Investigating the effects of large wood and forest management on flood risk and flood hydrology

Abstract

The changes to catchment scale flood risk following river restoration works, including the addition of large wood logjams to the channel, are poorly quantified in the literature. Key concerns following river restoration for river managers and other stakeholders are changes to flood hydrology at the reach and catchment scale and changes in the mobility of large wood pieces. The effects of accumulations of large wood (logjams) on local flood hydrology have been documented in the literature, showing logjams slow flood wave travel time and increase the duration and extent of local overbank inundation. Modelling studies conducted at a reach scale have shown that these local effects can slow flood wave travel time through a reach and delay the timing of flood peak discharge at the reach outflow. How these local and reach scale effects translate to the catchment scale remains to be illustrated in the literature. In this thesis a combination of field and modelling studies are used to; elucidate the link between logjam form and function, to quantify the mobility of pieces of large wood relative to their physical characteristics, to predict the changes in floodplain forest restoration over time and to provide predictions of changes to catchment scale flood hydrology following river restoration at a range of scales and locations. It is shown that logjams inducing a step in the water profile are most effective at creating diverse geomorphology and habitats. Logjams were found to account for 65% of flow resistance in forested river channels, rising to 75-98% of flow resistance where the logjam was inducing a step in the water profile. Large wood in small forested river channels was found to be highly mobile with 75% of pieces moving, with the longest transport length of 5.6km. Large wood mobility is governed primarily by the length of a piece of wood with wood in excess of 1.5x channel width a threshold for a lower probability of movement. Hydrological modelling using OVERFLOW shows that reach scale river restoration can lead to modest changes in catchment scale flood hydrology. It is concluded that flood risk management can incorporate river restoration, but that results are likely to be unpredictable if engineered logjams are used alone. Substantial benefits in reducing catchment outflow peak discharge (up to 5% reduction) are modelled for floodplain forest restoration at the sub-catchment scale (10-15% of catchment area), rising to up to 10% reductions as the forest matures and becomes more complex.