Robust Small Methane Plume Segmentation in Satellite Imagery

TL;DR

This paper introduces a deep learning method using U-Net with ResNet34 for detecting small methane plumes in satellite imagery, achieving high sensitivity and precision for plumes as small as 400 m2, aiding climate change mitigation.

Contribution

The paper presents a novel deep learning approach with spectral enhancement techniques that significantly improves detection of small methane plumes in satellite images.

Findings

Achieved 78.39% F1-score on validation set.

Detected plumes as small as 400 m2, surpassing traditional methods.

Outperformed existing remote sensing techniques in sensitivity and precision.

Abstract

This paper tackles the challenging problem of detecting methane plumes, a potent greenhouse gas, using Sentinel-2 imagery. This contributes to the mitigation of rapid climate change. We propose a novel deep learning solution based on U-Net with a ResNet34 encoder, integrating dual spectral enhancement techniques (Varon ratio and Sanchez regression) to optimise input features for heightened sensitivity. A key achievement is the ability to detect small plumes down to 400 m2 (i.e., for a single pixel at 20 m resolution), surpassing traditional methods limited to larger plumes. Experiments show our approach achieves a 78.39% F1-score on the validation set, demonstrating superior performance in sensitivity and precision over existing remote sensing techniques for automated methane monitoring, especially for small plumes.