Incorporating sedimentological data in UK flood frequency estimation

For the first time in the UK this study combines sedimentological flood data with instrumental river flow records into a unified flood frequency analysis. The results are then compared with flood estimates derived using traditional techniques and with systematic flow data alone. Published sedimentological flood data are re-examined for two contrasting catchments: the Nant Cwm-du, a small, steep upland ungauged stream in the Cambrian Mountains of Wales, and a piedmont reach of the River Severn in mid-Wales. Sedimentological flood estimates derived from a 129-year record of boulder-berm deposits in the Nant Cwm-du are incorporated into flood frequency analysis using a least-square method to fit a Generalised Logistic Distribution to the reconstructed flow data. Design estimates derived using these reconstructed flow data significantly exceed those derived from standard techniques. The River Severn is characterised by c. 5 m thick vertically accreted fine-grained overbank flood deposits, from which a c.3750-year flood series has been reconstructed using a geochemical grain-size proxy and 14C dating. The sedimentological flood data are combined with systematic river flow data and a flood frequency analysis undertaken using a maximum likelihood approach to estimate the parameters of the Generalised Logistic Distribution. In this case the incorporation of sedimentological data results in a reduction of design flow estimates when compared to those derived using systematic data alone. Recommendations for future research are identified and include the quantification of uncertainties associated with flood data derived from sedimentary sources, and the impact of the incorporation of sedimentological data on design estimate uncertainty. Future geomorphological studies should place more emphasis of developing robust estimates of palaeodischarge.