A regional frequency analysis of United Kingdom extreme rainfall from 1961 to 2000

Abstract

Multi-day rainfall events are an important cause of recent severe flooding in the UK, and any change in the magnitude of such events may have severe impacts upon urban structures such as dams, urban drainage systems and flood defences and cause failures to occur. Regional pooling of 1-, 2-, 5-and 10-day annual maxima for 1961 to 2000 from 204 sites across the UK is used in a standard regional frequency analysis to produce generalized extreme value growth curves for long return-period rainfall events for each of nine defined climatological regions. Temporal changes in 1-, 2-, 5-and 10-day annual maxima are examined with L-moments using both a 10 year moving window and the fixed decades of 1961-70, 1971-80, 1981-90 and 1991-2000. A bootstrap technique is then used to assess uncertainty in the fitted decadal growth curves and to identify significant trends in both distribution parameters and quantile estimates.There has been a two-part change in extreme rainfall event occurrence across the UK from 1961 to 2000. Little change is observed at 1 and 2 days duration, but significant decadal-level changes are seen in 5-and 10-day events in many regions. In the south of the UK, growth curves have flattened and 5-and 10-day annual maxima have decreased during the 1990s. However, in the north, the 10-day growth curve has steepened and annual maxima have risen during the 1990s. This is particularly evident in Scotland. The 50 year event in Scotland during 1961-90 has become an 8-year, 11-year and 25-year event in the East, South and North Scotland pooling regions respectively during the 1990s. In northern England the average recurrence interval has also halved. This may have severe implications for design and planning practices in flood control.

ENSO‐conditioned rainfall drought frequency analysis in northwest Baja California, Mexico

“…In addition, the L-moment regional frequency analysis technique has been widely used (e.g. Fowler and Kilsby, 2003;Wallis et al, 2007;Hallack-Alegria and Watkins, 2007) and 841 may be particularly valuable in other regions with short precipitation records. The results of this study may also have implications for other areas where rainfall variability has been linked to Pacific Ocean SSTs (e.g.…”

“…Easterling et al, 2000;Zhu and Lettenmaier 2007;Higgins et al, 2007) have noted that demonstrating ENSO's effect on rainfall variability in the region has been hampered by a lack of accurate and complete longterm precipitation records, a limitation that is partially overcome in this study through a regional frequency approach (e.g. Fowler and Kilsby, 2003;Trefry et al, 2005, Wallis et al, 2007.…”

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

“…There have, however, been relatively few attempts to establish whether this volatility reflects a trend, whether in mean values or in changing variance, that can be an equally important component of regime change. Most of the recent research in the UK has focused on extreme rainfall (Osborn et al, 2000;Fowler and Kilsby, 2003) and flooding (Robson et al, 1998;Robson 2002). The overall picture emerging from this work is that the evidence for climatic effects on flood flows is inconclusive (Robson et al, 1998;CEH and UKMO, 2001;Robson, 2002).…”

Shallow landslide initiation on terraced slopes: inferences from a physically based approach

“…cumulative rainfall at different hourly intervals). Specifically, we used the generalized extreme value (GEV) distribution (Jenkinson 1955), which is widely used in extreme event frequency analysis rather than the Gumbel distribution, as the literature increasingly suggests that the distribution of extreme events may be more heavily tailed (Fowler and Kilsby 2003). The cumulative distribution function of the GEV distribution is as follows:…”