University College London is advertising a competitive PhD on climate change and sewage-pollution risk using deep learning. UK and international applicants may compete for funding. The deadline is 13 January 2027. Applicants should consider the research fit and financial conditions together before deciding whether to apply.
Research Focus
The project builds a national dataset and models of combined sewer overflows. It examines how weather and infrastructure relate to spills. Future climate scenarios and explainable AI support the investigation of changing risk. Read the project description as a starting point for identifying a question that interests you, rather than assuming that every possible method will appear in the eventual thesis.
Location and Academic Team
The university base is London, United Kingdom. The research concerns England’s sewer network. Earth Sciences hosts the advertised opportunity. It forms part of UNRISK. Dr Alex Lipp is the named supervisor. The project connects environmental data science with water-sector applications.
Who Can Apply
Quantitatively trained STEM applicants with machine-learning interests and computational skills are relevant; environmental-engineering training will be provided. Check the current programme requirements for the precise qualifications and supporting documents needed. A relevant subject or an eligible nationality does not by itself guarantee admission or an award.
Funding and International Conditions
UNRISK offers 15–18 awards across the programme, covering full tuition, a UKRI-rate stipend, £6,000 individual research support and £5,000 cohort training per student. International awards are limited, with visa and Immigration Health Surcharge expenses remaining personal costs. Confirm the applicable stipend and written award terms before making financial commitments.
Exploring the Research Problem
For optional preparation, imagine a model that performs well at frequently monitored locations but has not been assessed elsewhere. Ask what information would be needed before using it to compare areas with different monitoring histories. Consider whether a missing observation means that an event did not occur or simply that it was not recorded. This is a hypothetical research-thinking exercise, not an assessment task, a statement about a real water system or a guarantee about the data available to the project.
You could extend the example by comparing performance during ordinary conditions with performance during unusual inputs. Explain why an apparently strong overall score might not establish that the model answers the particular risk question you care about. Identify a comparison that would help expose that limitation and what uncertainty would remain afterwards. A bounded explanation can demonstrate thoughtful interest in environmental machine learning without suggesting that a preliminary model is ready to determine practical engineering or regulatory decisions.
Presenting Evidence of Preparation
If you have worked with time-series data or a predictive model, explain the question, the data, your individual contribution and the evaluation you actually performed. Describe how you checked for a misleading comparison or a result that depended on an unexpected assumption. If your previous application area is different, identify the transferable computational skills rather than claiming specialist knowledge of hydraulic systems. Connect those strengths to a realistic learning goal and distinguish a completed exercise from an independently validated research tool or deployed operational system. Keep a precise account of which checks you completed and which possible extensions remained outside the scope of the work.
If supporting materials are requested, make your individual contribution clear when discussing work completed with other people. Distinguish a class exercise, an unpublished report and a peer-reviewed research output instead of presenting them as equivalent achievements. Explain what your evidence can establish and what remained unresolved, including any limitation that would matter to the conclusion. Adapt these preparation suggestions to the actual application questions, since optional advice does not add requirements to the programme’s document checklist or guarantee a particular assessment outcome.
How to Apply
Use the current programme instructions to establish the submission route and complete supporting-document checklist. Review the exact project title and academic contacts wherever the form asks you to identify your research preference. An informal academic conversation can clarify fit, but it does not replace a formal application or constitute an admission or funding offer.
- View scholarship details and follow the linked institutional application instructions.
- Identify this named project and check the current academic and funding requirements.
- Submit through the specified route by 13 January 2027 and retain confirmation.
Before Submitting
Information was checked on 23 October 2026. Reconfirm the deadline and current terms before submission, particularly where your ability to enrol depends on receiving an international award. Keep a copy of the final form and any written clarification that materially affects your eligibility, financial expectations or decision to accept an offer.