Practicalities for implementing regionalised allowances for climate change on flood flows: initial results

Project FD2020 ‘Regionalised impacts of climate change on flood flows’ provided a methodological framework designed to enable the quick estimation of the impact of a set of climate change scenarios on the flood flows of a catchment. The method separates the climate change that a catchment may be exposed to (the hazard) from the catchment response to changes in the climate (the vulnerability, in terms of change in four flood indicators), through use of a sensitivity framework. Re-combining vulnerability and hazard then leads to an estimate of the risk of change in flood flows. This extension to project FD2020 uses FD2020’s representation of vulnerability, along with sets of climate change scenarios from UK Climate Projections 09 (UKCP09), to estimate the probabilistic risk for each of the National River Flow Archive (NRFA) catchments in England and Wales. This report describes the methodology applied to estimate the vulnerability of each of the NRFA catchments for each flood indicator (flood peaks with a return period of 2, 10, 20 and 50 years), summarises the hazard as defined by UKCP09 for river-basin regions over England and Wales, and presents the resulting estimates of risk (with uncertainty ranges). The vulnerability of each catchment is estimated through the use of the FD2020 response types, and the catchment properties which characterise those response types (via decision trees), rather than through direct hydrological modelling of all NRFA catchments. Minor modifications are made to the FD2020 decision trees before using them to estimate the response type of each NRFA catchment, to make them more robust for use with a wider range of combinations of catchment properties. The hazard is derived from the UKCP09 climate scenarios, for the 2080s time-horizon under the A1B (Medium) emissions scenario, for the 12 river-basin regions covering England and Wales: Northumbria, Humber, Anglian, Thames, South-East England, South-West England, Severn, West Wales, North-West England, Solway and Tweed. In each case, a single harmonic function is fitted to each of the 10,000 sets of monthly changes in precipitation and temperature. The distributions of the three parameters of these fitted harmonic functions (mean, amplitude and phase) are assessed against the sets of changes applied in the FD2020 sensitivity framework. In general, the two sets compare favourably. The hazard for a given river-basin region is combined with the vulnerability for each response type. This is done by using the mean and amplitude of the fitted precipitation harmonics (hazard) to extract the estimated impact from the key response patterns (vulnerability) for each response type and flood indicator. The set of 10,000 extracted impacts then represents an initial estimate of the range of risk (due to climate modelling uncertainty and natural variability) in each case. The appropriate uncertainty allowances are added, to get a more robust estimate of the range of risk. The results are presented by river-basin region. The use of the UKCP09 Sampled Data for river-basin regions simplifies the results, as there is one set of results for each response type in each river-basin region at each return period, rather than for each catchment at each return period. For each region, a regional risk is also calculated at each return period, based on the number of NRFA catchments of each response type. The results present a wealth of information to support the update of guidance on flooding and climate change, and decisions by the policy-maker on what level of detail/complexity is most appropriate: • Revised nationwide allowances; • New regional allowances; • New regional / sub-regional allowances by response type; • Local decision-making, where the tools are provided to evaluate changes to river flows from user-defined future climate scenarios and catchment characteristics. The benefits and weaknesses of each level are discussed, including the balance between simplicity of guidance and the possibility of over/under-adaptation. Decisions also have to be made by the policy-maker on the preferred level of protection and how much uncertainty information is taken into account. The next part of the project will look at alternative time-horizons and emissions scenarios, and consider the extension to higher return periods