EOMF-11. PyFDEO - a new Python tool for Fire Danger prediction from Earth Observations

Abstract
Here we present PyFDEO - a new Python tool for predicting fire danger from Earth observations. PyFDEO generates a forecasted fire danger map from automatic model feeding and prediction using near-real-time data. It is intended to help firefighters plan and optimize resource allocation across the Conterminous United States (CONUS). The tool collects available remotely-sensed data on Vapor Pressure Deficit (VPD) from the Atmospheric Infrared Sounder (AIRS), Surface Soil Moisture (SSM) from the Gravity Recovery and Climate Experiment (GRACE), and Enhanced Vegetation Index (EVI) from the Moderate Resolution Imaging Spectroradiometer (MODIS) to feed probabilistic models. The models are land cover specific and were trained and validated using National Fire Program Analysis Fire-Occurrence Database (FPA FOD) previously. The output of our tool is a CONUS-wide, 0.25-degree gridded cell map of two-month forecasted fire danger, i.e. whether it is expected to be below normal, normal, or above normal. This open-source tool can also be adapted for use in other countries and may help society to respond to wildfires that are increasing under climate change.