On the development of benchmarking for eddy covariance data quality checking schemes

0503821 - ÚFA 2020 DE eng A - Abstrakt
Šigut, L. - Sedlák, Pavel - Fišer, M. - Jocher, G. - Pavelka, M. - Mauder, M.
On the development of benchmarking for eddy covariance data quality checking schemes.
Geophysical Research Abstracts. Vol. 21. Göttingen: European Geosciences Union, 2019. EGU2019-7216.
[EGU General Assembly 2019. 07.04.2019-12.04.2019, Vienna]
Institucionální podpora: RVO:68378289
Klíčová slova: eddy covariance measurement * post-processing * data quality checking
Obor OECD: Meteorology and atmospheric sciences
https://meetingorganizer.copernicus.org/EGU2019/EGU2019-6898.pdf

With the development of a standardized eddy covariance (EC) data processing software it is now possible to create a high-throughput post-processing chain that allows to assess various statistics extracted from the datasets. The effect of changes in quality checking (QC) procedures can be thus effectively analysed. It is still difficult to establish what constitutes an effective QC test or filter, in other words, how exactly to define such filtering efficiency. One of the main obstacles is the design of such filters that can focus on issues that span from very specific problems to general tests of basic assumptions of the EC method. Here we propose a framework to benchmark different QC schemes based on a combination of three main measures extracted from the dataset – 1) difference in the aggregated EC fluxes, 2) change in the aggregated uncertainty and 3) data exclusion fraction. It is clear that relaxing of QC cannot be justified simply by the effort to retain data coverage in case of frequent spurious records. However, inclusion of a filter that increases exclusion fraction while the post-processed data do not differ in aggregated fluxes and uncertainties can be hardly accepted. In order to push this topic forward it would be also beneficial to provide easily available and reliable QC tools for the EC community and collect the feedback. This was already accomplished by publishing the R package openeddy (https://github.com/lsigut/openeddy) and related workflows. The feedback can help to make the codes more robust and also draw attention to new QC related issues not addressed before. The software could be therefore a platform for an exchange of experience from EC measurements across different ecosystem types and environmental conditions.
This work was supported by the Ministry of Education, Youth and Sports of the Czech Republic within the CzeCOS program, grant number LM2015061, and within the National Sustainability Program I (NPU I), grant number LO1415.
Trvalý link: http://hdl.handle.net/11104/0295606