Biomathematics & Statistics Scotland
Extreme environmental events, such as flooding, drought and landslides, pose significant risks to human life, ecosystems and economic activity. It is predicted that climate change will lead to an increase in the magnitude and frequency of extreme events. Hence there is much interest in establishing whether such changes have already occurred and quantifying the changes that are likely to occur in future.
Extreme value theory (EVT) provides a rigorous mathematical basis for using data on moderately rare events to make statements about the probabilities of very rare events which have at best occurred infrequently in the past. EVT tells us, for example, that the generalised Pareto distribution can be used to model the distribution of values that exceed a high threshold. Our work focuses on using EVT to model changes over time in the magnitude and frequency of extreme events, involving the development of new methodology and its application to data on river flows and storm surge heights.
Daily spring discharge data
for the River Dee at Cairnton
(Aberdeenshire), together with
estimated 100 year return levels
(solid blue line) and associated 95%
confidence intervals (dashed blue
lines) that result from fitting an
extreme value model with linear
time trend to data above a threshold
of 200m3/s (red line).
Further details from: Adam Butler
