Will climate change make sewage pollution worse?

genuinely hard: sewer networks are vast, unevenly monitored and hydraulically complex, while future rainfall carries deep uncertainty. Advances in deep learning for hydrological time series, alongside growing monitoring datasets, are opening new ways to predict spill risk at scale and propagate climate uncertainty through to real-world outcomes. This is a fast-moving research area spanning climate, hydraulic engineering, and data science, with potential to inform regulation and adaptation policy.

PhD project

This PhD project will reduce the uncertainty around how climate change impacts the pollution risk posed by our sewer network. Combined sewer overflows (CSOs) already spill untreated wastewater into UK rivers thousands of times a year, but our understanding of how a warmer, more erratic climate will change this is limited. Will sewer spills become more frequent, longer, more impactful?

This project builds a national deep-learning pipeline to answer these questions, then applies it to future climate projections. You will assemble a national ‘sewer-shed’ dataset, matching England’s ~20,000 monitored CSOs to environmental and infrastructural attributes, creating a database purpose-built for training deep-learning models. Using millions of observed spills, you will apply methods from hydrological deep-learning to learn how meteorological inputs and infrastructure traits combine to trigger spills. Explainable-AI will reveal which environmental variables and drivers control today’s spill patterns – a key insight in its own right. This trained model then becomes a fast emulator, rerun under future climate scenarios. Feeding it projected future rainfall, you will produce national projections of how spill frequency, duration and drivers shift as rainfall grows more extreme, identifying regions facing the greatest escalation in risk under a warming climate, and what could mitigate it. You will develop tools reducing uncertainty around this key freshwater stressor, helping protect our rivers as climate pressures grow.

You’ll gain expertise in machine learning (LSTMs, explainable AI), environmental data science, and scientific software development (Python, HPC/cloud), applied to a live, high-profile policy problem. Working with the supervisory team and drawing on their experience in public outreach (e.g., www.sewagemap.co.uk) and network of contacts across the water industry, you will have the chance to translate your outputs into real-world impact.

Applicant Profile

This project would be accessible to any quantitatively trained STEM student with a strong interest in machine learning and computational skills looking to apply their expertise to solving a real world environmental problem. The project focus on environmental engineering would be particular accessible to students with training in civil, or hydraulic engineering, but training on this specific domain will be provided by the supervisory team.

Funding Notes

This project is part of the UNRISK CDT, which offers 15-18 fully-funded NERC studentships, covering full university tuition fees; a personal stipend at standard UKRI rates; £6000 individual research and training costs; £5000 (per student) of cohort-level training; and a ‘Flexible Fund’ for special projects.

International applicants will need to cover costs related to applying for a student visa and the international health surcharge (IHS).

Applications are open to UK and international applicants. The number of awards for international applicants is limited by UKRI rules.

More information is available on the UNRISK website: View Website

References

https://www.sewagemap.co.uk/ – Developed by supervisory team as public oriented data-viz of sewage spill pollution
“Infiltration drives a heavy–tailed distribution of combined sewer overflow spill durations” – https://iopscience.iop.org/article/10.1088/2515-7620/ae60cc
“Rainfall–runoff modelling using Long Short-Term Memory (LSTM) networks” – https://hess.copernicus.org/articles/22/6005/2018/