Abstract
We present an analysis of three datasets of 10 min metocean measurement statistics and our resulting recommendations to both producers and users of such datasets. Many of our recommendations are more generally of interest to all numerical measurement data producers. The datasets analyzed originate from offshore meteorological masts installed to support offshore wind farm planning and design: the Dutch OWEZ and MMIJ and the German FINO1. Our analysis shows that such datasets contain issues that users should look out for and whose prevalence can be reduced by producers. We also present expressions to derive uncertainty and bias values for the statistics from information typically available about sample uncertainty. We also observe that the format in which the data are disseminated is sub-optimal from the users' perspective and discuss how producers can create more immediately useful dataset files. Effectively, we advocate using an established binary format (HDF5 or netCDF4) instead of the typical text-based one (comma-separated values), as this allows for the inclusion of relevant metadata and the creation of significantly smaller directly accessible dataset files. Our conclusion is that datasets such as the ones we analyzed can be improved substantially in usefulness and convenience with limited effort.
Introduction: The Foundation of Offshore Wind
Offshore wind energy is a multi-billion dollar industry built on a foundation of "metocean" data—10-minute statistics of wind speed, wave height, and atmospheric pressure. If this data is faulty, wind farms are either under-designed (risking failure) or over-designed (wasting money). This paper provides a critical "health check" for these datasets, identifying systemic issues in data quality and dissemination, and offering a mathematical framework for uncertainty propagation.
Analyzed Datasets
The study evaluated three major offshore meteorological mast datasets located in the North Sea, which are critical for European wind energy planning:
- OWEZ: Offshore Windpark Egmond aan Zee (Netherlands).
- MMIJ: Measuring mast IJmuiden (Netherlands).
- FINO1: Research platform in the German Bight.
Methodology for Data Quality
The researchers performed a deep dive into these datasets using a mix of manual and automated techniques to identify "hidden" faulty data that often passes through initial quality filters:
- Visual Inspection: Using interactive plots to find anomalous extremes that automated filters might miss.
- Consistency Checks: Developing scripts to find physical impossibilities, such as a minimum value being higher than a mean, or values exceeding physical limits (e.g., wind speeds recorded at nearly Mach 5).
- Uncertainty Propagation: Deriving new mathematical expressions to propagate measurement noise into the 10-minute statistics (min, max, mean, and standard deviation).
- Format Transformation: Converting legacy text-based datasets into structured HDF5 and netCDF4 formats to improve accessibility and metadata integration.
Key Insights: Moving Beyond CSVs
A primary finding is the "Tidy Data" trap. While CSV files are common, they are inefficient for large-scale metocean data and lack machine-readable metadata (like sensor height or orientation), forcing researchers to hunt through external PDF manuals.
Furthermore, the paper highlights that the sample standard deviation is a biased estimator when measurement noise is present. The authors provide a correction formula that can reduce turbulence intensity estimates by up to 20%—a massive difference for turbine fatigue calculations.
Results and Recommendations
The transition to binary formats like netCDF4 allowed for an 8x reduction in file size compared to uncompressed text while making the data "self-describing." The authors conclude with a checklist for the industry:
- Project Owners: Require binary formats (HDF5/netCDF4) in commercial contracts.
- Producers: Implement automated consistency checks and include comprehensive metadata within data files.
- Users: Adopt modern binary workflows to improve efficiency and FAIR (Findable, Accessible, Interoperable, Reusable) compliance.
Limitations and Future Work
A major hurdle identified is data licensing; restrictive usage agreements often prevent the public release of improved, "cleaned" versions of datasets. Future work should apply these same rigorous standards to high-frequency raw data (1–50 Hz) to further improve the reliability of offshore wind designs.
Summary updated