Wilson Chan – wilson.chan@pgr.reading.ac.uk  

Hydrological droughts are periods of below normal river flows. These events negatively impact public water supply and the natural environment. The UK is commonly perceived as wet and rainy with low risk of water supply shortages. A recent report explored the “Great British Rain Paradox” by showing that this perception does not hold true given past severe droughts and vulnerability to future droughts under climate change. The latest UKCP18 projections suggest the potential for more frequent and intense droughts across the UK.  

“Top-down” and “bottom-up” approaches 

There has been a lot of research carried out on the possible impacts of climate change on river flows in the UK. In our recently published review paper, we reviewed over 100 papers published over the past three decades and found that there is relative certainty among studies over a possible reduction in summer river flows for catchments across the UK. There is also evidence to suggest that slow-responding groundwater-dominated catchments in the southeast, particularly important for public water supplies, may experience a reduction in river flows across all seasons.  

There remains considerable uncertainty over the magnitude of change and the temporal evolution of future droughts. In our review, we find that studies following a traditional “top-down” assessment approach may not be able to fully address key research gaps. Most of the papers we reviewed followed this approach where output from global climate models (GCMs) are fed through hydrological models of varying complexities to simulate river flows (Figure 1a). This approach often aims to analyze as many components within the impact modelling chain as possible and incurs the cascade of uncertainty (Figure 1b). Outcomes depend on the many choices made along the way (e.g. climate models, emission scenarios, hydrological models etc.) which often results in wide uncertainty ranges that are not conducive to decision-making. A large part of this uncertainty is due to differences in the atmospheric circulation response to climate change across different climate models. Studies following a “top-down” approach are therefore limited when considering plausible worst-cases (low likelihood, high impact outcomes) and the information produced often cannot be easily used in practical water resources planning. 

We also identified several approaches that have been developed to address drawbacks of “top-down” approaches. They do not seek to replace the traditional “top-down” approaches but instead aim to explore “top-down” projections from a wider “bottom-up” framework. For example, the scenario-neutral approach does not rely on GCM simulations and explores the sensitivity of hydrological systems to a much wider range of plausible futures. The storyline approach is another example of approaches designed to explicitly understand plausible worst cases and navigate the uncertainty cascade from a decision-making context. Storylines can be seen as plausible pathways conditioned on a discrete set of changes (e.g. in atmospheric circulation, management measures or event characteristics). They are informed by multiple lines of evidence (incl. process understanding, historical reconstructions and traditional GCM projections).  

Storylines 

The second paper of my PhD, published in Hydrology and Earth Systems Sciences, demonstrates how the storyline approach can be applied to understand UK droughts. We used an observed event, the 2010-12 drought, as the basis for developing a range of storylines. The drought is one of the top 10 most significant multi-year UK droughts. Temporary water use restrictions affected 20 million customers and drought conditions led to agricultural and industrial losses of over GBP400 million. The drought was characterized by two consecutive dry winters and terminated rapidly in early 2012 with record-breaking rainfall over spring 2012. Motivated by a series of “what-if” questions, we created downward counterfactual storylines of the 2010-12 drought to reimagine how the event could have turned out worse. 

 In our study, we created storylines quantifying what would happen if… 

1. Hydrological preconditions of the drought were drier 

2. Continued dry conditions persisted from a third dry winter instead of the observed rapid drought termination  

3. The drought was to unfold in a warmer climate.  

We showed that the 2010-12 drought was highly influenced by catchment preconditions. Storylines of drier preconditions showed that catchment preconditions prior to drought inception aggravated drought conditions for some of the most affected catchments. Progressively drier preconditions could have led to short but more intense conditions for fast responding catchments in Scotland and a lag and lengthening of drought conditions in slow responding catchments in lowland England.  

The observed 2010-12 drought was characterized by two consecutive dry winters. Weather forecasts and water companies at the time widely anticipated dry conditions to continue through 2012. The prospect of three consecutive dry winters is a well-known concern in the water resources industry and can lead to significant reduction in reservoir storage. This is especially important for slow-responding catchments as groundwater reserves are normally recharged during winter. Storylines of the 2010-12 drought given an additional dry year with dry winter conditions either before or after the observed drought showed the vulnerability of catchments to a “three dry winters” situation. Figure 2 shows that drought conditions could still have intensified with even lower river flows for catchments that were already the most affected.  

Applying the UKCP18 regional climate projections to the observed 2010-12 drought sequence, drought conditions are projected to worsen with temperature rise. Notably, the magnitude of change is lower for catchments in western Scotland due to the compensating effects of wetter winters in general although summer months are projected to become drier with temperature rise. Benchmark severe droughts such as the 1975-76 and the 1989-92 droughts are regularly used to test the feasibility of water management plans. Given a third dry winter or a >2°C temperature rise, the different counterfactual storylines of the 2010-12 drought could have led to worse conditions than both the selected benchmark droughts (Figure 3 for slow-responding catchments in southeast England relative to the 1989-92 drought).  

Event storylines created from plausible alterations made to past observed droughts can help water resources planners stress test hydrological systems against unrealised droughts. “Bottom-up” approaches exploring specific conditions relevant to water resources planning (e.g. three dry winters) can complement traditional “top-down” projections to better understand worst-cases and consider how future extreme droughts can unfold.