Dresden 2026 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

DY: Fachverband Dynamik und Statistische Physik

DY 15: Focus Session: Large Deviations and Rare Events II

DY 15.12: Vortrag

Montag, 9. März 2026, 18:15–18:30, ZEU/0114

Forecasting Extreme Events in Atmospheric Turbulence — •Finn Köhne and Joachim Peinke — Institute of Physics and ForWind, University of Oldenburg, Küpkersweg 70, D-26129 Oldenburg, Germany

Extreme wind speed fluctuations on time scales of seconds to minutes generate potentially significant mechanical loads for example on wind turbines and may lead to strong fluctuations in the power output [1,2]. Reliable short-term forecasts of such turbulent events are therefore of high relevance for many applications.

In this contribution, we use a stochastic framework based on the Fokker-Planck-equation (FPE) to forecast wind speed fluctuations in the atmospheric turbulence. We estimate drift and diffusion from measurement data to derive FPEs that describe the probabilistic evolution of wind speed increments without requiring data normalization [3].

The resulting model is used to perform probabilistic forecasts for extreme wind speed fluctuations. The quality of the forecast is quantified using receiver operating characteristic curves. The results demonstrate that the Fokker-Planck-based approach provides reliable short-term predictions of extreme events in the atmospheric turbulence.

[1] Davenport, A. G.; Proc. Inst. Civ. Eng. 19 (1961), Nr. 4, S. 449-472.

[2] DeMarco, A.; Basu, S.; Wind Energy 21 (2018), Nr. 10, S. 892-905.

[3] Peinke, J.; Tabar, M. R. R.; Wächter, M.; Annu. Rev. Condens. Matter Phys. 10 (2019), S. 107-132.

Keywords: Fokker-Planck; Markov; Turbulence; Extreme events