Two things to note: firstly, droughts are not currently measurably worse.
Secondly, short-term summer droughts may become more severe. But long-term droughts will become less frequent and less severe.
A sensible response is water management based on catchment areas, as we had for the majority of our history before politicians started messing with it.
Changes in drought frequency, severity and duration for the British Isles projected by the PRUDENCE regional climate models
Blenkinsop, S. & Fowler, H.J.
2007
Journal: Journal of Hydrology
Issue: Vol.342
Pages: 50-71
Using a number of climate models in impacts assessment allows the uncertainty in the projections of change to be examined. Six regional climate models (RCMs) from the PRUDENCE project (Prediction of regional scenarios and uncertainties for defining European climate change risks and effects http://prudence.dmi.dk/) were assessed to i) determine the ability of the model to accurately reproduce the observed climate statistics for the 1961-90 period; ii) examine the range of uncertainty in future changes for the SRES A2 emissions scenario for the 2071-2100 period; and iii) assess the ability of RCMs to reproduce observed drought frequencies. Future drought frequencies are also discussed.
It was found that the RCMs are able to re-produce the observed spatially averaged annual precipitation cycle over the British Isles but may have difficulty capturing important physical processes responsible for precipitation (such as regional variations within the UK). The RCMs were unable to simulate the observed frequency of drought events.
The RCMs indicate that future projections of change in mean precipitation decrease during summer months and increase during winter months. Short-term summer drought is projected to increase in all areas except Scotland and Northern Ireland, although there is large uncertainty associated with such changes. Projected changes in long-term droughts are highly uncertain, although the longest droughts are projected to become much shorter and less severe. Differences between the model simulations are also discussed by the authors and it is noted that the choice of criteria for weighting models to produce probabilistic climate change scenarios is not easy.
These results suggest under future climate change water companies may need to plan for more intense short-term droughts, but may experience fewer longer duration events.
Major droughts in England and Wales, 1800 – 2006
Marsh, T., Cole, G. and Wilby, R
2007
Journal: Weather
Issue: Vol.62, No.4
Pages: 87-93
Marsh et al examined historical rainfall patterns (back to 1800) in the UK to determine whether more recent drought conditions (2004 to 2006) were exceptional. Data from a range of sources was used in the research, including the CET (Central England Temperature series), groundwater level data and historical and documentary evidence of drought impact.
The results suggested that the more recent drought events where by no means exceptional in terms of the longer historical period considered by the research. The record displays many examples of multi-year droughts, with the most extreme examples being the 1798 to 1808 and 1890 to 1910 droughts. The 1890 – 1910 drought displayed a rainfall deficiency greater than that recorded in the 2004 to 2006 drought period.
The authors conclude that society has been able to adapt to these times of water stress in the past, even with an increasing water demand through population increase. However, considering the rise in temperatures, water demand and future expectations of water users, any repeat of the rainfall patterns of the 19th century would present a very significant challenge to the water industry.