Risk CDT - Drought Risk: developing a better understanding in drought magnitude, frequency and severity

PLEASE APPLY ONLINE TO THE SCHOOL OF ENGINEERING, PROVIDING THE PROJECT TITLE, NAME OF THE PRIMARY SUPERVISOR AND SELECT THE PROGRAMME CODE "EGPR" (PHD - SCHOOL OF ENGINEERING)

This is a project within the multi-disciplinary EPSRC and ESRC Centre for Doctoral Training (CDT) on Quantification and Management of Risk & Uncertainty in Complex Systems & Environments, within the Institute for Risk and Uncertainty. The studentship is granted for 4 years and includes, in the first year, a Master in Decision Making under Risk & Uncertainty. The project includes extensive collaboration with prime industry to build an optimal basis for employability.

Whilst considerable work has focused on the risks and uncertainties associated with flooding, risks from droughts are often underestimated. This proposal would address this issue by developing a better understanding in drought magnitude, frequency and severity; whilst the approaches used in assessing drought magnitude are similar to the approaches used in examining floods, the frequency and severity of droughts require different approaches.

Droughts are not characterised as discrete events in the manner that flood can be considered, as droughts are cumulative and the severity of a drought is not simply a reflection of the magnitude, but also the duration, as such any assessment of frequency needs to consider this. Previous research within the group has focused examining droughts have predominantly focussed on drought structure (Lennard et al., 2014), and the wealth of long records often available from which drought indices can be generated (Todd et al., 2013).

This research will use specific drought indices in the form of the Palmer Drought Severity Index (PDSI) and Standardised Precipitation Index (SPI) to derive long drought indices from which new models of analysis can examine the uncertainties and risks associated with droughts. For a realistic prediction of droughts we aim at developing a suitable and powerful numerical model. From a mathematical point of view the occurrence of droughts appears as a random process with quite significant memory features. These memory features refer, in particular, to the cumulative characteristics of the drought phenomena. This excludes the utilisation of the commonly used and well-established Markov type models for process simulation and prediction.

The process model must, further, be formulated in the context of climate data, such as precipitation and temperature, and it must be able to reflect non-stationarity due to climate change effects. Stochastic process models for numerical simulation which satisfy these requirements ad hoc are not available.

Our development will start from a probabilistic modelling of physical mechanisms as far as these are known, examine past drought events and utilise conditional physical input from available climate models and supplement this basis by statistical means. For modelling in time-frequency domain, wavelet transform appears promising as it basically provides the flexibility required. To cope with complexity and limited physical insight, neural networks provide a promising module for the process model as sufficient data are available for their training.

Remaining model uncertainty, for example in the climate prediction and in dependencies and memory features, can be addressed with methods of imprecise probabilities. Imprecise probabilistic models cover an entire set of plausible probabilistic models and help to reduce the risk of missing critical situations.

The outputs of this work are considerable as they will contribute significantly to better policy development and risk assessment in water resource management.

Any special features: (e.g. equipment, collaboration, industrial links, underpinning expertise)

The project requires a sound mathematical/analytical background, curiosity, creativity and a strong interest to work in a multi-disciplinary set-up. Applicants should have a background in the Environmental Sciences or Engineering (preferably Civil or Systems Engineering) or in other relevant fields. A combined education in these fields would be an advantage.

The student will join a multi-disciplinary research group in the Institute for Risk and Uncertainty.