Plume Segmentation from MethaneSAT with Cross-Sensor Transfer Learning and Physics-Informed Postprocessing

2026-05-22

AI Summary

This study addresses the challenges of scarce annotated data and unreliable inference under complex environmental conditions in detecting methane plumes from MethaneSAT satellite imagery. To this end, the authors propose a dual-mode detection framework that integrates cross-sensor transfer learning with physics-informed post-processing. Built upon a Mask R-CNN architecture with a ResNet-50 backbone, the model is pretrained using MethaneAIR and synthetic data, then refined through physically constrained post-processing steps—including morphological filtering, spatial proximity merging, and distribution-based classification—to enable robust emission source identification under both high-sensitivity and high-precision operating modes. Experimental results demonstrate that the proposed method achieves an instance-level recall of 0.98 on MethaneSAT data, with precision/recall of 0.71/0.94 in high-sensitivity mode and 0.92/0.70 in high-precision mode, substantially enhancing both detection performance and practical applicability.

Abstract

Automated detection and masking of individual methane plumes from satellite imagery is important for operational emission attribution and quantification. We present a machine learning framework for plume detection from MethaneSAT retrieved column-averaged dry-air mole fractions of methane. We address two core challenges: the scarcity of labeled MethaneSAT data and the need for inference reliability across diverse atmospheric and surface conditions. We first demonstrate that Mask R-CNN with a ResNet-50 backbone outperforms U-Net semantic segmentation on both MethaneAIR (an airborne version of MethaneSAT) and MethaneSAT data, with pixel-level F1 score gains of 10.49 and 5.48 respectively. To address MethaneSAT data scarcity, we evaluate three cross-sensor transfer strategies leveraging MethaneAIR flights and synthetic plumes. Mask R-CNN with ResNet-50 fine-tuned from MethaneAIR pre-trained weights is the most effective strategy, achieving instance-level precision of 0.60 and a near-perfect recall of 0.98 at the baseline operating point. A physics-informed post-processing pipeline converts detections into two operationally distinct modes. The first is a high-sensitivity mode that applies morphological filtering and proximity-based merging for comprehensive emission screening, achieving precision of 0.71 and recall of 0.94. The second is a high-precision mode that additionally applies a distribution-based classifier for confident source attribution, achieving precision of 0.92 and recall of 0.70. Manual review of detections classified as false positives against our wavelet-based ground truth labels reveals that a meaningful fraction of cases correspond to real methane enhancements excluded by conservative labeling criteria, indicating that precision values reported are lower bounds on true detection performance... Our data and code are available at: https://doi.org/10.7910/DVN/FR959H

Problem

• plume segmentation
• methane detection
• satellite imagery
• data scarcity
• emission attribution

Innovation

• cross-sensor transfer learning
• physics-informed postprocessing
• Mask R-CNN
• methane plume segmentation
• MethaneSAT