Methane Detection On Board Satellites from Unorthorectified Imagery

Luca Marini; Maggie Chen; Hala Lamdouar; Laura Martínez-Ferrer; Dr C. P. Bridges; Giacomo Acciarini

cs.CVSep 4, 2026

Methane Detection On Board Satellites from Unorthorectified Imagery

Authors: Luca Marini, Maggie Chen, Hala Lamdouar, Laura Martínez-Ferrer, Dr C. P. Bridges, Giacomo Acciarini

Organizations: Delft University of Technology, Delft, The Netherlands · University of Oxford, Oxford, United Kingdom · Universitat de València, València, Spain · University of Surrey, Guildford, United Kingdom · European Space Agency, Advanced Concepts Team, Noordwijk, The Netherlands

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.

Figures & tables

Dataset	Instrument	Bands	Spectral range (nm)	Unorthorectified
STARCOP [ 23 ]	AVIRIS-NG	125	RGB, 1573–1699, 2004–2480	✗
OxHyperSyntheticCH4 [ 22 ]	EMIT	86	RGB, 1573–1699, 2004–2478	✗
OxHyperRealCH4 [ 22 ]	EMIT	86	RGB, 1573–1699, 2004–2478	✗
UnorthoDOS (Ours)	EMIT	86	RGB, 1573–1699, 2004–2478	✓

Table 1: Comparison of hyperspectral datasets for methane plume detection, with emphasis on the presence or absence of orthorectification preprocessing.

Figure 1: Overview of the UnorthoDOS pipeline for synthesizing unorthorectified hyperspectral data from orthorectified EMIT data.

Figure 2: Overview of the detection pipeline. EMIT images are preprocessed into orthorectified and unorthorectified datasets for ML model training. Trained ML models simulate a tip and cue setting to perform plume classification and segmentation.

Setting	Model	Threshold (ppm m)	Precision	Recall	F1-Score	Accuracy (%)
Orthorectified	mag1c [ 5 ]	N/A	52.55	94.12	67.45	54.58
Orthorectified	mag1c [ 5 ]	$\geq$ 900	39.60	95.24	55.94	46.84
Orthorectified	U-Net	N/A	95.56	56.21	70.78	76.80
Orthorectified	U-Net	$\geq$ 900	94.60	83.33	88.61	92.41
Unorthorectified	U-Net	N/A	89.19	48.89	63.16	71.48
Unorthorectified	U-Net	$\geq$ 900	87.88	71.61	78.91	85.65

Table 2: Image classification performance. Precision, recall, F1-score, and accuracy are reported for mag1c [ 5 ] and U-Net, trained/evaluated on orthorectified and unorthorectified data. Best result per column within each block is highlighted in bold .

Setting	Model	Threshold (ppm m)	Precision	Recall	F1-Score	IoU
Orthorectified	mag1c [ 5 ]	N/A	41.67	15.69	22.80	4.76
Orthorectified	mag1c [ 5 ]	$\geq$ 900	44.91	15.69	23.25	4.84
Orthorectified	U-Net	N/A	79.23	19.91	31.82	18.47
Orthorectified	U-Net	$\geq$ 900	82.77	32.52	46.69	30.80
Unorthorectified	U-Net	N/A	88.41	19.85	32.42	16.91
Unorthorectified	U-Net	$\geq$ 900	89.08	33.01	48.16	26.17

Table 3: Semantic segmentation performance. Precision, recall, F1-score, and IoU are reported for mag1c [ 5 ] and U-Net, trained/evaluated on orthorectified and unorthorectified data. Best result per column within each block is highlighted in bold .

Panel of Figure 3 — Figure 3: Visualisation of semantic segmentation results on 2 example tiles from the orthorectified ( fig. 3(a) ) and the unorthorectified ( fig. 3(b) ) datasets each. From Left to Right in each sub-figure: L1B tiles (only RGB bands shown for visualisation) overlaid with ground truth methane plume annotations; predicted semantic segmentation plume masks from U-Net; segmentation predictions from mag1c [ 5 ] .

		PyTorch	32-bit ONNX	16-bit ONNX	8-bit Quant
File size (MB)		3.30	3.25	1.70	0.84
Max relative difference (%)	Orthorectified	0	1.66 $\times 10^{-4}$	0.30	11.93
Max relative difference (%)	Unorthorectified	0	2.19 $\times 10^{-4}$	0.25	10.70

Table 4: File sizes in MB, and the maximum relative difference in the outputs of the original PyTorch models trained on the orthorectifed and unorthorectified datasets and ONNX models compressed to varying precisions.

Explore similar work

A Fast Methane Detection Pipeline on Board Satellites Based on Mag1c-SAS and LinkNet
Jun 2, 2026Jonáš Herec, Vít Růžička, Rado Pitoňák +1Remote Sensing Image Segmentation Remote Sensing
Zaitra s.r.o., Brno, Czech Republic · Faculty of Informatics, Masaryk University, Brno, Czech Republic · NASA JPL, California, USA
FLAME: Physics-Guided Neural Operators for Onboard Satellite Methane Detection in Hyperspectral Imagery
Jun 1, 2026Junhyuk Heo, Junhwan Park, Sancheol Sim +2Edge Inference Remote Sensing
TelePIX, Seoul, Republic of Korea
Plume Segmentation from MethaneSAT with Cross-Sensor Transfer Learning and Physics-Informed Postprocessing
May 22, 2026Manuel Pérez-Carrasco, Maya Nasr, Zhan Zhang +16Remote Sensing Image Segmentation Instance Segmentation
Center for Astrophysics | Harvard & Smithsonian, Cambridge, MA · Environmental Defense Fund, Washington, D.C., USA · Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA, USA +7

Jun 2, 2026J/K move · Enter open · S save

A Fast Methane Detection Pipeline on Board Satellites Based on Mag1c-SAS and LinkNet

Jonáš Herec, Vít Růžička, Rado Pitoňák, Jan Sedmidubsky

Zaitra s.r.o., Brno, Czech Republic · Faculty of Informatics, Masaryk University, Brno, Czech Republic · NASA JPL, California, USA

Remote Sensing Image Segmentation Remote Sensing Hyperspectral Imaging

Open paper AskPDF arXiv

Methane is a potent greenhouse gas, and detecting leaks early via hyperspectral satellite imagery can help climate change mitigation efforts. Meanwhile, many existing hyperspectral missions only capture areas manually targeted by operators, thus missing potential events of interest. To overcome slow downlink rates cost-effectively, onboard detection is a viable solution. However, traditional methane detection methods are too computationally demanding for resource-limited onboard hardware. This work accelerates methane detection by focusing on efficient, low-power algorithms. In particular, we test fast target detection ACE and CEM methods that have not been previously used for methane detection and propose Mag1c-SAS -- a significantly faster variant of the current state-of-the-art Mag1c algorithm. To explore their detection potential, we integrate them with a machine learning model based on U-Net and LinkNet. We evaluate our methods on the STARCOP dataset and a novel EMIT-MSeg dataset, which we introduce and open-source alongside a high-quality annotation strategy. The proposed Mag1c-SAS approach proves highly effective by operating ~80x faster than the original Mag1c approach, providing a visually similar, but noisier result. When additionally paired with the lightweight LinkNet approach, it effectively reduces noise, achieving AUPRC score improvements of over 30 pp on EMIT-MSeg compared to the baseline Mag1c approach, and an F1 score on STARCOP ~4 pp higher. We evaluate two novel band selection strategies and confirm the system's onboard viability through hardware profiling, demonstrating marginal power consumption and efficient CPU/RAM utilization. We release the final system in a user-friendly and lightweight PyPI library at: https://pypi.org/project/onboard-methane-detection/, alongside all experimental code, models, and data at: https://github.com/zaitra/methane-filters-benchmark.

Abstract

Explore similar work

A Fast Methane Detection Pipeline on Board Satellites Based on Mag1c-SAS and LinkNet

FLAME: Physics-Guided Neural Operators for Onboard Satellite Methane Detection in Hyperspectral Imagery

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