SeasFire as a Multivariate Earth System Datacube for Wildfire Dynamics
This provides a new dataset for Earth system scientists and ML practitioners to improve wildfire anticipation, but it is incremental as it builds on existing Earth observation data.
The authors tackled the problem of understanding and predicting global wildfires by introducing the SeasFire datacube, a spatiotemporal dataset with 59 variables from 2001 to 2021, and demonstrated its use in modeling wildfire drivers and predicting patterns with a Deep Learning model.
The global occurrence, scale, and frequency of wildfires pose significant threats to ecosystem services and human livelihoods. To effectively quantify and attribute the antecedent conditions for wildfires, a thorough understanding of Earth system dynamics is imperative. In response, we introduce the SeasFire datacube, a meticulously curated spatiotemporal dataset tailored for global sub-seasonal to seasonal wildfire modeling via Earth observation. The SeasFire datacube comprises of 59 variables encompassing climate, vegetation, oceanic indices, and human factors, has an 8-day temporal resolution and a spatial resolution of 0.25$^{\circ}$, and spans from 2001 to 2021. We showcase the versatility of SeasFire for exploring the variability and seasonality of wildfire drivers, modeling causal links between ocean-climate teleconnections and wildfires, and predicting sub-seasonal wildfire patterns across multiple timescales with a Deep Learning model. We publicly release the SeasFire datacube and appeal to Earth system scientists and Machine Learning practitioners to use it for an improved understanding and anticipation of wildfires.