Author(s): Qiuhua Liang, Xiaodong Ming, Xilin Xia
Linked Author(s): Qiuhua Liang, Xiaodong Ming, Xilin Xia
Keywords: Hydrodynamic model; Flood forecasting; Numerical weather prediction; UKV model; GPU computing;
Abstract: Flooding is the most damaging and costly type of hydrometeorological hazards affecting millions of people every year across the world. Flood forecasting, and the subsequent warnings, with sufficient lead-time is an important means to enable society to reduce risk and mitigate loss. Over the last few decades, major advances have been made in the fields of numerical weather prediction (NWP) and high-performance flood inundation modelling. However, no effort has been made to integrate high-resolution NWP with the latest high-performance hydrodynamic models to step-change the current flood forecasting practice. This paper aims to fill this gap and develop and test a real-time forecasting system for short-range and intense rainfall induced floods across a large catchment. Driven by the NWP outputs from the UK Variable (UKV) resolution model, the full-scale flooding process from rainfall to inundation is predicted using a high-performance full 2D hydrodynamic model that is substantially accelerated using modern GPUs. On a 10m uniform grid, the proposed system is capable of providing accurate flood extent and water depth and providing 34 hours of lead time for the severe flood event induced by 2015 Storm Desmond over the 2500km2 Eden Catchment in England. This demonstrates that the proposed forecasting system is feasible for predicting intense rainfall induced flooding at a catchment scale and may be applied to provide operational services.