Forecast Assimilation For Resilient Water Resource Planning And Operations

This is an ideal project for a resourceful and driven individual. The successful candidate will develop and perfect first-rate analytical, computational and communication skills that will enable them to choose their subsequent career path, whether in academia or industry.

Weather and climate forecasting capabilities are quickly improving, with implications for predicting floods and droughts at a range of timescales. This generates a large quantity of information. For stakeholders concerned with avoiding floods, droughts and other adverse water-related events, the challenge is to process this information.

Most research in the use of forecasts to inform water resources planning and management asks: “what are the expected (average) benefits of a forecast product X for water system Y”? This PhD thesis proposes to ask instead: “What are the forecast characteristics that would increase the resilience of water system Y to climate-related risks? What variables? What lead times? And with what accuracy?” For instance, one would expect a probabilistic seasonal forecast for drought anticipation and mitigation to be most beneficial in dry years, so the relevant question is not what its benefits are on average, but how and when to extract early-warning indicators that will inform mitigation measures in a timely manner.

The work will typically comprise the following steps:

[1] On a case-study (to be chosen among several possibilities either in the UK or in a developing country), machine learning techniques will tell us which part of the available forecast information is the best predictor of adverse events.

[2] In a second step, mitigation measures are tested to understand how these predictors help to trigger them, and what the benefits are. This step will also test how improvements to prediction accuracy may benefit mitigation.

[3] Finally, the research will investigate infrastructure investments to adapt to climate change. Integrating forecasts in how new infrastructure is operated will help planners to make better decisions.

Note this plan is indicative: there will be opportunities for the successful candidate to co-shape the project and start collaborating with top research groups in the UK and abroad.

Pre-requisite qualification
Suitable for candidates holding or anticipating award of an MSc, or 1st/2.1 undergraduate degree in an engineering or numerical/physical sciences discipline. A taste for programming is essential and prior experience is a plus.