Vacancy terms: Full time or job share, fixed term from 01 January 2027 until 31 March 2028 About the role
As Research Associate in AI for Environmental Forecasting (SENTINEL project), you will lead the AI development strand of SENTINEL, a UKRI-funded project developing a physics-informed AI forecasting system for chemical fugitive emissions. You will build a multi-modal training dataset and design, train and evaluate the Receptor Transfer Function Learner, combining generative AI downscaling, graph neural networks and explainable AI to forecast community exposure to fugitive emissions 24 to 72 hours ahead. You will work with real-world environmental, meteorological, operational and community data, applying rigorous, reproducible research methods and using GPU and high-performance computing facilities. Working within the School of Engineering, you will collaborate with an interdisciplinary team spanning AI, atmospheric dispersion, chemical fate and public health, as well as project partners including the UK Health Security Agency and the Met Office. You will benchmark SENTINEL against established dispersion models, contribute to high-quality research publications and dissemination, and help develop collaborations and future research opportunities. The role sits within a research environment where AI is being applied to complex datasets and global challenges, including environmental monitoring and sustainability. About you You will have a PhD, equivalent professional qualification or equivalent experience in computer science, machine learning, data science, environmental data science, physics, engineering or another related quantitative discipline. You will bring demonstrable experience of developing and training machine learning or deep learning models using real-world, noisy or time-series data, alongside strong analytical problem-solving skills and experience contributing to peer-reviewed publications or equivalent research outputs. You will have strong Python programming skills, including experience with modern AI or deep learning tools such as PyTorch, TensorFlow or JAX and scientific Python tools such as NumPy, pandas or xarray. You will be able to work effectively both independently and as part of a collaborative research team, communicate complex information clearly and engage confidently with internal and external stakeholders. Experience with graph neural networks, generative or diffusion models, physics-informed machine learning, explainable AI, weather or geospatial data, atmospheric dispersion modelling, or training models using GPU clusters or HPC would be advantageous. You will join a research environment in which high-performance AI, physics-informed approaches and digital twins are being developed for applications including environment, health and engineering. £40,704 per annum, pro-rata if part-time: Grade 7
