aiqbench analysis
This research introduces UnorthoDOS, a dataset enabling machine learning models to detect methane plumes directly from unorthorectified hyperspectral imagery, bypassing traditional ground-based preprocessing steps. By training U-Net architectures on raw data, the approach achieves performance comparable to orthorectified baselines while significantly reducing model size for feasible onboard satellite deployment. The work advances remote sensing capabilities through efficient machine learning techniques tailored for edge computing in climate monitoring applications.
Methane Detection On Board Satellites from Unorthorectified Imagery
arXiv:2609.04906v1 Announce Type: cross Abstract: As a potent greenhouse gas, methane is a major driver of climate change. Its effective mitigation relies on timely detection. Conventional detection methods rely on orthorectification to correct geometric distortions and matched filters to enhance plume signals, which are steps designed for ground processing and poorly suited to onboard execution. We introduce UnorthoDOS, a dataset and approach for training machine learning models directly on unorthorectified hyperspectral imagery, bypassing both orthorectification and matched-filter products. Our U-Net models trained on unorthorectified data approach the performance of models trained on orthorectified data (IoU 16.91% vs. 18.47% on all plumes), while both substantially outperform the mag1c matched-filter baseline (IoU 4.76%). We further demonstrate the feasibility of onboard deployment: FP16 compression halves model size with under 0.3% output deviation. The trained ML models and two ML-ready datasets -- orthorectified and unorthorectified hyperspectral imagery from the EMIT sensor -- are publicly available at https://huggingface.co/datasets/SpaceML/UnorthoDOS, with code at https://github.com/spaceml-org/plume-hunter.
Abstract:As a potent greenhouse gas, methane is a major driver of climate change. Its effective mitigation relies on timely detection. Conventional detection methods rely on orthorectification to correct geometric distortions and matched filters to enhance plume signals, which are steps designed for ground processing and poorly suited to onboard execution. We introduce UnorthoDOS, a dataset and approach for training machine learning models directly on unorthorectified hyperspectral imagery, bypassing both orthorectification and matched-filter products. Our U-Net models trained on unorthorectified data approach the performance of models trained on orthorectified data (IoU 16.91% vs. 18.47% on all plumes), while both substantially outperform the mag1c matched-filter baseline (IoU 4.76%). We further demonstrate the feasibility of onboard deployment: FP16 compression halves model size with under 0.3% output deviation. The trained ML models and two ML-ready datasets -- orthorectified and unorthorectified hyperspectral imagery from the EMIT sensor -- are publicly available at this https URL, with code at this https URL.
Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
arXiv:2609.04906 [cs.CV]
(or arXiv:2609.04906v1 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2609.04906 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 259 words
Related coverage
Forbes Asia 100 To Watch 2026
forbes.com · Aug 24, 2026
MethaneFuse: Learning from Multi-Sensor Satellite Observations for Methane Plume Detection
arXiv · Sep 10, 2026
On-Board Anomaly Detection for Efficient Marine Environmental Monitoring
arXiv · Oct 6, 2026
Mapping global methane emissions from space with deep learning
Google Research · Sep 1, 2026