An assessment of trends in UK runoff and low flows using a network of undisturbed catchments

Abstract

In many parts of the world attempts to discern climatic-driven changes in river flow patterns are hindered by the combined impact of other, more direct anthropogenic disturbances such as abstraction and impoundments. This study capitalises on a newly defined 'benchmark' network of natural catchments in the UK to discern natural variability in flow regimes. Trend tests were applied to time series of runoff and indicators of low-flow magnitude and duration for two study periods to assess the sensitivity of trends to the frame of reference over which tests were conducted. Notwithstanding the volatility of the recent past, the results demonstrate a general stability in runoff and low flows since the early 1960s. The strongest signal to emerge from the study is a positive runoff trend for catchments in Scotland, which was resilient to the effect of the study periods. There was also some indication of increasing runoff in maritime western areas of England and Wales. These increases in maritime areas are likely to reflect the dominance of westerly airflows in the recent past, associated with an increase in the North Atlantic Oscillation (NAO) Index. For low flows, there were no compelling trends -significant positive trends over the 1973-2002 period are influenced by a sequence of notably dry years at the start of the period and were not observed over a 40-year time-frame. There are some indications of a tendency towards decreasing low flows in some eastern catchments, but this is supported by few significant results. The results of this study provide a baseline against which to assess longer-term change from historical flow time series and to monitor future change in the benchmark network.

The British river of the future: How climate change and human activity might affect two contrasting river ecosystems in England

“…• Winter river water temperatures could rise by 1.5-3 o C from the current typical 5 o C. characterised by considerable year-on-year variability (Hannaford and Marsh, 2006). These 199 hydrological scenarios have not specifically examined the risk of particular headwater 200 streams drying out in summer.…”

Spatial cross‐correlation patterns of European low, mean and high flows

“…Hannaford and Marsh, 2006;Mudelsee et al, 2003;Kingston et al, 2011]. Most of these studies analysed trends of miscellaneous annual summary statistics such as mean annual runoff, seasonal trends and trends in flood and drought characteristics.…”

“…Trends in annual low flow [e.g. Hannaford and Marsh, 2006;Hisdal et al, 2001;Stahl et al, 2010] and high flow [e.g. Kundzewicz et al, 2005;Hannaford and Marsh, 2008] statistics often follow the patterns of the corresponding seasonal trends.…”

Rural land management impacts on catchment scale flood risk

“…Huh et al, (2005) found that to be able to detect step changes in flooding frequency, record lengths needed to be at least 40 years for flood frequency to be assessed and 60 years for low flow trends to be tested. A benchmark network of gauging stations collated by Hannaford and Marsh (2006) had an average length of 33.7 years. However, the record length needed to accurately assess flood frequency depends both upon whether flood events are clustered and the temporal resolution of the data used.…”

“…It was also stated that a common period of data record is best, so that comparisons between stations can be made. Hannaford and Marsh (2006) argued that using a fixed period for relatively short records is problematic, as trend analysis is sensitive to the chosen period. This is illustrated by a study by Hisdal et al (2001), where for a single station, using different 30 year periods, both positive and negative significant trends were found.…”