Christoph Frei and his coauthors at the Swiss Federal Institute of Technology Zurich (ETH) and at the University of Reading in the United Kingdom used a unique set of regional climate model simulations and statistical analysis tools from a pair of European Union projects--PRUDENCE and STARDEX--in six regional climate models to quantify the changes in exceptionally strong precipitation events over the next 100 years.
Their analysis shows that Alpine regions and northern European locations above 45 degrees latitude (including such major cities as London, Berlin, and Stockholm) are likely to experience increases in the frequency and strength of fall, winter and springtime extreme precipitation events by the year 2100. They report, for example, that in Scandinavia, unusually strong events that are now expected to occur once per century will occur at approximately 20-40 year intervals.
Global circulation and disconnected regional models had previously forecast increases in extreme precipitation, as higher atmospheric greenhouse gas concentrations are thought to heighten the frequency of such events worldwide. The combination of regional European models used in the current study adds to the detail available to researchers and provides improved estimates for the pattern, magnitude, and uncertainty of precipitation changes, as compared with larger, more general models.
"There are several implications for climate change research from this study," notes Frei, now at the Swiss Federal Office of Meteorology and Climatology (MeteoSwiss). "First, it confirms the prospects of regional climate models as tools for deriving future scenarios of climate extremes. This has great impact on the design of future ensemble climate modeling projects."
The authors confirmed their results using the extensive rain gauge network already in place in the European Alps. They say that although the Alps cover only a limited part of the model's domain, and its results cannot be extrapolated to other regions, their model analysis showed exceptional accuracy when compared with observational data at fine spatial scales that are not resolved in current global models.
By combining a number of different models, the researchers were able to use the varying techniques employed in regional climate modeling. They note that their study does not account for all sources of uncertainty, and should be interpreted as a possible scenario of future extreme precipitation events, but one with higher reliability than was previously possible.
The research was sponsored by the European Union, the Swiss Ministry for Education and Science, and the Swiss National Science Foundation.