[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-117942-en":3,"doc-seo-117942-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},117942,1649267921044,"Ava Thompson","https://us-avatar.wpscdn.com/avatar/1800007509477c92dfb?_k=1782875107921204101",8,"Research & Report","Semantic Segmentation of Methane Plumes with Hyperspectral Machine Learning Models","Methane is a key greenhouse gas, and reducing active point-source emissions is a fast-impact mitigation strategy. Remote sensing can detect methane plumes, but current methods suffer from high false positives and depend on manual intervention, while machine learning is constrained by limited annotated datasets. This work releases a machine learning-ready dataset with refined methane plume annotations, labelled AVIRIS-NG hyperspectral data and simulated WorldView-3 multispectral views for benchmarking. The proposed HyperSTARCOP model improves over a matched filter baseline by more than 25% in F1 and cuts false positives per classified tile by over 41.83%, also demonstrating zero-shot generalisation on EMIT data.","[www. nature.com/scientificreports](www. nature.com/scientificreports)  \nOPEN  \nSemantic segmentation of methane plumes  \nwith hyperspectral machine learning models  \nVít Růžička1,2*, Gonzalo Mateo‑Garcia2,3, Luis Gómez‑Chova3, Anna Vaughan4, Luis Guanter5,6 & Andrew Markham1  \nMethane is the second most important greenhouse gas contributor to climate change; atthe sametime its reduction has been denoted as one of the fastest pathways to preventing temperature growth due to its short atmospheric lifetime. In particular, the mitigation of active point‑sources associated with the fossil fuel industry has a strong and cost‑effective mitigation potential. Detection of methane plumes in remote sensing data is possible, but the existing approaches exhibit high false positive rates and need manual intervention. Machine learning research in this area is limited due to the lack of large real‑world annotated datasets. In this work, we are publicly releasing a machine learning ready dataset with manually refined annotation of methane plumes. We present labelled hyperspectral data from the AVIRIS‑NG sensor and provide simulated multispectral WorldView‑3 views of the same data to allow for model benchmarking across hyperspectral and multispectral sensors. We propose sensor agnostic machine learning architectures, using classical methane enhancement products as input features. Our HyperSTARCOP model outperforms strong matched filter baseline by over 25% in F1 score, while reducing its false positive rate per classified tile by over 41.83%. Additionally, we demonstrate zero‑shot generalisation of our trained model on data from the EMIT hyperspectral instrument, despite the differences in the spectral and spatial resolution between the two sensors:  \nin an annotated subset of EMIT images HyperSTARCOP achieves a 40% gain in F1 score over the baseline.  \nMethane leak detection from anthropogenic sources has seen increasing attention, as it is regarded as one of the most viable targets for preventing catastrophic scenarios in temperature increase due to climate change related effects1. Given methane’s short atmospheric lifetime, its removal from the atmosphere would have a very rapid effect in reducing global warming over the next decades. Large leaks, the so-called super-emitters, have been shown to contribute disproportionately to the concentration of methane in the atmosphere: Lavaux et al.2 recently showed that 12% of all oil and gas (O &G) methane emissions are episodic ultra-emission events that in many cases are caused by equipment failures in oil rigs, pipelines or well pads. Additionally, those emissions are highly underestimated: Alvarez et al.3 reported that O &G supply chain emissions in 2015 were 60% higher than bottom up estimates from the United States Environmental Protection Agency, and Zhang et al.4 reported that observed emissions using satellite data are two times higher than bottom-up inventories in the Permian basin. This is due to the fact that bottom-up inventories often underestimate emissions, which can be improved with the use of satellite-based information.  \nUsing different multispectral and hyperspectral satellite instruments several works5–9 have proposed methods for detection and identification of point sources of medium to large methane emissions (>100kg/h) . However these methods still require a significant amount of manual intervention: for hyperspectral instruments, methods based on a matched filter, such as mag1c10, produce reliable enhancements, however, they are still prone to high false detection rates. Meanwhile, methods for multispectral data have not been automated and existing approaches6,8 require manual inspection by human experts looking at pre-computed spectral ratio products. Furthermore, there is no standard dataset for the task of methane plume detection; existing works5–9 report  \n1University of Oxford, Oxford, UK. 2Trillium Technologies, London, UK. 3University of Valencia, Valencia, Spain. 4U","cbCaimMoPoNbsYi0","https://ap.wps.com/l/cbCaimMoPoNbsYi0","pdf",2546111,1,14,"English","en",105,"# Overview\n## Motivation and challenge\n## Released dataset and sensor benchmarking\n## Proposed sensor-agnostic models\n## Results and generalisation to EMIT\n## Dataset annotation and experimental setup","[{\"question\":\"Why is methane plume detection challenging in remote sensing data?\",\"answer\":\"Existing approaches can produce high false positive rates and require manual intervention. The task also lacks large real-world annotated datasets for training robust models.\"},{\"question\":\"What data and annotations does the work release?\",\"answer\":\"It releases manually refined, machine learning-ready labelled hyperspectral AVIRIS-NG data and provides simulated multispectral WorldView-3 views to support model benchmarking across sensors.\"},{\"question\":\"How does HyperSTARCOP perform compared with a matched filter baseline?\",\"answer\":\"HyperSTARCOP outperforms the matched filter baseline by over 25% in F1 score, while reducing false positives per classified tile by over 41.83%. It also shows zero-shot generalisation on EMIT data with a reported 40% F1 gain on an annotated subset.\"}]","Semantic Segmentation of Methane Plumes with Hyperspectral Machine Learning Models | PDF",1785680469,35,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"semantic-segmentation-of-methane-plumes-with-hyperspectral-machine-learning-models","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/semantic-segmentation-of-methane-plumes-with-hyperspectral-machine-learning-models/117942/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-02",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why is methane plume detection challenging in remote sensing data?","Question",{"text":75,"@type":76},"Existing approaches can produce high false positive rates and require manual intervention. The task also lacks large real-world annotated datasets for training robust models.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What data and annotations does the work release?",{"text":80,"@type":76},"It releases manually refined, machine learning-ready labelled hyperspectral AVIRIS-NG data and provides simulated multispectral WorldView-3 views to support model benchmarking across sensors.",{"name":82,"@type":73,"acceptedAnswer":83},"How does HyperSTARCOP perform compared with a matched filter baseline?",{"text":84,"@type":76},"HyperSTARCOP outperforms the matched filter baseline by over 25% in F1 score, while reducing false positives per classified tile by over 41.83%. It also shows zero-shot generalisation on EMIT data with a reported 40% F1 gain on an annotated subset.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]