The use of catchment-scale riparian intervention measures in downstream flood hazard mitigation

Abstract

In recent years there has been debate over increasing flood risk in Britain, whilst the perception of increased flood risk is becoming more prevalent. This comes at a time when funding for flood defence construction and maintenance is thought to be insufficient, but public spending is facing contraction.
This study explores the potential of diffuse, small-scale interventions placed throughout the Uck catchment to reduce the flood peak downstream at Uckfield. Catchment Riparian Intervention Measures (CRIMs) in this study take the form of woody debris dams and riparian vegetation, designed to reduce flow conveyance, and attenuate flow through local flooding. Few studies have investigated the catchment-scale effects of spatially diffuse flood risk reduction measures due in part to the large computational requirement of modelling numerous spatial arrangements of potentially a wide variety of intervention measures.
The reduced complexity hydrological model Overflow was therefore chosen for this study. In contrast to more complex models, Overflow allows the spatial arrangement of flood risk reduction measures to be investigated rapidly, and with ease.
First, the performance of Overflow is evaluated by carrying out sensitivity and uncertainty analyses. It was found that the presence of a number of homogenous parameters, namely roughness parameters, do not prevent useful results from being obtained. However, other parameters are more important to model output. This led to the partial calibration of Overflow to improve the temporal representation of catchment hydrological response throughout the simulated storm event. Following this, Overflow simulated to a good degree the hydrograph observed downstream of Uckfield during the October 2000 flood event.
Second, the effect of the spatial arrangement of CRIMs placed throughout the Uck catchment was explored. The location of a CRIM in the catchment had an important influence on its effectiveness as a risk reduction measure, conditioned to a small degree by local channel and floodplain properties. In a number of reaches, the placement of a CRIM increased downstream peak flow, due primarily to modification of the relative flood wave timing. When several CRIMs were added to reaches throughout the catchment, interaction effects became important. The effect of a CRIM when others were placed throughout the catchment could not be predicted by its effect in isolation. An intervention strategy is developed. Overall, by placing CRIMs in 52 reaches throughout the Uck catchment, simulated peak flow at Uckfield was reduced by 12.5 cumecs, from an initial peak of 124.7 cumecs.