The simulation of floodplain flows: Implications for floodplain management and river restoration strategies

The high incidence of devastating flood events around the world in the 1990s and the growing interest in the restoration of river systems have raised technical questions about the selection of appropriate flood simulation tools. The restoration and integrated management of rivers and floodplains requires careful judgment of the consequences of alterations to channels and the natural physical processes. A range of numerical models are available to simulate the hydrodynamic characteristics of floodplain flows. One-dimensional models are simple and require less data but may be inadequate to capture the complex floodplain processes. An emerging question within the scientific community is whether the potential reduction in flood risk due to floodplain restoration can be accurately assessed by a simple 1-D formulation. Two-dimensional models are more complex and capture most of the processes, but require a finer spatial grid and calibration data as well as a larger computational effort. These constraints limit the application of 2-D models to small reaches, but integrated management requires a holistic analysis to comprehend the effect of management actions not only within local project areas but also at the regional and watershed scale.; This study develops a 2-D model formulation for the analysis of complex floodplain processes at a large scale and establishes guidelines for the selection of 1-D or 2-D model for specified hydraulic and topographic conditions. A 1-D parameterization of the main channel has been incorporated into a fully dynamic 2-D algorithm for the floodplain flows by modifying the conventional form of the parameterization in the depth-integrated shallow water equations. This reduces the computational effort and allows large scale 2-D modeling. The guidelines for appropriate model formulation combine the concept of section coherence and non-dimensional analysis to generate criteria that account for stage dependent hydraulic and topographic variables. By adoption of a consistent roughness formulation, the results of 1-D and 2-D models can be compared directly. The validity of both the model and the guidelines are evaluated for the laboratory data of the Flood Control Facility (FCF) in the United Kingdom and the Lower Boise River of Idaho.