Learning Wildfire Model from Incomplete State Observations

TL;DR

This paper develops a deep neural network model for wildfire prediction using incomplete remote sensing data, emphasizing dynamic online estimation and addressing confounding factors across multiple locations.

Contribution

It introduces a novel deep learning approach that handles incomplete observations and confounding factors for improved wildfire prediction accuracy.

Findings

Higher prediction performance than existing methods.

Effective handling of incomplete state observations.

Applicable to multiple locations with confounding factors.

Abstract

As wildfires are expected to become more frequent and severe, improved prediction models are vital to mitigating risk and allocating resources. With remote sensing data, valuable spatiotemporal statistical models can be created and used for resource management practices. In this paper, we create a dynamic model for future wildfire predictions of five locations within the western United States through a deep neural network via historical burned area and climate data. The proposed model has distinct features that address the characteristic need in prediction evaluations, including dynamic online estimation and time-series modeling. Between locations, local fire event triggers are not isolated, and there are confounding factors when local data is analyzed due to incomplete state observations. When compared to existing approaches that do not account for incomplete state observation within wildfire time-series data, on average, we are able to achieve higher prediction performances.