Transforming more than 25 years of turbulence measurements along the Cabauw tower into a high-value research data set.

Turbulence measurements up to 180 m have been performed for over 25 years along the Cabauw tower, The Netherlands, providing a unique data set. Such long-term observations are essential to better understand the turbulent structures in various weather conditions (e.g. convective transport, gravity waves, dominating timescales and eddy sizes) and provide a foundation for weather modelling. For example, such in-situ measurements are increasingly important for the wind energy industry as the turbulence profiles can provide invaluable insight into the impacts of wind (shear) and turbulence intensity which impacts turbine loading and safe operation.

The observations are obtained via a 3D sonic anemometer, providing high-frequency wind measurements and sonic temperature and an open-path gas analyser that measures water vapour and carbon dioxide fluctuations. Over the course of this measurement programme, the measurements have been subject to transition. For instance, the instrumentation is intermittently exchanged for calibration or maintenance, and they have been renewed, upgraded or repaired. Also, data acquisition systems are modernised, set-ups have been moved or altered due to environmental changes or innovation, and processing software to obtain the fluxes has been upgraded with new insights. Now, after 20 years of measurements with similar instrumentation, the sonic anemometer will be replaced by a different brand and different processing software will be incorporated.

The current renewals and frequent requests for turbulence data have encouraged us to investigate the historic data sets, towards standardising and aligning the different time periods into a high-value data set that improves data usability and efficiency. To that end, we first build a complete change log from existing documentation to identify where inhomogeneities can occur. Then we reprocess the data to have the same naming convention and units, apply calibration factors, perform the same coordinate rotation, corrections, averaging, etc. Afterwards we will detect and quantify the discontinuities by visual inspection and statistical tests and find methods to. Last, a consistent quality control will be applied. All processing steps are documented and motivated.

This poster gives an overview of the complex process as well as the challenges we face in making this data set research and poses research questions to which this data is well suited to explore.