New Methods for Modeling and Monitoring Wildfires Using Multiple Data Sources: Smartfire v2

While forest and other wildland fires may be naturally beneficial to the ecosystem, they can be catastrophic from the human perspective. To detect, monitor, and map wildland fires, satellite observations have been used for many years, and multiple satellite-derived products with various attributes (hot spot detection vs. burn scar, polar orbiting vs. geostationary, etc.) have been developed. Each product provides something useful that might not be available from the others. In some jurisdictions, land management agencies also provide useful fire information from ground reports and aircraft surveillance. The Smartfire system was originally developed in 2007 to combine fire activity data from satellite observations and ground reports into a single reconciled data set of best available fire activity in near real-time. Smartfire v1 combines satellite observations from the National Oceanic and Atmospheric Administration's Hazard Mapping System and ground reports from the National Interagency Fire Center. Smartfire v1 has been used for real-time fire and smoke modeling and for developing the wildland fire emissions inventory for the U.S. Environmental Protection Agency. Smartfire has also proven beneficial during emergencies when rapid response is needed, such as large wildfires with impacts on residential areas. The Smartfire system has been comprehensively redesigned to be more flexible, expandable, and accurate. Smartfire v2 provides a framework that intelligently combines an indefinite number of fire information data sources into custom-reconciled data streams for modeling, monitoring, emergency response, retrospective analysis, and emissions inventory development. We will present an overview of Smartfire v2 and assess how well it performs using only data sets available in near-real time.