Detection of inhomogeneities in daily climate records to study trends in extreme weather (daily stew project)
Zusammenfassung der Projektergebnisse
Long instrumental climate records are usually affected by non-climatic factors, which complicate assessments of trends in extreme and mean weather. These are especially important for daily data, where our abilities to homogenize the distribution are very limited and the requirements for studies on changes in extremes and weather variability are very high. The limited information we have from parallel measurements and our physical understanding makes a good case that non-climatic changes in the tails of the distribution are more affected by non-climatic changes than the mean of the distribution. For example, temperature measurements at Kremsmünster, Austria, on a north-facing wall show a mean bias of 2 °C in June, with a bias of 5 °C in the 99th percentile due to insolation at dawn. It is custom to detect breaks in daily data by investigating changes in the mean only. Breaks in the tail of the distribution that do not change the mean significantly thus go undetected and uncorrected. Given that the breaks in the tails are likely strong this is a major problem for studies on climatic changes in extreme weather using daily station data. In this project we wanted to start with the detection of breaks in the tails of the distribution, both because of their importance for changes in extremes and in the hope of being able to find more subtle break that would affect the mean as well. This turned out not to work, even if breaks in the tails are strong, the noise in such difference time series is even stronger. In total the signal to noise ratio is mostly smaller and we noticed that our state-of-the-art homogenization methods have problems in case of small signal to noise ratios. We hardly found any correlation between the breaks in the mean and in the tails. If this is not due the methodological problem we found with current break detection methods, this makes the current praxis of only considering breaks in the mean even more worrisome. In our three papers on the multiple breakpoint problem we now understand why low signal to noise ratios are a problem for multiple breakpoint detection methods. The problem of break detection in case of multiple breakpoints is an important general result that is also valuable for other sciences that use historical inhomogeneous data. The next step would be to develop a new statistical test that takes the break variance into account. This would allow us to revisit the problem of breaks in the tails of the distribution. To confidently distinguish climatic from non-climatic changes we may well have to study parallel measurements with current and historical measurement set-ups. The research on daily homogenization would be much facilitated by a global reference database with parallel measurements. If the problems with daily homogenization methods cannot be resolved, such datasets can provide us with the necessary information to separate climatic and non-climatic changes. Setting up a climate reference network at pristine locations that are not likely to change in the next century should be a high priority to provide accurate data on the unique experiment we are performing with the Earth’s climate.
Projektbezogene Publikationen (Auswahl)
-
2011: What is the correct number of breakpoints hidden in a climate record? Final Meeting of Cost Action HOME, October 24 2011, Budapest, Hungary
Lindau, Ralf and Victor Venema
-
2013: Efficiencies of homogenisation methods: our present knowledge and its limitation. Proceedings of the Seventh seminar for homogenization and quality control in climatological databases, Budapest, Hungary, 24 – 28 October 2011, WMO report, Climate data and monitoring, WCDMP-No. 78, pp. 11-24
Domonkos, Peter, Victor Venema, Olivier Mestre
-
2012: Homogenisation on monthly and daily temperature and precipitation data. 9. Deutsche Klimatagung, Albert-Ludwigs-Universität Freiburg, 9.-12. Oktober 2012
Venema, Victor, Ralf Lindau, Elke Rustemeier, Alice Kapala, Clemens Simmer
-
2012: Separation of true from spurious breaks. 12th Annual Meeting of the European Meteorological, September 13 2012, Lodz, Poland
Lindau, Ralf and Victor Venema
-
2013: Break position errors in climate records. 12th International Meeting on Statistical Climatology, June 24 2013, Jeju, Korea
Lindau, Ralf and Victor Venema
-
2013: Detection and correction of breakpoints in long-term German precipitation series. Proceedings of the Seventh seminar for homogenization and quality control in climatological databases, Budapest, Hungary, 24 – 28 October 2011, WMO report, Climate data and monitoring, WCDMP-No. 78, pp. 54-63
Rustemeier, Elke, Alice Kapala, Victor Venema and Clemens Simmer
-
2013: Homogenisation of monthly and daily temperature and precipitation data. Meteorologisches Kolloquium Freie Universität Berlin, 14 January
Venema, Victor
-
2013: On the multiple breakpoint problem and the number of significant breaks in homogenization of climate records, Idöjaras - Quarterly Journal of the Hungarian Meteorological Service, 117, No. 1, 1-34
Lindau, Ralf and Victor Venema
-
2013: Parallel measurements to study inhomogeneities in daily data. 12th International Meeting on Statistical Climatology, IMSC2013, Jeju, South Korea, 24-28 June 2013
Venema, Victor, Enric Aguilar, Renate Auchmann, Ingeborg Auer, Theo Brandsma, Barbara Chimani, Alba Gilabert, Olivier Mestre, Andrea Toreti, and Gregor Vertacnik
-
2013: Parallel measurements to study inhomogeneities in daily data. Data management workshop, DMW2013, San Lorenzo del Escorial, Spain, 6 – 8 November 2013
Venema, Victor, Enric Aguilar, Renate Auchmann, Ingeborg Auer, Theo Brandsma, Barbara Chimani, Alba Gilabert, Olivier Mestre, Andrea Toreti, and Gregor Vertacnik
-
2013: Parallel measurements to study inhomogeneities in daily data. European Geosciences Union, General Assembly 2013, Vienna, Austria, 07 – 12 April 2013
Venema, Victor, Enric Aguilar, Renate Auchmann, Ingebor Auer, Theo Brandsma, Barbara Chimani, Alba Gilabert, Olivier Mestre, and Andrea Toreti
-
2014: Which homogenisation method is appropriate for daily time series of relative humidity? European Geosciences Union, General Assembly, Vienna, Austria, 27 April – 02 May 2014
Chimani, Barbara, Johanna Nemec, Barbara Chimani, Ingeborg Auer, Victor Venema
-
The uncertainty of break positions detected by homogenization algorithms in climate records. International Journal of Climatology, Vol 36 Issue 2, February 2016, Pages 576-589. First published: 13 May 2015
Lindau, Ralf and Victor Venema
-
The joint influence of break and noise variance on the break detection capability in time series homogenization. Adv. Stat. Clim. Meteorol. Oceanogr. 4, 1-18, 2018
Lindau, Ralf and Victor Venema
