This section places the hydrological conditions experienced in 2005 in the context of the recent past, and within a broader historical perspective.
2005 continued a sequence of years, beginning in the late-1980s, which taken together exhibit a greater degree of hydrological volatility than is captured in the previous records for most river flow and groundwater level monitoring stations. On the basis of climatological data and more limited hydrometric evidence, comparably volatile episodes with extended periods of depressed runoff and recharge rates can be identified in the historical record (e.g. the 1850s, 1890-1910, and the 1940s). There are, however, no close modern parallels to the very notable departures from the normal seasonal patterns in river flows and rates of aquifer recharge experienced over the last 15-20 years. An enduring feature of hydrological conditions across the UK is the large contrasts often experienced between northern and southern Britain - these differences have achieved an extreme expression in a number of recent years.
Following extended drought conditions in 1988-92, which were punctuated by the exceptionally wet winter of 1989/90, a wet interlude heralded a second protracted drought (1995-97) which impacted most severely on southern Britain - groundwater resources in particular. The drought terminated in the autumn of 1997 to be followed by the wettest five- year sequence on record for the UK. England & Wales registered its highest 5-year rainfall total in a series from 17661. Severe flooding occurred in April 1998 (across the Midlands), throughout most of southern Britain in 2000/01, and again in early 2003. Existing maximum recorded flows were widely eclipsed - mostly by modest margins - and groundwater levels, responding to unprecedented rates of aquifer recharge (especially in the winter of 2000/01), exceeded previous maxima for extended periods in many southern outcrop areas. With most rain-bearing frontal systems following more southerly tracks than normal a further drought episode developed in northern Britain and Northern Ireland during 2001. More intense drought conditions extended across much of the UK during the exceptionally hot spring and summer of 2003. Sustained rainfall during the late autumn had a moderating effect but very limited replenishment of groundwater resources in the late winter and early spring of 2004 heralded the recent winter drought episodes which impacted most severely on parts of eastern, central and southern England in 2005 (and 2006).
Considering the last 15 years as a whole rainfall for Scotland has been around 9% higher ( 4% higher for Northern Ireland) than the preceding average with winters contributing most to the additional rainfall. Notwithstanding the recent drought periods, rainfall for England & Wales over the 1991-2005 period is very close to the preceding average. Figure 1 allows the rainfall during the recent past to be compared with rainfall variability since 1914. The outstanding wetness of the 1998-2002 period in southern Britain, together with the notable dryness of much of the 1970s, has served to exaggerate the apparent increasing trend in rainfall over the last 30- 40 years. This is particularly true of Scotland where the wetness of the last 20 years has seen an extension in the range of recorded variation in annual rainfall totals. A modest contributory factor to the increase in rainfall may be the decreasing proportion of snowfall in the overall precipitation totals (the underestimation of snowfall is normally greater than that for rainfall). The erratic increase in rainfall since the mid-1970s was not maintained over the 2003-05 period, underlining the dangers of extrapolation of time series characterized by protracted perturbations about a relatively stable long term mean.
A more extended historical perspective is provided by the 232-year England & Wales rainfall series1. It exhibits no overall trend but long term tendencies to increasing winter and decreasing summer rainfall may be identified. The last 35-years in particular has seen a clearer partitioning between winter half-year (November-April) and summer half-year (May-October) rainfall across much of the UK. For most of the 20th century England & Wales, winter half- year (November-April) rainfall exceeded those for the summer half-year. By contrast, prior to the 1st World War summers were very often wetter than winters and clusters of dry winters were common, in the 1850s and 1890-1905 period particularly2. Any repetition of similar rainfall patterns across southern Britain would represent a considerable water resource challenge given modern levels of water demand.
A compelling climatological feature of the recent past has been the persistence of seasonally high temperatures, especially over the last 10 years. 1996 is the only year since 1988 that the annual mean Central England Temperature (CET)3 has failed to exceed the average for the preceding record. Nine of the ten warmest years in the 337-year CET series cluster in the post- 1988 period and the average CET for the last ten years is around 1.0°C greater than for a century ago. The recent very appreciable warming is evident in Figure 2 which illustrates winter (December-February) and summer (June-August) temperature and rainfall anomalies (relative to the 1845-1974 average) for England & Wales. The red diamonds show the plotting positions for the most recent 30-years; a high proportion, including the winter and summer half-years for 2004/05, have registered positive temperature anomalies. The last 30 years is notable for the above average frequency of mild, wet winters and hot, dry summers. However, the dryness of the 2004/05 winter and the limited summer rainfall deficiency underline the inherent variability of the British climate.
The warm conditions have contributed to significant increases in evaporative demands over the last 40 years. Figure 3 shows the 5-year running mean annual potential (PE) and actual (AE) evaporation losses for England & Wales and for Scotland (based on MORECS4 data. PE losses for England & Wales have increased erratically but recent annual totals have been around 30 mm greater than those that characterized the 1960s. AE losses exhibit a broadly similar trend although in the drier regions of England little overall increase can be detected. This is largely due the warmer (and, often, drier) summers with correspondingly very dry soil conditions - inhibiting transpiration rates and, thereby, moderating overall AE losses. As a consequence, changing evaporative demands have had only a limited impact on the water balances of most catchments. In Scotland the generally higher rainfall implies that transpiration rates are normally restricted for only very short periods away from sheltered areas along the eastern seaboard. Correspondingly, countrywide annual AE losses closely approach PE totals in most years. Actual evaporation losses have increased by a similar margin to those for England & Wales since the 1960s but the impact on water balances in most catchments has been more than counterbalanced by increases in annual precipitation totals over the same timespan.
River flows, reservoirs stocks and groundwater resources are sustained and replenished not by rainfall directly but by that proportion which remains after allowing for evaporative demands. Runoff therefore provides the best index of the health of water resources. Figure 3 provides evidence of substantially greater runoff from Scotland over the post-1980 period with a notable range of annual variation over the last decade and a downturn over the 2003-5 period. The latter is more prominent in Northern Ireland where the 2003-05 runoff is considerably lower than that for any three-year accumulation in a series from 1981. For England & Wales, notwithstanding the enhanced evaporative demands, total runoff over the five years beginning in 1998 was unprecedented over the post-1960 period (and, very probably, throughout the instrumented era - which stretches back into the 18th century). Subsequently, the dry winters have contributed to much more modest overall runoff and aquifer recharge, especially in southern Britain where an exceptional contrast may be recognized between groundwater resources in early 2003 and at the end of 2005 when depressed levels in parts of the Chalk and Permo-Triassic sandstones (of the Midlands) resulted in very low flows in many spring-fed streams and rivers.
Water resources and environmental stress, as well as impacts on the community, are influenced more by the frequency and magnitude of extreme events than changes in mean runoff or average groundwater levels. Over the last 40 years, statistically significant increases in runoff and flood magnitude have been identified over various timespans for some rivers (mostly in northern Britain) and examples of both positive and negative trends in low flows can readily be found3. However, given the natural range of hydrological variability and the pervasive impact of artificial influences (e.g. abstractions, river regulation, inter-basin transfers and flood alleviation measures) on extreme flows, any attribution to climatic change can be only tentative. Of particular significance in the UK is the fact that the great majority of river flow and groundwater level data has been collected over the post-1960 period. In relative terms, the early part of this period (up to the late-1970s) can be broadly characterised as quiescent - with below average runoff and a low frequency of major flood events across much of the country. By contrast, the recent past has been much more volatile. Correspondingly, trend analyses embracing datasets covering the last 30-40 years will inevitably detect significant trends. In assessing their implications it is important to recognize how sensitive trend recognition is to the timeframe over which the analysis is undertaken; even modest changes in the timeframe used can produce considerable differences in both the sign and magnitude of any apparent trend . When analyses are confined to the limited number of flow records which extend over 80 or more years, few compelling examples of long term trends can be discerned5,6.
1. Wigley, T. M., Lough, T. M. & Jones, P. D. (1984). Spatial patterns of precipitation in England and Wales and a revised homogeneous England and Wales precipitation series. Int J. Climatol., 1-27
2. Marsh, T. J., Cole, G. A. and Wilby, R. L. Major droughts in England and Wales, 1800- 2006, Weather (In press)
3. Manley, G. (1974) Central England Temperatures: monthly means 1659 to 1973. Quart. J. Roy. Met. Soc., 100, 389-405
4. Hough, M., Palmer, S., Weir, A., Lee, M. and Barrie, I. 1995. The Meteorological Office Rainfall and Evaporation Calculation System: MORECS Version 2.0. An update to Hydrological Memo. No. 45, Met Office
5. Hannaford, J. and Marsh, T.J. 2006. An assessment of trends in UK runoff and low flows using a network of undisturbed catchments. Int. Jour. Clim. 26, 1237-1253
6. Hannaford, J. and Marsh, T.J. 2006. High and Low Flow trends in a national network of undisturbed indicator catchments in the UK. In: Climate Variability and Change - Hydrological Impacts. IAHS Publication 308, 496-501