Towards Methane Detection Onboard Satellites

2025-08-30

AI Summary

Methane—a potent greenhouse gas—requires real-time, low-cost onboard detection from space to support climate mitigation efforts. This paper proposes an end-to-end onboard methane detection framework tailored for satellite platforms. Departing from conventional pipelines, it trains a lightweight machine learning model directly on unorthorectified hyperspectral imagery, eliminating computationally intensive preprocessing steps such as orthorectification and atmospheric correction. Leveraging data from the Earth Surface Mineral Dust Source Investigation (EMIT) sensor, we construct a standardized, ML-ready dataset. Experimental results demonstrate that our model achieves detection accuracy comparable to the orthorectified + matched-filter (mag1c) baseline on uncorrected data, and significantly outperforms it on orthorectified data. The approach substantially reduces onboard computational load and downlink bandwidth requirements. To foster reproducibility and adoption, we publicly release the trained models, dataset, and source code.

Technology Category

• Machine Learning: Hardware-aware ML
• Computer Vision: Remote Sensing / Geospatial AI
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Web Mining and Content Analysis: Large pretrained models with web data
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Methane is a potent greenhouse gas and a major driver of climate change, making its timely detection critical for effective mitigation. Machine learning (ML) deployed onboard satellites can enable rapid detection while reducing downlink costs, supporting faster response systems. Conventional methane detection methods often rely on image processing techniques, such as orthorectification to correct geometric distortions and matched filters to enhance plume signals. We introduce a novel approach that bypasses these preprocessing steps by using extit{unorthorectified} data (UnorthoDOS). We find that ML models trained on this dataset achieve performance comparable to those trained on orthorectified data. Moreover, we also train models on an orthorectified dataset, showing that they can outperform the matched filter baseline (mag1c). We release model checkpoints and two ML-ready datasets comprising orthorectified and unorthorectified hyperspectral images from the Earth Surface Mineral Dust Source Investigation (EMIT) sensor at https://huggingface.co/datasets/SpaceML/UnorthoDOS , along with code at https://github.com/spaceml-org/plume-hunter.

Problem

• Detecting methane emissions using satellite data
• Bypassing preprocessing steps with unorthorectified data
• Comparing machine learning models to traditional methods

Innovation

• Uses unorthorectified satellite data for methane detection
• Machine learning models bypass traditional preprocessing steps
• Achieves comparable performance to orthorectified data methods