Extreme precipitation has recently caused unexpectedly devastating floods in Switzerland. Such meteorological events cannot be avoided, but their consequences can be mitigated by adequate planning of infrastructural or financial nature. This can be done with the help of statistical analyses of extreme precipitation events.
At present, the standard reference work for return levels of rare events currently available to hydrologists, natural hazard prevention, and engineers in Switzerland («Blauen Bände» or «Röthlisberger-Ordner»), is long out of date, both in terms of its data, which only extends into the 1980s, and of the statistical methods employed.
The goal of the project "Extremniederschläge" - a joint project between MeteoSwiss and BAFU - is to create a web platform for public use presenting statistical analyses of extreme precipitation. The project is scheduled to last from 2013 to 2015. The climatological data base will be updated, and state-of-the-art statistical methods implemented or developed.
1. Climatological data
At our disposal we have precipitation observation recorded at different intervals (10 min, 1 hour, 1 day, as well as multiples of it). Which temporal resolution as well as time period (1864 - today, 1901 - today, 1961 - today ) to use for what analysis will be decided within the project.
The project will make various extreme value analyses at rain gauges available, ranging from simple return level plots to intensity-duration-frequency diagrams. To this end we implement and refine the latest statistical methods. Hence we can give confidence intervals to the estimates, which are not supplied in the "blue books". This is of relevance, since in the context of extreme value statistics confidence intervals tend to be large, especially for long return period.
The goal is to design, in an intensive exchange with the users, a web platform, capable of development.This web platform will replace the "blue books" and can be updated without any problems.
4. Longterm plan
Within the two-year project only elementary needs of the users can be addressed. An important objective consists in drafting a plan for the continuation of the service and the extension of the web platform. Next to customer service and maintenance, new ways of extreme value analysis will be contemplated. Regionalisation of results at any location in Switzerland and development of extreme value statistics for spatial means (e.g. of catchment areas) are matters of particular interest.