Wildfire smoke plume segmentation using geostationary satellite imagery
This work addresses the challenge of accurately attributing particulate matter to wildfire smoke for public health research, though it is incremental as it applies existing deep learning methods to a specific domain.
The paper tackles the problem of segmenting wildfire smoke plumes from geostationary satellite imagery using deep convolutional neural networks, and finds that the predicted segmentations explain more variation in EPA-measured surface-level PM2.5 compared to noisy manual annotations.
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 ($\textrm{PM}_{2.5}$).