Dealing with Dirty Data: A Blueprint for Analyzing Climate Variability

Andrew Rhines, Harvard University

The impact of climate change on human populations is disproportionately sensitive to the frequency of extreme events such as heat waves and cold snaps.  As a result, there is considerable interest in determining how the full distribution of temperature has changed over recent decades. Model ensembles project that extratropical land temperature variance will decrease by 2070, and this is consistent with simple physical arguments. However, observational studies have thus far come to conflicting conclusions. I will discuss how several analytical pitfalls in interpreting observational records lead to these discrepancies, and highlight the importance of accounting for non-normality and the effects of filtering, time-averaging, gridding, and smoothing. I will then present a set of methods designed to overcome these challenges, which I apply to a large set of daily temperature observations to show that a decrease in temperature variability is already robustly detectable in the extratropics.

Link to Recording: