Modelling the interactions between geomorphological processes and Natural Flood Management

Natural Flood Management (NFM) measures are being implemented across the UK and Europe in an effort to reduce flood impacts in a cost-effective and sustainable manner. Presently, NFM measures are often constructed without consideration of their geomorphological impact. There is also little evidence for NFM measures such as Runoff Attenuation Features (RAFs), which include leaky barriers, having an impact at larger (> 10 km2) catchment scales and for extreme events (> 100 year return period). This thesis examines both the hydrological and geomorphological effects of RAFs through; (i) morphodynamic modelling of RAFs of differing shapes, sizes and quantities at catchment scale (~41 km2) for an extreme event (120 year return period), (ii) comparison between a morphodynamic model and geomorphological processes estimated from outputs of a hydraulic model, which enables (iii) hydraulic modelling of differences in leaky barrier design.

The results indicate that a RAF can be designed to increase water storage and floodplain connectivity. However, hydrologically beneficial designs can produce erosion and deposition and scour to the feature itself. These effects are all localised and at catchment scale no notable peak discharge reduction was observed. Thus relying solely on RAFs will only likely reduce localised, low magnitude flooding and future research should look to strengthening this argument alongside increasing understanding of structure failure within networks of RAFs.

Qualitatively, the geomorphological outputs derived from CAESAR-Lisflood (a morphodynamic model) and HEC-RAS 2D (a hydraulic model) agreed with each other. However, methodological refinement is needed before hydraulic model outputs can be repurposed into geomorphological process estimations on a wider, national scale with little detailed validation. Overall, the HEC-RAS 2D methodology should be used to consider local problems requiring a high spatial resolution including changes in infrastructure design. CAESAR-Lisflood should be used to consider catchment dynamics and where NFM measures induce bed morphology changes.