Summary of results for national scale recharge modelling under conditions of predicted climate change

This report describes the application of the BGS distributed recharge model ZOODRM to produce recharge values (potential recharge) for Great Britain (England, Scotland and Wales). This model has been run with the rainfall and potential evaporation for the Future Flows Climate datasets (11 ensembles of the HadCM3 Regional Climate Model or RCM). The following results have been produced: • For groundwater bodies in England and Wales: o The mean, standard deviation and the following percentiles: 10, 25, 50, 75, 90 (absolute values of annual recharge produced by ranking annual recharge values) have been produced for annual recharge totals for the following periods: simulated historic (1950-2009), 2020s (2010 - 2039), 2050s (2040 - 2069) and 2080s (2070 - 2099). o The 25th percentile and 75th percentile for the simulated historic recharge for each month have been calculated. The estimated daily recharge values were aggregated to monthly values first and the analysis was undertaken using these monthly values. Further, a proportion of recharge values above and below these values for the future climate has been calculated. o Mean monthly recharge values were calculated for each month for the simulated historic period. The change in recharge value for each month in absolute terms compared to monthly value calculated for the historic simulation was calculated for the 2020s (2010 - 2039), 2050s (2040 - 2069) and 2080s (2070 - 2099). o Monthly change factors (percentage difference between monthly average recharge for future climate and historic simulation) for each groundwater body for each of the 11 ensembles were produced. These have been summarised in maps of England and Wales, which illustrate for each month the minimum, maximum and median monthly change factor from all the ensembles for each groundwater body. • River Basin Management Districts (RBMD) in England and Wales: o The mean monthly recharge value was calculated for each month for the RBMD. The change in recharge value in absolute terms was calculated for the 2020s (2010 - 2039), 2050s (2040 - 2069) and 2080s (2070 - 2099). o The total recharge volume for the RBMD for the time periods 1961-90, 1971-00 and for the 2020s (2010 - 2039), 2050s (2040 - 2069) and 2080s (2070 - 2099) was calculated. o Empirical cumulative distribution functions (ECDF) have been produced for seasonal (spring, summer, autumn and winter) as well as monthly averages for historic simulation (both 1961-1990 and 1971-2000) as well as for the 2020s, 2050s and 2080s. Generally the recharge season is shorter in the future. For the historical simulation (1950-2009) the recharge season is between five to seven months each year (September to April). It appears that this is reduced to three to four months for the future climate predictions. This is seen in both the changes in 25% / 75% recharge values and the monthly differences. There appears to be agreement between ensemble outputs. This could make aquifers more vulnerable to droughts if rainfall fails in one or two months rather than a prolonged dry winter as can occur now. When recharge volumes were produced for the RBMDs then the volumes tend to increase from the historical simulation to the 2020s/2050s, but more significantly in the 2080s. For example in the Thames RBMD the average recharge volume increases from 67 x 106 Ml/d in the 2020s/2050s to just over 73 x 106 Ml/d in the 2080s. However, the range of possible outcomes also increases and so one possible future outcome is that recharge volumes could reduce. The recharge season appears to be forecast to become shorter, but with greater amount of recharge “squeezed” into fewer months. This is acceptable for ensuring that recharge for groundwater water resources is maintained from a water balance perspective, but could result in greater “lumpiness” of the recharge signal. This increased “lumpiness” could result in flashier groundwater level response and potentially greater drought vulnerability. Groundwater drought could, therefore, occur if rainfall “fails” for one month, i.e. recharge totals are reliant on fewer months of rainfall. Finally, the results show that the balance between climate variability and climate change shifts towards the end of the future period (2010-2099) with a stronger climate signal being observed in changes to the recharge values in the 2080s than either of the 2020s or 2050s. Given the amount of data produced, a more detailed examination of the results for groundwater bodies would enable more value to be gained from the work. Alongside this, understanding how water balances for the RBMD varies in the future would be beneficial. Three issues should be examined: 1) Disaggregation of recharge volumes for the River Basin Management Districts to examine how recharge to individual aquifers may change; 2) Shortening of recharge season and vulnerability to drought; and 3) Variability of results from the ensembles and likely worse cases. Finally, whilst the initial analysis has focussed on how recharge will change for water resources, no consideration of groundwater flooding has been included and this should be examined