The Cooperative Institute for Satellite Earth System Studies (CISESS) is working with the National Environmental Satellite, Data, and Information Service and NOAA National Centers for Environmental Information to advance the predictions of environmental conditions conducive to wildfires across the United States on sub-seasonal to seasonal scales. CISESS has resources such as cloud storage and cloud-based high-performance computing to bring together many environmental data products relevant to wildfire prediction. Additionally, this research effort will employ AI/ML techniques to aid in the prediction and assist in the understanding of these forecasts. The development of new datasets (e.g., soil moisture, atmospheric humidity, vegetation, etc.) are already underway as part of this work with the aim of enhancing the AI/ML models.
Most importantly, the scope of this research depends on the input and feedback from the wildfire community. In particular, we would like to know: (1) what forecast lead times and spatial scales are most needed to effectively mitigate wildfire risk for your organization, and (2) what environmental datasets/variables, and probabilistic forecasts thereof, would be most helpful for your organization’s efforts to predict and account for wildfire risk?

