Location
Newcastle University

Supervisors: Hayley Fowler (NU), Amy Green (NU), Matthew Fry (UKCEH), Sara Alexander (EA)

Contact email: Hayley.Fowler@newcastle.ac.uk

Project rationale

Increasing frequency and intensity of extreme rainfall events from climate change pose significant challenges for UK flood risk management. Current precipitation datasets exhibit substantial limitations: radar precipitation estimates suffer from inaccuracies, while rain gauge networks, though more reliable, lack the density required to capture localised convective storms. The UK has a dense network of tipping bucket rain gauges (e.g., EA, SEPA, NRW), satellite coverage and comprehensive weather radar (Met Office). Blending these datasets to provide a multisource dataset can achieve much higher accuracy and broader spatial coverage for precipitation estimates, with potential benefits highlighted by successful initiatives for hourly datasets (e.g. UKGrsHP [1], CAMELS-GB2).

This research aims to explore optimal methods to provide blended high-resolution (15-min) UK precipitation datasets, developing bias-correction, disaggregation and quality control methods to preserve extreme rainfall statistics. This has been identified as a high impact and high priority action from the recent Environment Agency review of uncertainty for flood hydrology. Resulting precipitation datasets will significantly enhance understanding of rainfall dynamics, inform flood risk assessment, and support climate adaptation. This is critical for addressing the urgent need for reliable rainfall data in a rapidly changing climate, with wide-ranging implications for hydrodynamic flood modelling and infrastructure resilience.

Methodology

The project aims to produce a high-resolution, reliable precipitation dataset that addresses critical challenges in flood risk assessment and climate adaptation.

  1. Robust data preparation: Integration of large datasets from various sources (e.g., tipping bucket gauges, radar QPE, and satellite data) using coding best practice to ensure reusability.
  2. Quality control and assessment: Develop comprehensive QC framework: evaluate data quality metrics, remove suspicious data, implement interpolation techniques, correct for systematic biases.
  3. Multisource blending of data: Informed by (2), assess multisource blending methods (e.g., Kriging, Gauss blending, probabilistic merging [2], ML).
  4. Uncertainty quantification: Integrate information on multiple sources of uncertainty (measurement accuracy, QC, network density, spatial data, blending method) and evaluate impacts on blended dataset. Develop methods to provide information to users on uncertainty (e.g., ensemble, data confidence intervals, etc). This will include assessment of the use of AI/ML. Demonstrate impact on flood predictions by propagating uncertainty through hydrological models.
  5. Open access data and applications: Deliver final blended dataset and open QC tools to users via FDRI to facilitate a wide range of applications. The student will engage with industry partners to demonstrate dataset’s utility through applications focused on climate change impacts or flood dynamics, ensuring dataset meets real-world needs and can contribute to improved flood risk management strategies. 

Background reading

[1] Yu, J., Li, X.-F., Lewis, E., Blenkinsop, S., & Fowler, H. J. (2020) UKGrsHP: a UK high-resolution gauge–radar–satellite merged hourly precipitation analysis dataset. Climate Dynamics, 54(5), 2919–2940. https://doi.org/10.1007/s00382-020-05144-2

[2] Ochoa-Rodriguez, S., Wang, L.-P., Willems, P., & Onof, C. (2019) A Review of Radar-Rain Gauge Data Merging Methods and Their Potential for Urban Hydrological Applications. Water Resources Research, 55(8), 6356–6391. https://doi.org/10.1029/2018WR023332

[3] Kossieris, P., Tsoukalas, I., Brocca, L., Mosaffa, H., Makropoulos, C. and Anghelea, A. (2024) Precipitation data merging via machine learning: Revisiting conceptual and technical aspects. Journal of Hydrology, p131424. https://doi.org/10.1016/j.jhydrol.2024.131424.

FLOOD-CDT

This PhD is being advertised as part of the Centre for Doctoral Training for Resilient Flood Futures (FLOOD-CDT). Further details about FLOOD-CDT can be seen here https://flood-cdt.ac.uk. Please note, that your application will be assessed upon:

  1. Motivation and Career Aspirations;
  2. Potential & Intellectual Excellence;
  3. Suitability for specific project and
  4. Fit to FLOOD-CDT.

So please familiarise yourselves with FLOOD-CDT before applying. During the application process candidates will need to upload:

  • a one-page statement of your research interests in flooding and FLOOD-CDT and your rationale for your choice of project;
  • a curriculum vitae giving details of your academic record and stating your research interests;
  • name two current academic referees together with an institutional email addresses; on submission of your online application your referees will be automatically emailed requesting they send a reference to us directly by email;
  • academic transcripts and degree certificates (translated if not in English) – if you have completed both a BSc and an MSc, we require both; and
  • a IELTS/TOEFL certificate, if applicable.

Please upload all documents in PDF format. You are encouraged to contact potential supervisors by email to discuss project-specific aspects of the proposed prior to submitting your application. If you have any general questions please contact floodcdt@soton.ac.uk.

Apply

Apply for this PhD here: https://www.ncl.ac.uk/postgraduate/fees-funding/search-funding/?code=FL…