, while preserving the consistency of the smoke emission estimates. This talk describes the asynchronous data ingest methodology, including latency statistics for the constellation. We also provide an overview and show results from the system we have developed to normalize multi-sensor fire detection for consistency.
