Research Assistant (Fixed Term)

We are seeking a highly motivated Research Assistant to join an exciting interdisciplinary project at the intersection of ecology and computer science. This collaborative project, involving researchers from the Department of Plant Sciences and Computer Science at the University of Cambridge, focuses on using Geospatial Foundation Modelling to generate a globally consistent high-resolution habitat map.

About the Role

Global efforts to address nature-related risks and achieve ambitious conservation targets are being hampered by a fundamental data gap: there is still no accurate fine-grained, globally consistent map of the world's remaining natural and human-modified habitats, classified by a robust ecological typology. There is also no reliable system for tracking habitat change. These shortfalls affect a wide range of stakeholders conservationists cannot track species ranges accurately, companies struggle to report on nature-related risks, governments lack credible datasets to guide their 30 x 30 conservation commitments under the UN Convention on Biological Diversity, and philanthropic foundations and NGOs are unable to prioritize ecosystem investments with confidence or evaluate the effectiveness of their interventions. Existing attempts to map habitats have fallen short, largely due to a lack of high-quality, regionally relevant training data that can be used to train models to recognize and contextualize diverse land cover types at the ecoregion level. Creating an open-source library of ecoregional training data will enable the next generation of AI-powered mapping tools to deliver unprecedented accuracy and consistency in habitat classification and change detection. By acting now, we can finally provide the data foundation needed to mobilize sustainable finance, guide effective conservation, and secure the world's most important natural habitats for future generations.

Recent advances in geospatial foundation models, including Cambridge University's TESSERA model, offer a transformative opportunity to overcome these challenges provided they are supported by comprehensive, expertly curated ground truth datasets, which the CCI community is well positioned to provide. Geospatial foundation models are powerful AI systems trained on huge amounts of satellite images that can understand patterns and changes on Earth, such as land use and land cover change mapping. Instead of building a new model for every task, foundation models can be quickly adapted to answer different questions about our planet, saving time and effort. This makes it much easier for scientists and decision-makers to get reliable insights from geospatial data, whether they're monitoring habitats, crops, cities, or natural disasters.

Fixed-term: The funds for this post are available for 4 months in the first instance.

To apply online for this vacancy and to view further information about the role, please click 'Apply' above.

Please notice that if you have not received any news from us 1 month after the closing date you should consider that on this occasion your application has not been successful.

Please quote reference PD46354 on your application and in any correspondence about this vacancy.

The University actively supports equality, diversity and inclusion and encourages applications from all sections of society.

The University has a responsibility to ensure that all employees are eligible to live and work in the UK.

Advert information