Analysis and control of extreme events using Wetropolis flood demonstrator


The Wetropolis flood demonstrator [1] is a portable set-up designed to provide the general public. As such, it visualises what a return period is for extreme rainfall and flooding events. Unfortunately, people have difficulty comprehending what it means when a flood is classified as a 1 in 100 year flood (or one with an Annual Event Probability of 1%). It means that, on average, a flood of that magnitude occurs once in 100 years. That is based on data analysis and extreme-value theory. Note that it does not mean that once such a flood happens, it takes another 100 years for such a flood of that magnitude to happen again. Wetropolis showcases extreme rainfall and flood events with a 1 in 6:06min return period. This is short enough to wait for on average 6:06min and long enough to appreciate that it is a rare event on the time scale of viewing the set-up. Unexpectedly, Wetropolis drew the attention of flood professionals and academics. Consequently, they suggested to use it for research on flood control, data assimilation [2], effects of climate change, education and as a flood-science policy tool. In particular, we have now partly developed such a policy tool [3]. Notably, the project allows exploration of research in (one or a few of) these various directions. We can determine these together as a team.
Desired project outcomes
Given the above, we can formulate various exciting research aims formulated. Herein, Wetropolis’ flood demonstrator will function as a test bed for validation, as follows, on:
(i) developing flood control;  real-time flood control using machine-learning tools is a hot topic with direct applications in the UK; e.g., active control of reservoirs to buffer severe rainfall;
(ii) using data assimilation; flood prediction includes parameter estimation of properties such as riverbed roughness, required for accuracy but difficult to obtain;
(iii) assessing effects of climate change; flood classifications change due to climate change; better understanding of combined fluid dynamics and statistics is needed;
(iv) stimulating education; Wetropolis allows educational modules to be developed for outreach and education, involving full predictions and environmental fluid dynamics [2]; possibly involving a partnership with the School of Education, see also
(v) providing a validation environment for 3D/2D CFD; and,
(vi) enhancing flood-science policy; see [3].
[1] OB, Hicks, Zweers, Kent 2020: Wetropolis extreme rainfall and flood demonstrator: from mathematical design to outreach and research. Hydrology and Earth System Sciences 24, 2483-2503.
[2] Kent and OB 2020: Developments in hydraulic modelling and visualisation
[3] OB, Kelmanson, Kent 2020: Evidence entitled “A new tool for communicating cost-effectiveness of flood-mitigation schemes for the UK Government inquiry into flooding”.