A model for wheat cultivars and optimisation for climate scenarios – Sim Farm 2030 (PhD)

The development and assessment of new crop cultivars is essential to ensure food security in a changing climate. In UK wheat, a relatively simple comparison of cultivars, at a relatively small number of sites, is used. However, this approach is unlikely to fully account for variation in crop performance, particularly variation linked to weather and soil. It is also unable to predict varietal performance under climate change scenarios. Recent research has established the potential to simulate yields under various scenarios. Named ‘Sim Farm 2030’, the model requires further development and validation before it can be used to guide decision making (e.g. in crop breeding). This PhD studentship project will develop Sim Farm 2030, so that it can provide decision support to aid the optimal selection of wheat cultivars for UK conditions and potential climate scenarios. The work will apply cutting-edge machine-learning, data-driven techniques to model the yield of wheat cultivars, as a function of meteorological (e.g. temperature and precipitation) and environmental (e.g. pollution and soil) variables. Pilot work (funded by the STFC Food Network+), which developed a simple US maize-yield model and preliminary testing on UK wheat cultivars, has established proof of concept for the approach. This PhD project is an unusual interdisciplinary enterprise, supervised jointly by an astronomer with extensive data science skill and a crop scientist. The student will make regular visits to the Met Office, interfacing the tools with the latest climate projections, and an additional six months on a placement with Quant Foundry exploring the integration with commercial tools and platforms.