Methane Plume

Methane plume detection aims to identify and quantify atmospheric methane releases, primarily to mitigate their contribution to climate change. Current research heavily utilizes deep learning, employing architectures like U-Net, Mask R-CNN, and transformers, often trained on large datasets of satellite imagery (e.g., Sentinel-2, Landsat, PRISMA, AVIRIS-NG) and augmented with simulated data to improve model accuracy and robustness. These advancements enable near real-time monitoring of super-emitters, facilitating rapid response and informing policy decisions, significantly improving the efficiency and scale of methane emission tracking compared to previous methods.