Modelling biodiversity responses to climate and human-induced land change

Year
2025
Primary Supervisor
Mattia Mancini <m.c.mancini@exeter.ac.uk>
Institution
University of Exeter (Land, Environment, Economics and Policy )
CASE Partner
Academic Supervisors
Amy Binner <a.r.binner@exeter.ac.uk> (University of Exeter, Land, Environment, Economics and Policy ), Daniel Williamson <d.williamson@exeter.ac.uk> (University of Exeter, Land, Environment, Economics and Policy )
Research Themes
None
Project Partners
None
Research Keywords
None
Relevant Degree Courses
None

Background

Biodiversity loss is threatening ecosystems and the services they provide to human society. To reverse this decline, we need to understand how human activities impact biodiversity, especially under future climate and societal change. Within this project, we aim to develop decision support tools based on state-of-the-art process-based models combined with AI techniques such as emulation and history matching. These approaches allow us to deal with unprecedented levels of computational complexity, enabling us to 1) model how individual species respond to climate and human-induced stressors with fine-grained detail at very large scales, 2) rigorously quantify uncertainties, and 3) dynamically re-calibrate models as new data emerges. By advancing these cutting-edge approaches, our goal is to provide decision-makers and land managers with robust, fast running decision support tools that allow for the systematic exploration of pathways to halt biodiversity loss and promote nature recovery.

PhD opportunity

Bending the curve of biodiversity loss requires an integrated approach that combines the refinement of the understanding of the processes governing biodiversity change from a natural standpoint, as well as the impacts of human activities on nature within evolving societal and policy contexts. We seek projects that address either or both themes, emphasizing the development of innovative tools that integrate existing or novel process-based biodiversity models with advanced AI techniques. These techniques may include model emulation using deep Gaussian processes, network emulation, or history matching and calibration methods that explicitly quantify and propagate uncertainty from diverse sources.
As our interest extends beyond the technical advancement of such modelling tools to their real-world applications, we welcome work that aims to develop and use such modelling tools, among others, to:
– Explore future trends in biodiversity changes under climate and policy uncertainty.
– Identify policy solutions that can reverse biodiversity loss and restore ecosystems even under uncertain future conditions.
– Assess the economic and social implications of strategies to achieve nature recovery.
We are particularly keen on research that emphasize the integration of such modelling tools into decision support tools co-designed with practitioners, policymakers, and land managers with a focus on how to better understand and tackle uncertainty in decision-making for nature recovery.

Other information

Applicant profile: Students with a strong interest in trans-disciplinary research and a solid background in mathematics, statistics and and computer modelling. We are open to consider students with a variety of backgrounds such as Economics, Mathematics, Statistics, Computer Science, Ecology, Geography.