Posters

Abstract

Data availability, defined as the ratio between the number of valid 10-minute mean wind speed values recorded during a measurement campaign relative to the total possible number of 10-minute intervals, is commonly used as an indicator whether a campaign provides a representative measure of the prevailing wind conditions, e.g.: in the OWA (Offshore Wind Accelerator) roadmap for floating lidar systems and the German technical guideline TR6. Data availability can be reduced by factors such as low aerosol content, which reduces the back scattered signal resulting in data being filtered by quality control processes, or when one of more lidar beams are blocked, e.g.: by fouling of the lidar window. Vaisala has developed two algorithms to increase data availability: the 3-LOS algorithm, that reconstructs the horizontal wind speed based on 3 rather than 4 beams, and the range boosting algorithm that compares adjacent radial wind speeds along a line of sight and assess the validity based on a continuity principal. Furthermore, a data quality filter is recommended based on a 10-minute data availability parameter (the total number of valid individual measurements within a 10-minute averaging period). DNV have conducted a study at their lidar verification test site in China, where weather conditions during the winter months commonly lead to very low data availabilities. The algorithms were tested individually and in combination with two different data availability thresholds at the 10-minute level, and the impact on both the campaign data availability and the measurement uncertainty were investigated. The trials show that the algorithms can significantly increase data availability with, in most cases, a modest impact on the measurement uncertainty. For any given measurement campaign, users can balance data completeness and uncertainty according to project needs.