Statistical evaluation of the correlation pattern between rising global temperature and stable water isotopes in precipitation.

Statistical evaluation of the correlation pattern between rising global temperature and stable water isotopes in precipitation.

Lukas Ditzel, Jonas Schramm, Matthias Gassmann

Department of Hydrology and Substance Balance, University of Kassel, Kurt-Wolters-Strasse 3, 34125 Kassel, Germany

 

Stable water isotopes in precipitation on the northern hemisphere are usually following a predictable pattern throughout the year, with high amounts in summer and low amounts of deuterium and ¹⁸O in the winter season. Backed by a richness of available date from the International Atomic Energy Agency (IAEA), one can mostly expect an annual sinusoidal form of isotope data, when looking at data for a certain region in the northern hemisphere.

Since the driving factor for isotopic enrichment or depletion is isotopic fractionation, the seasonal behavior is strongly correlated to air-temperature. The correlation between temperature and fractionation is strong enough to explain most of the greater deviations from the sinusoidal form like in arid regions. It occurs that globally rising temperatures, initiated by climate change, should have an impact on the sinusoidal form of the stable water isotope time series. We assumed, that rising temperatures will lead to higher contents of deuterium and ¹⁸O in the precipitation of the northern hemisphere. Due to the availability of data and the long time series, which are needed for robust answers, we focused our work on European and North-American data. First analyses showed a positive correlation between rising air-temperatures and isotopic content, but not all regions. Other effects like the elevation- and continental-effect were dampening the effect of rising global temperatures, especially in coastal regions or islands such as Ireland. More continental regions, however, are showing a rise for isotopic enrichment in precipitation. We analyzed this trend by the calculation of the trend-components of these time-series via Loess and validated them by using the Mann-Kendall-Test. Furthermore, we separated sets of data into monthly clusters and looked for rising temperature trends in every month over the size of the available time series. This second analysis was performed for the time series from weather stations in Berlin, Vienna and Krakow covering almost 40 years of monthly isotope data.