Spatiotemporally resolved turbulent wind fields from atmospheric turbulence measurements
Abstract
In earth sciences, astrophysics or turbulent environments such as wind parks, many relevant physical quantities fluctuate in a non-Gaussian fashion. This oftentimes imposes major challenges for the acquisition of finely-resolved measurements which would be necessary for a detailed analysis or modeling of the underlying physical processes or as a foundation for machine learning training. E.g., comprehensive assessments of operational performance and maintenance levels (and ultimately cost efficiency) of wind parks, are in strong need of an accurate determination of wind speeds with a temporal resolution of approximately 0.1 seconds. As of today, however, industrial standards record solely the statistics of ten-minute- means of wind speed and turbulence intensity. It is therefore, of the utmost importance to develop methods that allow for an accurate reconstructions of wind fields from sparse wind data. In this contribution, we present paths to generate synthetic signals from sparsely sampled data points, by a recently proposed stochastic interpolation in combination with a method to generate non-Gaussian fluctuations at small scales which follow the Kolmogorov-Oboukhov model of turbulence. Potential applications of the proposed method range from better statistical estimates for power output and fatigue loads on individual wind turbines to mesoscale models of meteorologic turbulence simulations. E.g., the stochastic interpolation might be used to resolve small time scales of individual wind turbine measurements, for instance from cup anemometers. This may ultimately lead to a more genuine assessment of past (for instance, for the past 5 years) extreme fatigue loads on rotor and blades which may help preventing unscheduled maintenance. Furthermore, spatio-temporal wind fields of entire wind plants might be reconstructed on the basis of simultaneous wind speed measurements at multiple space points, for example by laser doppler anemometry (LIDAR). We also discuss the reconstruction of Eulerian velocity fields (in a fixed frame of reference) from a couple of Lagrangian trajectories by means of spatiotemporal fractional Brownian bridge processes.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFMNG35B0449F