The quality of meteorological measurements deserves continuous attention. KNMI therefore follows as close as possible the regulations of the World Meteorological Organization and the guidelines for climate monitoring for these measurements. Sometimes it is possible to improve the quality of historical measurements with retroactive effect, for instance due to methodological improvements. This is the case for the study in this report.
At the start of this work, it was known that for the five principal stations in the Netherlands several methods had been used in the 1901-1970 period to calculate daily minimum and maximum temperatures (Tn and Tx) from the available measurements. It was, however, not completely clear which methods had been used for each of the stations and how this affected the time series. In addition, for the 1946-1970 period climatological and synoptical hourly temperature data were available for the principal stations. So far, it was unclear which data type is closest to the real temperature of that time. The type of hourly data may affect, however, the Tn and Tx and the daily mean temperature (Tmean). The heterogeneity of data and methods is undesirable and limits the usefulness of the time series. The work described in this report was performed to standardize the data and the methods used for calculating Tmean, Tn and Tx and thus obtain a more robust dataset.
A reconstruction is made here of the data and methods used to calculate Tmean, Tn and Tx for the five principal stations. The methods and data are standardized and new daily Tmean, Tn and Tx time series are calculated for the five principal stations. These time series will be implemented by KNMI as the new operational daily temperature series for the five stations. The update improves the quality of the existing series substantially.
Measured daily temperature time series are commonly used when informing the general public on temperature extremes.
The standardized time series can be considered as an update of the measured time series. The updated series will now serve as basis for the calculation of extreme values. The changes in extremes are described in this report. The standardization does not remove inhomogeneities due to for instance station relocations. Currently there is no generally accepted method to correct for inhomogeneities on a daily basis.
Theo Brandsma, Rudmer Jilderda and Rob Sluijter. Standardization of data and methods for calculating daily Tmean, Tn and Tx in the Netherlands for the 1901-1970 period
KNMI number: TR-340, Year: 2013, Pages: 48