Modified

27 October 2026

Raw Data: Eddy Covariance

The raw eddy covariance data are not part of FP2026.1. This page describes the data the fluxes were calculated from.

Measurement

  • Height and rate. The sonic anemometer and the gas analysers are mounted at 47 m on the tower and sampled at 20 Hz. The sonic orientation is 209° (see Level 0).
  • Gas analysers.
    • IRGA72, the enclosed-path LI-7200, from 11 January 2016. The fluxes in FP2026.1 are calculated from this analyser.
    • IRGA75, the open-path LI-7500, from 2004 to 12 December 2017, and from 17 January to 20 March 2019 as a stand-in for the defective LI-7200. Its fluxes are in progress for FP2026.2 and are not part of FP2026.1.
    Both analysers measured in parallel from 11 January 2016 to 12 December 2017. Models, serial numbers and exchanges of all instruments are listed on Instrumentation.

Files

The raw data are stored as binary files, each typically covering 6 hours. The file name gives the start of the file: 2013070615.b02 starts on 6 July 2013 at 15:00.

The setup changes over time. Each change of hardware or configuration starts a new setup period, labelled by year and number (2016_2). The setup table on Yearly Notes lists every setup period with its first and last raw file, its instruments and their separations. Its numbered notes are explained on the SwissFluxNet page on the raw binary format.

Conversion to ASCII

The binary files are converted to ASCII with bico before the flux calculation. Two corrections are made in this step:

  • Wrong calibration gas. All IRGA72 CO2 concentrations from 14 December 2017 to 15 March 2019 were calibrated against the wrong reference gas. They are multiplied by 0.974 during the conversion, so every flux calculated from them is corrected.
  • File names in July 2013. The names of the files from 2013070615.b02 to 2013071214.b00 give a start time that is wrong by about 6 hours. The converted files are renamed after the conversion. This affects only the IRGA75 record.

Known limitations

  • No fluxes from 13 December 2021 to 7 February 2022. The raw data of setup periods 2021_2 and 2022_1 were logged in an unknown format, and the conversion fails on an unrecognised analyser record. The data may be corrupted.
Back to top