BAS AI Lab
Lead Research Organisation:
British Antarctic Survey
Department Name: UNLISTED
Abstract
Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.
Organisations
- British Antarctic Survey (Collaboration, Lead Research Organisation)
- Alan Turing Institute (Collaboration)
- University of Washington (Collaboration)
- UK CENTRE FOR ECOLOGY & HYDROLOGY (Collaboration)
- National Oceanography Centre (Collaboration)
- National Institute of Water and Atmospheric Research (NIWA, New Zealand) (Collaboration)
- UNIVERSITY COLLEGE LONDON (Collaboration)
- University of Glasgow (Collaboration)
- University of Cambridge (Collaboration)
- World Wide Fund for Nature (Collaboration)
- European Centre for Medium Range Weather Forecasting ECMWF (Collaboration)
- Amazon.com (Collaboration)
Publications
Andersson T
(2023)
Environmental sensor placement with convolutional Gaussian neural processes
in Environmental Data Science
Andersson T
(2022)
Seasonal Arctic sea ice forecasting with probabilistic deep learning
Attard M
(2024)
Review of Satellite Remote Sensing and Unoccupied Aircraft Systems for Counting Wildlife on Land
in Remote Sensing
Bajada J
(2022)
Efficient Temporal Piecewise-Linear Numeric Planning With Lazy Consistency Checking
in IEEE Transactions on Artificial Intelligence
Bojko J
(2022)
Microsporidia: a new taxonomic, evolutionary, and ecological synthesis.
in Trends in parasitology
Bradley A
(2024)
WAVI.jl: Ice Sheet Modelling in Julia
in Journal of Open Source Software
Bruinsma WP
(2023)
Autoregressive Conditional Neural Processes
Christensen H
(2025)
Machine learning for stochastic parametrization
in Environmental Data Science
Coca-Castro A
(2025)
Improving the reproducibility in geoscientific papers: lessons learned from a Hackathon in climate science
in Environmental Data Science
| Description | We developed an AI to combine different types of satellite sensor data of sea ice which led to improved spatial resolution compared to the currently available sea ice satellite derived products. We also have use our sea ice forecasts to be able to forecast the onset of a migration of caribou across the Cornonation Gulf (Canada). Lastly, we are detecting and now tracking icebergs from fragmented satellite images which has implications for freshening of the ocean layers and therefore ecosystems, and also the risk to ships. |
| Exploitation Route | We are developing new tools and digital infrastructure that is integrated with one another, and as part of international software development efforts. Our aim is for these tools to be integrated as components within environmental digital twins. |
| Sectors | Aerospace Defence and Marine Digital/Communication/Information Technologies (including Software) Environment |
| URL | http://bas.ac.uk/ai |
| Description | Our findings have been used to support two key projects: * to identify the most fuel efficient route through sea ice in the polar regions to meet the UKRI net zero agenda * to support local government in Canada and WWF with wildlife surveys, to support conservation of Caribou and Polar Bears. |
| First Year Of Impact | 2024 |
| Sector | Aerospace, Defence and Marine,Digital/Communication/Information Technologies (including Software),Environment,Transport |
| Impact Types | Societal Economic |
| Title | Annotated dataset of wandering albatrosses detected in Very High Resolution (VHR) satellite imagery of South Georgia, Crozet Islands and Kerguelen Islands, from 2015-2017. |
| Description | This annotated dataset comprises locational data of wandering albatrosses at colonies on Bird Island and Annenkov Island in South Georgia, Apotres in the Crozet Islands and Grande Coulee on the west coast of Kerguelen Island. Albatrosses were detected from Very High Resolution (VHR) satellite imagery, which was acquired on the following dates: - Bird Island: January 10th, 2015 - Annenkov: February 3rd, 2017 - Apotres: March 3rd, 2017 - Grande Coulee: March 16th, 2017 Images were manually scanned and albatrosses detected through the use of grids. An experienced observer ('reference count') and five novice volunteers (observers 1-5) annotated the satellite images to assess interobserver variation. Scanning was carried out by collaborators and volunteers of the Albatrosses from Space project. Additional metadata includes information on image type and model, and image acquisition location. This dataset supports the advancement of remote monitoring for a vulnerable seabird species. This research was funded by the Natural Environment Reseach Council through the NEXUSS CDT, grant number NE/N012070/1, titled 'Automated UAV and Satellite Image Analysis for Wildlife Monitoring.' |
| Type Of Material | Database/Collection of data |
| Year Produced | 2024 |
| Provided To Others? | Yes |
| Impact | N/A |
| URL | https://data.bas.ac.uk/full-record.php?id=GB/NERC/BAS/PDC/01830 |
| Title | Downscaled ERA5 monthly precipitation data using Multi-Fidelity Gaussian Processes between 1980 and 2012 for the Upper Beas and Sutlej Basins, Himalayas |
| Description | The dataset is the output of a statistical model which downscales ERA5 monthly precipitation data using gauge measurements from the Upper Beas and Sutlej Basins in the Western Himalayas. Multi-Fidelity Gaussian Processes (MFGPs) are used to generate more accurate precipitation values between 1980 and 2012, including over ungauged areas of the basins. MFGPs are a probabilistic machine learning method that provides principled uncertainty estimates via the prediction of probability distributions. These predictions can therefore be used to estimate the likelihood of extreme precipitation events which have led to droughts, floods, and landslides. Funding from UK Engineering and Physical Sciences Research Council [grant number: 2270379]. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2023 |
| Provided To Others? | Yes |
| Impact | Not applicable |
| URL | https://data.bas.ac.uk/full-record.php?id=GB/NERC/BAS/PDC/01769 |
| Description | AMBER (AI-assisted Monitoring of Biodiversity with Edge-processing and Remote sensors) |
| Organisation | UK Centre for Ecology & Hydrology |
| Country | United Kingdom |
| Sector | Public |
| PI Contribution | This is an Abrdn-funded CEH/Turing biodiversity monitoring project. Our team plans to contribute new models and data to the scivision catalog |
| Collaborator Contribution | Providing datasets and domain expertise |
| Impact | The project is only just starting |
| Start Year | 2022 |
| Description | BAS AI Lab - implement models and datasets |
| Organisation | British Antarctic Survey |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | Assist in the preparation and publication of a demonstrator/notebook in the scivision gallery demonstrating the Vegetation Edge Dector (VEdge_Detector) model authored by Rogers et al. (2021). |
| Collaborator Contribution | Register sample data and pre-trained VEdge_Detector model in the scivision public catalog. |
| Impact | Scivision notebook repository (https://github.com/scivision-gallery/coastalveg-edge-detection) Pull request (https://github.com/alan-turing-institute/scivision/pull/155) Presentation at AI UK conference - March 2022 |
| Start Year | 2022 |
| Description | Cambridge AI for Environmental Risk CDT Collaboration |
| Organisation | University of Cambridge |
| Department | Department of Earth Sciences |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | Assist in the preparation and publication a demonstrator/notebook in the scivision gallery demonstrating the detectreeRGB model authored by Sebastian Hickman (Cambridge University) as part of his AI4ER MRes project, titled 'Detecting changes in tall tree height with machine learning, LiDAR, and RGB imagery'. |
| Collaborator Contribution | Register sample data and pre-trained detectreeRGB model in the scivision public catalog. |
| Impact | Scivision notebook repository (https://github.com/scivision-gallery/tree-crown-detection) Pull Request (https://github.com/alan-turing-institute/scivision/pull/267) |
| Start Year | 2022 |
| Description | DeepSensor - National Institute of Water and Atmospheric Research NIWA |
| Organisation | National Institute of Water and Atmospheric Research (NIWA, New Zealand) |
| Country | New Zealand |
| Sector | Public |
| PI Contribution | creating better gridded datasets from sensor networks and additionally informing sensor placements, with respect to a variety of scenarios based on rainfall, irrigation and soil moisture |
| Collaborator Contribution | Providing data and local knowledge for a new region and environmental conditions |
| Impact | None yet |
| Start Year | 2022 |
| Description | Engagement with PhD student at Cambridge University |
| Organisation | University of Cambridge |
| Department | Department of Engineering |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | Supporting a PhD student by shaping the work to develop synthetic data of sea ice imagery to help train machine learning / artificial intellegence models |
| Collaborator Contribution | develop the code and synthetic data |
| Impact | Accepted presentation (talk) at the European Geophysical Union general assembly in April 2025 |
| Start Year | 2024 |
| Description | IceNet (v1) Research Team |
| Organisation | Alan Turing Institute |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | The primary researcher on this grant (Tom Andersson) has put together a strong group of scientists to help collaborate on this multidisciplinary project. We have had meetings around once a month with many of those listed below, and together we have had conference poster accepted at Europe's largest environmental conference (European Geophysical Union, EGU). Tom has organised meetings at the Turing institute to get many of these people together. Tom Andersson, Scott Hosking, María Pérez-Ortiz, Brooks Paige, Chris Russell, Andrew Elliott, Stephen Law, Jeremy Wilkinson, Yevgeny Askenov, Will Tebbutt, and Emily Shuckburgh |
| Collaborator Contribution | The partners have made a huge impact on the methodological machine learning design for this project and the paper which is now in review. |
| Impact | Abstract accepted for the EGU General Assembly 2020: EGU2020-15481 - "Deep learning for monthly Arctic sea ice concentration prediction" by Tom Andersson et al. (Session ITS4.3/AS5.2 - Machine learning for Earth System modelling) Pre-print to our paper: https://eartharxiv.org/repository/view/2027/ |
| Start Year | 2019 |
| Description | IceNet (v1) Research Team |
| Organisation | Amazon.com |
| Department | Amazon Web Services |
| Country | United States |
| Sector | Private |
| PI Contribution | The primary researcher on this grant (Tom Andersson) has put together a strong group of scientists to help collaborate on this multidisciplinary project. We have had meetings around once a month with many of those listed below, and together we have had conference poster accepted at Europe's largest environmental conference (European Geophysical Union, EGU). Tom has organised meetings at the Turing institute to get many of these people together. Tom Andersson, Scott Hosking, María Pérez-Ortiz, Brooks Paige, Chris Russell, Andrew Elliott, Stephen Law, Jeremy Wilkinson, Yevgeny Askenov, Will Tebbutt, and Emily Shuckburgh |
| Collaborator Contribution | The partners have made a huge impact on the methodological machine learning design for this project and the paper which is now in review. |
| Impact | Abstract accepted for the EGU General Assembly 2020: EGU2020-15481 - "Deep learning for monthly Arctic sea ice concentration prediction" by Tom Andersson et al. (Session ITS4.3/AS5.2 - Machine learning for Earth System modelling) Pre-print to our paper: https://eartharxiv.org/repository/view/2027/ |
| Start Year | 2019 |
| Description | IceNet (v1) Research Team |
| Organisation | British Antarctic Survey |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | The primary researcher on this grant (Tom Andersson) has put together a strong group of scientists to help collaborate on this multidisciplinary project. We have had meetings around once a month with many of those listed below, and together we have had conference poster accepted at Europe's largest environmental conference (European Geophysical Union, EGU). Tom has organised meetings at the Turing institute to get many of these people together. Tom Andersson, Scott Hosking, María Pérez-Ortiz, Brooks Paige, Chris Russell, Andrew Elliott, Stephen Law, Jeremy Wilkinson, Yevgeny Askenov, Will Tebbutt, and Emily Shuckburgh |
| Collaborator Contribution | The partners have made a huge impact on the methodological machine learning design for this project and the paper which is now in review. |
| Impact | Abstract accepted for the EGU General Assembly 2020: EGU2020-15481 - "Deep learning for monthly Arctic sea ice concentration prediction" by Tom Andersson et al. (Session ITS4.3/AS5.2 - Machine learning for Earth System modelling) Pre-print to our paper: https://eartharxiv.org/repository/view/2027/ |
| Start Year | 2019 |
| Description | IceNet (v1) Research Team |
| Organisation | European Centre for Medium Range Weather Forecasting ECMWF |
| Country | United Kingdom |
| Sector | Public |
| PI Contribution | The primary researcher on this grant (Tom Andersson) has put together a strong group of scientists to help collaborate on this multidisciplinary project. We have had meetings around once a month with many of those listed below, and together we have had conference poster accepted at Europe's largest environmental conference (European Geophysical Union, EGU). Tom has organised meetings at the Turing institute to get many of these people together. Tom Andersson, Scott Hosking, María Pérez-Ortiz, Brooks Paige, Chris Russell, Andrew Elliott, Stephen Law, Jeremy Wilkinson, Yevgeny Askenov, Will Tebbutt, and Emily Shuckburgh |
| Collaborator Contribution | The partners have made a huge impact on the methodological machine learning design for this project and the paper which is now in review. |
| Impact | Abstract accepted for the EGU General Assembly 2020: EGU2020-15481 - "Deep learning for monthly Arctic sea ice concentration prediction" by Tom Andersson et al. (Session ITS4.3/AS5.2 - Machine learning for Earth System modelling) Pre-print to our paper: https://eartharxiv.org/repository/view/2027/ |
| Start Year | 2019 |
| Description | IceNet (v1) Research Team |
| Organisation | National Oceanography Centre |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | The primary researcher on this grant (Tom Andersson) has put together a strong group of scientists to help collaborate on this multidisciplinary project. We have had meetings around once a month with many of those listed below, and together we have had conference poster accepted at Europe's largest environmental conference (European Geophysical Union, EGU). Tom has organised meetings at the Turing institute to get many of these people together. Tom Andersson, Scott Hosking, María Pérez-Ortiz, Brooks Paige, Chris Russell, Andrew Elliott, Stephen Law, Jeremy Wilkinson, Yevgeny Askenov, Will Tebbutt, and Emily Shuckburgh |
| Collaborator Contribution | The partners have made a huge impact on the methodological machine learning design for this project and the paper which is now in review. |
| Impact | Abstract accepted for the EGU General Assembly 2020: EGU2020-15481 - "Deep learning for monthly Arctic sea ice concentration prediction" by Tom Andersson et al. (Session ITS4.3/AS5.2 - Machine learning for Earth System modelling) Pre-print to our paper: https://eartharxiv.org/repository/view/2027/ |
| Start Year | 2019 |
| Description | IceNet (v1) Research Team |
| Organisation | University College London |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | The primary researcher on this grant (Tom Andersson) has put together a strong group of scientists to help collaborate on this multidisciplinary project. We have had meetings around once a month with many of those listed below, and together we have had conference poster accepted at Europe's largest environmental conference (European Geophysical Union, EGU). Tom has organised meetings at the Turing institute to get many of these people together. Tom Andersson, Scott Hosking, María Pérez-Ortiz, Brooks Paige, Chris Russell, Andrew Elliott, Stephen Law, Jeremy Wilkinson, Yevgeny Askenov, Will Tebbutt, and Emily Shuckburgh |
| Collaborator Contribution | The partners have made a huge impact on the methodological machine learning design for this project and the paper which is now in review. |
| Impact | Abstract accepted for the EGU General Assembly 2020: EGU2020-15481 - "Deep learning for monthly Arctic sea ice concentration prediction" by Tom Andersson et al. (Session ITS4.3/AS5.2 - Machine learning for Earth System modelling) Pre-print to our paper: https://eartharxiv.org/repository/view/2027/ |
| Start Year | 2019 |
| Description | IceNet (v1) Research Team |
| Organisation | University of Cambridge |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | The primary researcher on this grant (Tom Andersson) has put together a strong group of scientists to help collaborate on this multidisciplinary project. We have had meetings around once a month with many of those listed below, and together we have had conference poster accepted at Europe's largest environmental conference (European Geophysical Union, EGU). Tom has organised meetings at the Turing institute to get many of these people together. Tom Andersson, Scott Hosking, María Pérez-Ortiz, Brooks Paige, Chris Russell, Andrew Elliott, Stephen Law, Jeremy Wilkinson, Yevgeny Askenov, Will Tebbutt, and Emily Shuckburgh |
| Collaborator Contribution | The partners have made a huge impact on the methodological machine learning design for this project and the paper which is now in review. |
| Impact | Abstract accepted for the EGU General Assembly 2020: EGU2020-15481 - "Deep learning for monthly Arctic sea ice concentration prediction" by Tom Andersson et al. (Session ITS4.3/AS5.2 - Machine learning for Earth System modelling) Pre-print to our paper: https://eartharxiv.org/repository/view/2027/ |
| Start Year | 2019 |
| Description | IceNet (v1) Research Team |
| Organisation | University of Glasgow |
| Department | School of Mathematics and Statistics |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | The primary researcher on this grant (Tom Andersson) has put together a strong group of scientists to help collaborate on this multidisciplinary project. We have had meetings around once a month with many of those listed below, and together we have had conference poster accepted at Europe's largest environmental conference (European Geophysical Union, EGU). Tom has organised meetings at the Turing institute to get many of these people together. Tom Andersson, Scott Hosking, María Pérez-Ortiz, Brooks Paige, Chris Russell, Andrew Elliott, Stephen Law, Jeremy Wilkinson, Yevgeny Askenov, Will Tebbutt, and Emily Shuckburgh |
| Collaborator Contribution | The partners have made a huge impact on the methodological machine learning design for this project and the paper which is now in review. |
| Impact | Abstract accepted for the EGU General Assembly 2020: EGU2020-15481 - "Deep learning for monthly Arctic sea ice concentration prediction" by Tom Andersson et al. (Session ITS4.3/AS5.2 - Machine learning for Earth System modelling) Pre-print to our paper: https://eartharxiv.org/repository/view/2027/ |
| Start Year | 2019 |
| Description | IceNet (v1) Research Team |
| Organisation | University of Washington |
| Country | United States |
| Sector | Academic/University |
| PI Contribution | The primary researcher on this grant (Tom Andersson) has put together a strong group of scientists to help collaborate on this multidisciplinary project. We have had meetings around once a month with many of those listed below, and together we have had conference poster accepted at Europe's largest environmental conference (European Geophysical Union, EGU). Tom has organised meetings at the Turing institute to get many of these people together. Tom Andersson, Scott Hosking, María Pérez-Ortiz, Brooks Paige, Chris Russell, Andrew Elliott, Stephen Law, Jeremy Wilkinson, Yevgeny Askenov, Will Tebbutt, and Emily Shuckburgh |
| Collaborator Contribution | The partners have made a huge impact on the methodological machine learning design for this project and the paper which is now in review. |
| Impact | Abstract accepted for the EGU General Assembly 2020: EGU2020-15481 - "Deep learning for monthly Arctic sea ice concentration prediction" by Tom Andersson et al. (Session ITS4.3/AS5.2 - Machine learning for Earth System modelling) Pre-print to our paper: https://eartharxiv.org/repository/view/2027/ |
| Start Year | 2019 |
| Description | IceNet (v1) Research Team |
| Organisation | World Wide Fund for Nature |
| Country | Switzerland |
| Sector | Charity/Non Profit |
| PI Contribution | The primary researcher on this grant (Tom Andersson) has put together a strong group of scientists to help collaborate on this multidisciplinary project. We have had meetings around once a month with many of those listed below, and together we have had conference poster accepted at Europe's largest environmental conference (European Geophysical Union, EGU). Tom has organised meetings at the Turing institute to get many of these people together. Tom Andersson, Scott Hosking, María Pérez-Ortiz, Brooks Paige, Chris Russell, Andrew Elliott, Stephen Law, Jeremy Wilkinson, Yevgeny Askenov, Will Tebbutt, and Emily Shuckburgh |
| Collaborator Contribution | The partners have made a huge impact on the methodological machine learning design for this project and the paper which is now in review. |
| Impact | Abstract accepted for the EGU General Assembly 2020: EGU2020-15481 - "Deep learning for monthly Arctic sea ice concentration prediction" by Tom Andersson et al. (Session ITS4.3/AS5.2 - Machine learning for Earth System modelling) Pre-print to our paper: https://eartharxiv.org/repository/view/2027/ |
| Start Year | 2019 |
| Description | Soil Moisture Prediction |
| Organisation | UK Centre for Ecology & Hydrology |
| Country | United Kingdom |
| Sector | Public |
| PI Contribution | We have setup workshops and discussions, and invited our new collaborators to exising meetings |
| Collaborator Contribution | Providing expertise and knowledge of the data |
| Impact | None yet |
| Start Year | 2021 |
| Description | Soil Moisture collaboration - UKCEH, Glasgow University |
| Organisation | UK Centre for Ecology & Hydrology |
| Country | United Kingdom |
| Sector | Public |
| PI Contribution | We have developed a close working relationship to develop new AI methods using data from UKCEH |
| Collaborator Contribution | Providing data and knowledge of the environmental challenge |
| Impact | The collaboration is multi-disciplinary, including machine learning engineers and hydrologists |
| Start Year | 2022 |
| Description | The Computational and Biological Learning (CBL) lab, Univ Cambridge |
| Organisation | University of Cambridge |
| Department | Department of Engineering |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | Providing expertise on the science challenge, data management, and code development. |
| Collaborator Contribution | Providing expertise on the design of the machine learning algorithms |
| Impact | We are submitting two papers to machine learning conferences in the next few weeks. NeurIPS 2022 workshop paper and ICLR 2023 conference paper |
| Start Year | 2021 |
| Title | C-H-Simpson/HarvestOccupationalHeat: |
| Description | The plots included are now more like the plots in the preprint. (The preprint was also recently updated.) The explanations are more detailed, and the plots are more consistent. |
| Type Of Technology | Software |
| Year Produced | 2021 |
| Open Source License? | Yes |
| Impact | This work led to new collaborations between BAS, Turing and the UK Met Office |
| URL | https://zenodo.org/record/4746385 |
| Title | Code associated with the paper: 'Seasonal Arctic sea ice forecasting with probabilistic deep learning' |
| Description | This is the first release of the codebase associated with the IceNet paper in Nature Communications. |
| Type Of Technology | Software |
| Year Produced | 2021 |
| Open Source License? | Yes |
| Impact | This tool is now being used by WWF for improving their ability for conservation efforts. It is also being used to reduce fuel usage in our research ship |
| URL | https://zenodo.org/record/5176573 |
| Title | Digital Infrastructure for Deploying IceNet Sea Ice Forecasts |
| Description | IceNet is a probabilistic, deep learning sea ice forecasting system developed by an international team and led by British Antarctic Survey and The Alan Turing Institute. |
| Type Of Technology | New/Improved Technique/Technology |
| Year Produced | 2022 |
| Impact | This digital framework is now being used to support wildlife conservation in the Canadian Arctic |