Elizabeth Cooper, University of Reading
Improving Flood Models Using Data Assimilation
CoauthorsSarah Dance, Javier Garcia-Pintado, Nancy Nichols, Polly Smith
Abstract: River flooding is a costly problem in the UK and worldwide. Real-time, accurate inundation forecasting can help to mitigate damage caused by such events by warning people where and when flood water is likely to affect them. Data assimilation can be used to combine satellite derived water line observations with a mathematical model describing the flow of flood waters in order to improve the quality of the model's forecast. We present results of applying an ensemble Kalman filter to an idealized river flood situation with synthetic observations and show that the forecast ability of the Kalman filter is linked to the speed at which water flows through the system. The Kalman filter is then used to retrieve the value of the model parameter describing friction in the river channel, and the results from this simultaneous state-parameter estimation are compared with those from the state only estimation case.