Wildfire smoke plume segmentation using geostationary satellite imagery

TL;DR

This paper develops a deep learning approach to segment wildfire smoke plumes from satellite images, improving the quantification of smoke's health impacts amid noisy annotations and spatial uncertainties.

Contribution

It introduces a CNN-based method for wildfire smoke segmentation from satellite imagery and evaluates its effectiveness against noisy annotations using causal inference.

Findings

CNN segmentation improves smoke detection accuracy

Quantifies smoke contribution to PM2.5 levels

Reduces reliance on manual annotations

Abstract

Wildfires have increased in frequency and severity over the past two decades, especially in the Western United States. Beyond physical infrastructure damage caused by these wildfire events, researchers have increasingly identified harmful impacts of particulate matter generated by wildfire smoke on respiratory, cardiovascular, and cognitive health. This inference is difficult due to the spatial and temporal uncertainty regarding how much particulate matter is specifically attributable to wildfire smoke. One factor contributing to this challenge is the reliance on manually drawn smoke plume annotations, which are often noisy representations limited to the United States. This work uses deep convolutional neural networks to segment smoke plumes from geostationary satellite imagery. We compare the performance of predicted plume segmentations versus the noisy annotations using causal inference methods to estimate the amount of variation each explains in Environmental Protection Agency (EPA) measured surface level particulate matter <2.5um in diameter ($\textrm{PM}_{2.5}$).