Huot Fantine ; Hu R. Lily ; Goyal Nita ; Sankar Tharun ; Ihme Matthias ; Chen Yi-Fan

Abstract

Predicting wildfire spread is critical for land management and disasterpreparedness. To this end, we present `Next Day Wildfire Spread,' a curated,large-scale, multivariate data set of historical wildfires aggregating nearly adecade of remote-sensing data across the United States. In contrast to existingfire data sets based on Earth observation satellites, our data set combines 2Dfire data with multiple explanatory variables (e.g., topography, vegetation,weather, drought index, population density) aligned over 2D regions, providinga feature-rich data set for machine learning. To demonstrate the usefulness ofthis data set, we implement a neural network that takes advantage of thespatial information of this data to predict wildfire spread. We compare theperformance of the neural network with other machine learning models: logisticregression and random forest. This data set can be used as a benchmark fordeveloping wildfire propagation models based on remote sensing data for a leadtime of one day.