Historic spatial patterns of compound flood events in UK estuaries

Abstract

<p>Combination estuarine flooding is driven by extreme sea-levels and river discharge occurring at the same time, or in close succession. We hypothesise the drivers of flooding rarely occur independently and operate and co-operate at sub-daily timescales. There is a need to accurately capture fluvial and sea-level interactions at a sub-daily timescale to understand the relative timing and duration of compound flooding hazards in estuaries to support forecasts and warnings, emergency response and long-term management plans. This research analyses the co-cooccurrence of extreme sea-level and river discharge events using historic river flow data from 126 gauges and tide gauge data from 27 locations across Britain, which were analysed to identify estuaries that are susceptible to compound flooding events. Daily mean and maximum river discharge, and river discharge at 15-minute frequency were analysed to identify extreme peaks in the records, and corresponding skew surge values identified within a time window based on average hydrograph duration. Results show that daily mean river discharge underestimates peaks in the record and does not accurately capture hydrograph behaviour. This research is the first time that 15-minute frequency river discharge data has been used to characterise hydrograph behaviour and identify i) peaks over threshold; ii) top 500 peaks and corresponding skew surges to determine dependence based on Kendall’s rank correlation and the number of occurrences between extreme drivers each storm season (May-June). Different methods of data selection and identification of peak river discharge events generates different results. The duration of river discharge peaks, total water level, lag time, and overlap between peaks is calculated to identify locations where co-occurrences are likely to happen. The results identify a clear east-west split in dependence, with gauges on the west coast of Britain showing stronger correlations. There are more co-occurrences of extreme sea and fluvial levels each storm season in Northwest England and West Scotland. Estuaries that are most susceptible to compound events based on the Kendall’s rank correlation τ, seasonal occurrences, and potential for river discharge and sea-level peaks to overlap based on duration and lag time are identified. The results identify 46 gauges that are highly susceptible to compound flooding, notably the Rivers Lune and Eden, with strongest correlation coefficients and highest number of seasonal occurrences. The results highlight spatial variability to the sensitivity of estuaries to combination flood hazard, and the impact of future changes in flood risk that are unresolved currently in management plans.</p>