OR/17/026 Summary of key results and narrative

Mansour, M M, and Hughes, A G. 2017. Summary of results for national scale recharge modelling under conditions of predicted climate change. British Geological Survey Internal Report, OR/17/026.

Introduction

Given the sheer volume of outputs produced (see Appendices) and to ensure that the report is as readable as possible, the main story has been summarised into a single narrative using a selected sub-set of model results. Whilst care has to be taken to avoid bias in selecting the results, using a reduced set of figures makes the story accessible to as wide an audience as possible.

The following results are selected:

• One set of seasonal (spring, summer, autumn and winter) recharge changes (as seasonal average value in mm/d) for each groundwater body for all of the 11 ensemble member for the 2020s, 2050s and 2080s (See Figures 3 to 5). These are based on the long-term average values for each period.
• One set of monthly averaged recharge changes (as a percentage) for each groundwater body for the median for all 11 ensemble member for the 2050s and 2080s (See Figures 6 and 7). These are based on the long-term average values for each period.
• Plots of changes in recharge (long-term monthly average as mm/d) for three (North-west, Humber and Thames) River Basin Management Districts for each ensemble member along with histograms of minimum, maximum and average of recharge totals (long-term average recharge values in 10⁶ x Ml/day) for all the 11 ensembles (See Figures 8 to 13).
• Plots of empirical cumulative distribution function (ECDF) for seasonal and monthly long-term average total recharge for all the RBMD (No. 2–12). ECDF is a way of producing cumulative distribution function curves by modelling the distribution of measured data. Recharge totals are produced for seasonal summaries (Figure 16): winter (DJF), spring (MAM), summer (JJA) and autumn (SON) as well as for monthly values (Figure 17) for all of the RBMDs covering England and Wales (Nos. 2–12).

Change in seasonal (winter, spring, summer and autumn) recharge for each ensemble

To correspond with the previous work for the Future Flows and Groundwater Level project (see for example: Prudhomme et al., 2012), seasonal averages expressed as mm/d for all groundwater bodies for England and Wales for all 11 ensembles were produced. This enables the results for all 11 ensembles to be presented in a digestible form and compared against each other. Figures 3 to 5 shows the results summarised by meteorological season (winter–December, January and February (DJF); spring–March, April and May (MAM); summer–June, July and August (JJA); autumn–September, October and November (SON)). The change in fraction of recharge for the 2020s, 2050s and 2080s are presented and discussed below.

2020s: (Figure 3) In general there is increasing winter recharge with a subsequent reduction in recharge for spring and summer. For the latter this is less important as there is limited potential recharge occurring between June and August. Importantly there is a mixed signal in spring with some ensembles showing a decrease in recharge and others an increase.

2050s: (Figure 4) There is an increasingly polarised picture compared to the 2020s with winter, for a vast majority of groundwater bodies, showing an increase for each ensemble. Recharge in summer shows a consistent reduction, although not as significant as for the 2020s described above. In spring four out of the 11 ensembles demonstrate increasing recharge which is repeated for autumn.

2080s: (Figure 5) The pattern is similar to the 2050s but with increases in the number of ensembles in autumn which show an increase in recharge (six in total). However, the spatial pattern of the results in spring is mixed, with equal numbers of ensembles showing increases and decreases.

File:OR17026fig3.jpg

Figure 3    Seasonal changes in recharge values as seasonal average (mm/d) for groundwater bodies for each ensemble member (2020s).

File:OR17026fig4.jpg

Figure 4    Seasonal changes in recharge values as seasonal average (mm/d) for groundwater bodies for each ensemble member (2050s).

File:OR17026fig5.jpg

Figure 5    Seasonal changes in recharge values as seasonal average (mm/d) for groundwater bodies for each ensemble member (2080s).

Monthly median percentage change for all 11 ensembles

To investigate the details of which months exhibit the greatest change in monthly average recharge (mm/d) for groundwater bodies, the median of the change of all the 11 ensembles was produced for the 2050s (Figure 6) and 2080s (Figure 7). Note that the 2020s were not included as they are thought to be overly influenced by the climatic variability rather than climate change. The changes are summarised in Table 2 below and demonstrate that for both the 2050s and 2080s recharge increases during winter and for November and decreases during summer. The pattern is much more mixed for both autumn and spring with both seasons exhibiting spatial variability.

Table 2    Summary of seasonal changes for median of the change for each ensemble.

Season	2050s	2080s
Winter	Widespread increases for all winter months confirm the pattern observed in seasonal summaries.	Consolidates patterns observed for 2050s.
Spring	March — spatially variable with central and southern England showing increases, rest decreases. April and May show widespread decreases.	Much more mixed picture (spatially varying increases and decreases).
Summer	Very significant and widespread reductions for all summer months.	Less pronounced change in June and July than 2020s and more spatially variable. August more consistent with 2020s except for parts of east Anglia which show increases.
Autumn	September and October also exhibit significant decreases. Increase is only seen for November.	September mainly decreased but some areas increase. November again has a significant increase.

File:OR17026fig6.jpg

Figure 6    The Median percentage change in monthly recharge of the ensemble members shown for each month (2050s).

File:OR17026fig7.jpg

Figure 7    Median values for percentage change in monthly recharge of all 11 ensemble members (2080s).

Total recharge for River Basin Management Districts

Recharge summaries for the North-west (RBMD No. 12), Humber (RBMD No. 4) and Thames (RBMD No. 6) were used to illustrate the general trends for the impact of climate change on potential recharge in different parts of England and Wales (Figures 8 to 13 and Tables 3 to 5). The plots are of monthly average recharge (as mm/d) for the historical simulation and changes in monthly average recharge (as mm/d) for the 2020s, 2050s and the 2080s. Histograms were provided of minimum, maximum and average recharge totals expressed as Ml (equivalent of annual average recharge).

The general view is that climate signal (however that manifests itself for each RBMD) predominates as the time slices go forward in time (i.e. 2020s to 2050s to 2080s).

• North-west: broad agreement across ensembles showing a decrease in recharge over summer/early autumn which becomes more prevalent from the 2020s to the 2050s and on to the 2080s. This is followed by an increase in winter recharge and a more mixed picture in spring. Overall, the total recharge volume (Table 3 and Figure 9) increases over the 2020s, 2050s and 2080s.
• Humber: generally more subdued response than the North-west and Thames. The ensembles show variable recharge over late winter and early spring recharge, followed by relatively small change predicted for late spring and early summer recharge. Consistent decreases in recharge are confined to August and to a lesser extent September, whilst consistent increases occur in late autumn/early winter. The variability is similar in all three time slices. There is an increase in the average recharge volume totals compared to the Historical Simulation with results from the 2050s showing greater totals than the 2020s and 2080s (Table 4 and Figure 11).
• Thames: Generally increasing within the recharge season, i.e. late autumn and winter. The greatest increase is observed in January and February. Average totals of recharge increase compared to Historic Simulation, but with a corresponding increase in range (minimum value to maximum) — see Table 5 and Figure 13.

Note that all ensembles are equally likely and that whilst the average increases from the 2020s to the 2050s and onto the 2080s there is an equal likelihood that recharge volumes could decrease.

In summary the results for the Humber RBMD shows that the response in the east of the country is more damped. The North-west RBMD sees a reduction in late summer/early autumn which could be interpreted as resulting from changes to the western predominance of weather systems. The recharge response in the Thames RBMD is similar to North-west RBMD with increases in recharge in the current recharge season. However, the response in the Thames RBMD in January and February is more pronounced than North-west RBMD possibly due to higher recharge signals in the west of the catchment and lower in the east of the catchment.

File:OR17026fig8.jpg

Figure 8    Comparison of historic and future ensemble monthly recharge results for North-west RBMD (12).

Table 3    Recharge volumes for Catchment 12: North-West.

	afgcx	afixa	afixc	afixh	afici	afixj	afixk	afixl	afixm	afixo	afixq	Min.	Max.	Average
1961–1990	105.24	102.95	107.38	109.71	107.93	104.77	105.67	109.42	105.76	108.67	106.99	102.95	109.71	106.77
1971–2000	106.12	105.32	108.08	113.47	109.50	105.05	108.34	108.32	109.14	109.39	104.03	104.03	113.47	107.89
20s	110.55	107.53	109.57	117.70	106.13	105.22	105.04	114.72	112.13	105.59	115.18	105.04	117.70	109.94
50s	114.34	111.27	110.51	117.64	111.08	103.00	103.01	110.64	113.48	104.72	107.71	103.00	117.64	109.76
80s	111.53	116.35	111.81	124.74	111.90	107.76	105.14	114.58	112.39	112.53	114.99	105.14	124.74	113.07

Note: Recharge values in 10⁶ x Ml/day

File:OR17026fig9.jpg

Figure 9    Changes to monthly recharge for North-west RBMD (12).

File:OR17026fig10.jpg

Figure 10    Comparison of historic and future ensemble monthly recharge results for Humber RBMD (4).

Table 4    Recharge volumes for Catchment 4: Humber.

	afgcx	afixa	afixc	afixh	afici	afixj	afixk	afixl	afixm	afixo	afixq	Min.	Max.	Average
1961–1990	112.41	108.98	118.00	123.10	115.07	109.50	112.26	121.23	112.65	117.86	112.63	108.98	123.10	114.88
1971–2000	116.05	107.27	119.98	127.61	118.39	116.32	118.13	122.55	109.22	114.74	107.39	107.27	127.61	116.15
20s	122.61	102.30	131.78	129.58	106.31	108.86	114.70	129.74	120.20	107.27	116.49	102.30	131.78	117.26
50s	116.75	106.98	124.18	131.78	118.36	105.00	100.14	121.62	118.44	102.94	115.79	100.14	131.78	114.73
80s	119.71	113.01	134.43	140.09	126.80	107.54	107.76	122.77	124.47	114.74	119.42	107.54	140.09	120.98

Note: Recharge values in 10⁶ x Ml/day

File:OR17026fig11.jpg

Figure 11    Comparison of historic and future ensemble monthly recharge values for North-west RBMD (12).

File:OR17026fig12.jpg

Figure 12    Monthly recharge for Thames RBMD (6).

Table 5    Recharge volumes for Catchment 6: Thames.

	afgcx	afixa	afixc	afixh	afici	afixj	afixk	afixl	afixm	afixo	afixq	Min.	Max.	Average
1961–1990	63.69	62.98	69.27	73.05	66.99	62.78	64.65	69.43	63.67	67.85	64.16	62.78	73.05	66.23
1971–2000	69.05	62.16	68.18	78.81	65.74	66.59	67.59	66.13	60.43	69.67	67.23	60.43	78.81	67.42
20s	67.70	54.82	74.84	80.72	62.00	64.77	64.99	75.64	62.23	71.78	62.96	54.82	80.72	67.49
50s	65.53	61.04	77.21	86.57	68.49	56.28	60.90	65.56	69.08	59.53	68.99	56.28	86.57	67.20
80s	74.65	62.53	82.65	93.39	83.39	63.01	66.50	72.57	64.53	70.82	73.65	62.53	93.39	73.43

Note: Recharge values in 10⁶ x Ml/day

File:OR17026fig13.jpg

Figure 13    Changes to monthly recharge for Thames RBMD (6).

Seasonal and monthly empricial distribution functions

ECDF plots for seasonal long-term recharge totals along with monthly totals for all the RBMD included in this report are presented in Figure 14 and 15 along with median values of long-term total recharge in Table 6. These plots are produced to examine the distribution of the recharge for each year along with how these change from the historic simulation to the future climate change for the 2050s and 2080s. Four 30 year periods are chosen: 1961–90, 1971–2000, 2050s and 2080s to enable direct comparison of the total recharge calculated over the RBMD.

Seasonal: Examining Figure 14 and Table 6 and focussing on the median (50%ile) for each ECDF curve shows a significant increase in winter, small variability during spring and autumn with a significant reduction in summer.

Monthly: Examining Figure 15 the following is highlighted:

• For winter (DJF): Future recharge is greater than the historic simulation.
• For spring (MAM): Similar profiles exist for historic simulation, 2050s and 2080s.
• For summer (JJA): Recharge reduces from historic simulation to future climate (2050s and 2080s).
• For autumn (SON): There is a switch from a reduction in the future in September, neutral in October and an increase in November.

The analysis of seasonal and monthly trends from the historic simulation highlights that summer will become a period of reduced potential recharge. The reduced potential recharge in September (historically the start of the recharge season) suggests that the period of low recharge could be extended by one to two months, thereby shortening the recharge period. This is an important trend to note as prolonged dry weather in a year could have a significant impact on groundwater storage recovery.

File:OR17026fig14.jpg

Figure 14    Empirical cumulative distribution function plots for recharge totals (RBMD No. 2–12) over the winter (DJF), spring (MAM), summer (JJA), and autumn (SON) seasons.

Table 6    Median values (50%tile) of total recharge (RBMD No. 2–12) for historical simulation and 2050s and 2080s (Ml).

Time period	Winter	Spring	Summer	Autumn
1961–90	15689416	6586268	2936518	9039051
1971–00	15790442	6440959	3010257	9147050
2050s	17705422	6012697	1905246	9012378
2080s	18470139	6316415	1649490	9176169

File:OR17026fig15.jpg

Figure 15    ECDF plots for monthly recharge totals (RBMD No. 2–12): historic simulation, 2050s and 2080s.

Summary

Once the climate change signal becomes more dominant, i.e. 2050s and 2080s, the overall picture is one of shorter recharge season with a similar or increased amount of potential recharge. There are, however, regional variations with basins in the west of England and Wales showing greater changes in late autumn/early winter. The reduction in recharge in the ‘shoulder’ of the recharge season means that more recharge occurs in fewer months. Whilst this means that the groundwater balance is maintained and so is ‘good news’ for water resources, it may make the system more vulnerable to drought if one or two months within the recharge season experience lower than average rainfall.