Extreme events in total ozone over Arosa – Part 1: Application of extreme value theory

摘要

In this study ideas from extreme value theory are for the first time appliedin the field of stratospheric ozone research, because statistical analysisshowed that previously used concepts assuming a Gaussian distribution (e.g.fixed deviations from mean values) of total ozone data do not adequatelyaddress the structure of the extremes. We show that statistical extremevalue methods are appropriate to identify ozone extremes and to describe thetails of the Arosa (Switzerland) total ozone time series. In order toaccommodate the seasonal cycle in total ozone, a daily moving threshold wasdetermined and used, with tools from extreme value theory, to analyse thefrequency of days with extreme low (termed ELOs) and high (termed EHOs)total ozone at Arosa. The analysis shows that the Generalized ParetoDistribution (GPD) provides an appropriate model for the frequencydistribution of total ozone above or below a mathematically well-definedthreshold, thus providing a statistical description of ELOs and EHOs. Theresults show an increase in ELOs and a decrease in EHOs during the lastdecades. The fitted model represents the tails of the total ozone data setwith high accuracy over the entire range (including absolute monthly minimaand maxima), and enables a precise computation of the frequency distributionof ozone mini-holes (using constant thresholds). Analyzing the tails insteadof a small fraction of days below constant thresholds provides deeperinsight into the time series properties. Fingerprints of dynamical (e.g. ENSO, NAO) and chemical features (e.g. strong polar vortex ozone loss), andmajor volcanic eruptions, can be identified in the observed frequency ofextreme events throughout the time series. Overall the new approach toanalysis of extremes provides more information on time series properties andvariability than previous approaches that use only monthly averages and/ormini-holes and mini-highs.