Linking Riemann and storage cell models for flood prediction

Abstract

As part of the nationally funded Flood Risk Management Research Consortium the use of linked one-dimensional and two-dimensional models has been investigated with a view to optimising the balance of accuracy, model complexity and run times. This has been developed in research software named TRENT using a Riemann approach for both dimensional representations. In this paper the details of the method are presented along with a storage cell scheme similar to that implemented in the software known as LISFLOOD-FP developed at the University of Bristol. Remotely sensed data are used to provide the digital terrain model for a test case at Upton-upon-Severn in the UK and airborne imagery is used for validation. The storage cell and Riemann approaches are compared and conclusions drawn.