By Richard Dallison
Water service providers in the UK face a vast array of challenges when it comes to planning their future operations and services; the impact of climate change on the amount of water available for supply to consumers is key among these. Changes in seasonal and annual average river flows are driven by average precipitation levels, while the extreme river flow events (both high and low flows) are effected by the frequency, duration and intensity of single precipitation events. Potential alterations in both averages and extreme flows are important to understand for water supply in order to be able to maintain a continued consistent supply of clean drinking water in the future.
Using the Soil and Water Assessment Tool (SWAT) hydrological model, we have projected and analysed future river flows from 2021 to 2079 under a worst-case scenario of future emissions, that being representative concentration pathway 8.5, as laid out by the Intergovernmental Panel on Climate Change. To account for uncertainty in future modelling, an ensemble of 12 regionally downscaled models derived from the Met Office Hadley Centre Global Environmental Model (HadGEM3), and supplied by UKCP18, have been used as the future climate inputs to SWAT. Five catchments have been studied, the Clwyd and Conwy in north Wales; the Dyfi in mid-Wales; and the Teifi and Tywi in south Wales; these systems represent a variety of catchment characteristics in terms of land use, soil types, underlying geology and topography.
Study areas and catchments showing streams larger than third order as defined by Strahler method.
Both annual/seasonal average, as well as extreme events, have been analysed for the full 59-year dataset and for all 12 model outputs, the results presented below are based on Mann-Kendall trend analysis of the average of these 12 models for each catchment. In terms of seasonal averages, in all catchments a statistically significant increase in average spring flows is observed across the timeframe, this is matched with a corresponding decrease in autumn flows, significant in all catchments except the Dyfi. Winter and summer flows are only significantly effected in the 3 most northern catchments, with winter flows increasing in the Conwy and Dyfi, but decreasing in the summer; the reverse trends are observed for the Clwyd. All catchments except the Dyfi display a downward trend in annual average river flows, albeit this trend is only statistically significant in the Clwyd catchment. Conversely, annual average flows in the Dyfi show a statistically significant increase through the study period.
When looking at extreme events we have studied four factors, and all have been analysed seasonally and annually: one-day maximum flow volume; one-day minimum flow volume; the number of days where flows are above the 95th percentile value for the full 59 year seasonal or annual dataset; and the number of days that are below the 5th percentile value for the full 59-year dataset. As with the average flows, the changes that are most consistent across all catchments are seen in autumn and spring. In particular, autumn flows are decreasing in volume on average over the study period, with trends towards a greater number of events below the 5th percentile, fewer events above the 95th percentile, and lower one-day minimum and maximum flow volumes. Spring extreme events correspond with the seasonal average results discussed above, with a significant increase in the number of days with a greater flow volume than the average value for the whole dataset through the study period.
Visual representation of the direction and significance of trends observed in seasonal and annual river flow volumes during the period of the hydrological years 2021 to 2079, as detected by Mann Kendall trend analysis.
Given that in four of the five catchments we see decreases in annual average river flows, as well as summer and autumn flows in particular, this could pose a problem for water supply in the future, especially when it is considered that this is usually the period of peak water demand. Increases in the average volume of spring flows in particular, as well as winter flows to a lesser extent, will help to remediate some scarcity, but this is only beneficial if there is sufficient capacity to store additional water in these periods. Further pressure could also be placed on water service providers if projected increases in the number of arid days of the period come to fruition. More frequent shallow flows could affect abstraction license conditions, meaning that water companies would be unable to abstract as much, or even any water from certain abstraction locations, as there would not be sufficient ‘hands off flow’ remaining. A further concern during the projected increased number of very large discharge events (in spring in particular) is changing incoming water quality. Large discharge events could put increased pressure on drinking water treatment plants due to increased turbidity and sediment loads in particular, as well as other pollutants washed from the land in such events. It is clear that the projected changes in the river flows could have a large impact on a variety of aspects of drinking water supply. Work is needed now to plan and mitigate against these potential changes in order to maintain continued supplies into the future.