Record models for temperature and rainfall

A record is an entry in a time series that is greater or smaller than all previous entries. The public perception of weather extremes is particularly sensitive to record-breaking events, which often receive extensive media coverage. As a consequence, a number of recent studies have been aimed at detecting and quantifying the
effects of a warming climate on the frequency of temperature records.

In comparison, the effects of climatic trends on precipitation records are more complex and have generally received less attention. After a brief introduction to the classic theory of records, the talk will describe two simple models describing the occurrence statistics of record-breaking daily temperatures and precipitation amounts, respectively. In the latter case our main finding is that the presence of dry days of zero precipitation induces nontrivial correlations between record-breaking events that can be detected in empirical data.