AI-enhanced Climate Storylines for Agricultural Adaptation and Resilience
Agricultural systems are increasingly exposed to climate extremes such as droughts, floods, and heatwaves, which are intensifying due to climate change and threatening food production, land use, and rural livelihoods. Conventional climate projections often lack the spatial resolution, sectoral relevance, and narrative framing needed for effective decision-making in agriculture. Climate storylines have emerged as a powerful tool for communicating climate risk and uncertainty, offering physically consistent narratives of extreme events or climate pathways. However, generating meaningful climate storylines for agriculture is challenging, as conventional climate models struggle to capture extreme events at the spatial and temporal scales relevant to farming decisions.
This PhD project at Coventry University’s Centre for Agroecology, Water and Resilience aims to develop a novel framework for generating AI-enhanced climate storylines tailored specifically to agricultural adaptation. The successful candidate will leverage recent advances in AI and machine learning-based downscaling of global climate model outputs, alongside other AI applications, to identify patterns of past and projected extremes. These data-driven storylines will be refined through participatory co-production with agricultural stakeholders, ensuring they reflect local realities, vulnerabilities, and decision contexts. The project will also explore the potential for agentic AI to autonomously explore high-resolution climate datasets and identify plausible projections of event types relevant to specific agricultural regions or stakeholder concerns.
Key research questions include: How can AI tools generate high-resolution, stakeholder-relevant climate storylines for agriculture? What combinations of climate drivers and impacts are most important for agricultural planning? How can local knowledge and agricultural thresholds be integrated into climate storyline development? What role can agentic AI play in supporting the co-creation and interpretation of climate narratives?
Supervision is provided by Associate Prof. Jonathan Eden (Coventry University), Associate Prof. Robert Faggian (Deakin University), Prof. Matthew England (Coventry University), and Dr. Bahareh Nakisa (Deakin University). The project offers tuition fees and a stipend, supporting the doctoral candidate throughout their research.
Applicants should have strong quantitative skills and experience in statistical modelling, machine learning, climate data analysis, or environmental modelling. Experience with scientific programming languages such as Python or R is required or willingness to develop these skills. Interest in climate change, agricultural systems, and climate risk communication is essential. The ability to work across disciplines and engage with stakeholders is important, as is good analytical skill and attention to detail when working with large and complex datasets. Entry requirements include a minimum of a 2:1 first degree in a relevant discipline with at least 60% in the project element or equivalent and a minimum 60% overall module average. English language proficiency (IELTS academic overall minimum score of 7.0 with a minimum of 6.5 in each component) is mandatory.
The application deadline is April 14, 2026. To apply, submit full supporting documentation, a covering letter, and a 2000-word supporting statement showing how your expertise and interests are relevant to the project. For further information, contact Associate Professor Jonathan Eden. Apply via the provided application link.
