The Statistics of Climate Change

Statisticians have been involved in climate change research in numerous ways, from the assessment of trends in climatic time series through to the processing of large spatial-temporal datasets both from observational networks and from climate models. In this talk, I shall give an overview of some of the different ways that statistical analysis has entered climate science, illustrated by some specific examples from my own research. The examples are likely to cover: (a) methods for detecting trends in correlated time series, (b) how to determine whether observed trends are due to human causes (detection and attribution), (c) paleoclimatology and the "hockey stick", (d) the problem of single-event attribution: how a single extreme event, such as Hurricane Katrina or the European heatwave of 2003, can be attributed to human as opposed to natural causes.