Improved Wildfire Spread Prediction with Time-Series Data and the WSTS+ Benchmark

The introduction of the WSTS+ benchmark marks a significant advancement in wildfire spread prediction, leveraging deep neural networks (DNNs) to enhance accuracy. By incorporating four additional years of historical wildfire data, the WSTS+ benchmark not only doubles the unique years of data available but also broadens its geographic scope. This expansion is crucial as it allows for more comprehensive modeling strategies. The research demonstrates that models utilizing time-series input yield the highest accuracy, underscoring the necessity of this method in future wildfire research. As wildfires continue to pose a growing threat globally, improving predictive capabilities through advanced benchmarks like WSTS+ is essential for effective management and response strategies.