Investigating the accuracy of Swiss water surface temperature measurements

For decades, the Federal Office for the Environment has observed an increase in water temperatures in many Swiss watercourses. Extensive studies to determine the statistical and systematic uncertainty of temperature measurements of watercourses are being undertaken to verify the validity of series of long-term measurements.


  • Lead department Engineering and information technology
  • Institute Statistical Modelling
  • Research unit Statistical Data Analysis
  • Duration (planned) 01.06.2017 - 31.01.2020
  • Project management IODA
  • Head of project Jasmin Wandel
  • Project staff Vidushi Bigler
    Olivier Mermoud
    Jasmin Wandel
  • Partner Fischwerk
    BFH PV-Lab (Thomas Schott)
    Federal Office for the Environment FOEN (Begleitung Thilo Herold)
  • Keywords Long-term Measurement, Measurement Uncertainty, Precision, Watercourses, Time Series

Initial Situation

The Federal Office for the Environment has been measuring the temperature of various watercourses in Switzerland since the 1950s. For decades, the average annual temperature of many large rivers has been rising, along with the frequency of limit value violations in the water temperature. Due to external factors (e.g. sunshine, shade, water levels) and the limited accuracy of the measuring instruments, however, the temperature measurements of watercourses are subject to uncertainty. This project takes an extensive look at the validity of temperature measurements in bodies of water. This is done by taking comparative in situ measurements and analysing statistics to ascertain the significance and scope of various influencing factors on temperature measurement.


The use of statistical methods and mathematical models should permit reliable statements to be made with regard to measurement accuracy. The insights gained will help formulate measures to enhance measurement accuracy (e.g. positioning of measuring equipment). The next step is to integrate cantonal data and look at whether to extend the project to include smaller bodies of water.


To start with, the project evaluated suitable statistical methods for estimating the measurement uncertainty. This was followed by a pilot study that generated high-resolution data, allowing an initial estimate of the measurement uncertainty to be made. Appropriate mathematical models were used to model the data.


The pilot project found indications that external factors have a temporary influence on water temperature measurements. It is proving difficult to precisely quantify the measurement inaccuracy because the measurement configuration has not yet been optimally defined. This therefore requires further investigations with adjusted measurement configurations.