Clustering of Station Observations for Extreme Value Analysis

The analysis of precipitation extremes with decadal return periods requires long data records; longer than is typically available from station observations. To address this problem, a method is proposed for clustering station records based on the similarity of their precipitation climatologies. It is demonstrated that the proposed aggregation method is superior to naive aggregation methods for Extreme Value Analysis and enables the detection of changes in the historical record that would otherwise not be distinguishable from noise.