Next Day Wildfire Spread: A Machine Learning Data Set to Predict Wildfire Spreading from Remote-Sensing Data

  title={Next Day Wildfire Spread: A Machine Learning Data Set to Predict Wildfire Spreading from Remote-Sensing Data},
  author={Fantine Huot and R. Lily Hu and Nitakshi Goyal and T. Ravi Sankar and Matthias Ihme and Yi-Fan Chen},
Predicting wildfire spread is critical for land management and disaster preparedness. To this end, we present ‘Next Day Wildfire Spread,’ a curated, large-scale, multivariate data set of historical wildfires aggregating nearly a decade of remotesensing data across the United States. In contrast to existing fire data sets based on Earth observation satellites, our data set combines 2D fire data with multiple explanatory variables (e.g., topography, vegetation, weather, drought index, population… 
Collaborative Auto-Curricula Multi-Agent Reinforcement Learning with Graph Neural Network Communication Layer for Open-ended Wildfire-Management Resource Distribution
This work builds on a recently proposed Multi-Agent Reinforcement Learning (MARL) mechanism with a Graph Neural Network (GNN) communication layer and introduces a procedural training environment accommodating auto-curricula and openendedness towards better generalizability.
  • 2022


Predictive modeling of wildfires: A new dataset and machine learning approach
Deep Learning Models for Predicting Wildfires from Historical Remote-Sensing Data
This work creates a data set by aggregating nearly a decade of remote-sensing data and historical fire records to predict wildfires and demonstrates that deep learning models can successfully identify areas of high fire likelihood using aggregated data about vegetation, weather, and topography.
Wildland Fire Spread Modeling Using Convolutional Neural Networks
The machine learning DCIGN approach was able to provide fire spread predictions at a computational cost three orders of magnitude less than current models and was shown to be capable of predicting burn maps further in the future by recursively using previous predictions as inputs to the DCIGN.
FireCast: Leveraging Deep Learning to Predict Wildfire Spread
FireCast predicts which areas surrounding a burning wildfire have high-risk of near-future wildfire spread, based on historical fire data and using modest computational resources, compared to a random prediction model and a commonly used wildfire spread model, Farsite.
A global wildfire dataset for the analysis of fire regimes and fire behaviour
This work presents and test a data mining work flow to create a global database of single fires that allows for the characterization of fire types and fire regimes worldwide.
Estimating the risk of fire outbreaks in the natural environment
This study builds predictive models to estimate the risk of fire outbreaks in Slovenia, using data from a GIS, Remote Sensing imagery and the weather prediction model ALADIN, and explores the influence of fire fuel information on the performance of the models.
A review of machine learning applications in wildfire science and management
A scoping review of ML in wildfire science and management, identified 298 relevant publications, where the most frequently used ML methods included random forests, MaxEnt, artificial neural networks, decision trees, support vector machines, and genetic algorithms.
High-severity fire: evaluating its key drivers and mapping its probability across western US forests
Wildland fire is a critical process in forests of the western United States (US). Variation in fire behavior, which is heavily influenced by fuel loading, terrain, weather, and vegetation type, leads
Characterizing wildfire regimes in the United States.
It is found that wildfires exhibit robust frequency-area power-law behavior in 18 different ecoregions, allowing for the classification of wildfire regimes for probabilistic hazard estimation in the same vein as is now used for earthquakes.
Large wildfire trends in the western United States, 1984–2011
We used a database capturing large wildfires (> 405 ha) in the western U.S. to document regional trends in fire occurrence, total fire area, fire size, and day of year of ignition for 1984–2011. Over