Abstract:We consider the problem of short-term forecasting of surface wind speed probability distribution. Our approach simply consists in predicting the parameters of a given probability density function by training a neural network model whose loss function is chosen as the log-likelihood provided by this distribution. We compare different possibilities among a set of distributions that have been previously considered in the context of modeling wind fluctuations. Our results rely on two different hourly wind speed datasets: the first one has been recorded by Météo-France in Corsica (South France), a very mountainous Mediterranean island while the other one relies on KNMI database that provides records of various stations over The Netherlands, a very flat country in Northwestern Europe. A first part of our work globally unveils the superiority of the so-called "Multifractal Rice" (M-Rice) distribution over alternative parametric models, showcasing its potential as a reliable tool for wind speed forecasting. This family of distributions has been proposed in the context of modeling wind speed fluctuations as a random cascade model along the same picture as fully developed turbulence. For all stations in both regions, it consistently provides better results regardless of the considered probabilistic scoring rule or forecasting horizon. Our second findings demonstrate significant enhancements in forecasting accuracy when one incorporates wind speed data from proximate weather stations, in full agreement with the results obtained formerly for point-wise wind speed prediction. Moreover, we reveal that the incorporation of ERA5 reanalysis of 10 m wind data from neighboring grid points contributes to a substantial improvement at time horizon h=6 hours while at a shorter time horizon ( h=1 h) it does not lead to any noticeable improvement. It turns out that accounting for pertinent features and explanatory factors, notably those related the spatial distribution and wind speed and direction, emerges as a more critical factor in enhancing accuracy than the choice of the "optimal" parametric distribution. We also find out that accounting for more explanatory factors mainly increases the resolution performances while it does not change the reliability contribution to the prediction performance metric considered (CRPS).

Improving probabilistic wind speed forecasting using M-Rice distribution and spatial data integration

Improving probabilistic wind speed forecasting using M-Rice distribution and spatial data integration

Probabilistic and deterministic wind speed forecasting based on non-parametric approaches and wind characteristics information

Efficient modeling of sub-kilometer surface wind with Gaussian processes and neural networks

Leveraging data from nearby stations to improve short-term wind speed forecasts

SHM-based Probabilistic Representation of Wind Properties: Bayesian Inference and Model Optimization

Probabilistic Method for Wind Speed Prediction and Statistics Distribution Inference Based on SHM Data-Driven

Data-Driven Day-ahead Probabilistic Forecasting of Wind Power Based on Features Sensitivity Analysis and Meteorological Scenario Classification

Wind power predictions from nowcasts to 4-hour forecasts: A learning approach with variable selection

A robust error correction method for numerical weather prediction wind speed based on Bayesian optimization, variational mode decomposition, principal component analysis, and random forest: VMD-PCA-RF (version 1.0.0)

Short-Term Probabilistic Forecasting Method for Wind Speed Combining Long Short-Term Memory and Gaussian Mixture Model

Gaussian mixture models for the optimal sparse sampling of offshore wind resource

Smart Urban Wind Power Forecasting: Integrating Weibull Distribution, Recurrent Neural Networks, and Numerical Weather Prediction

Statistical Modelling of 2m Temperature and 10m Wind Speed Forecast Errors

Joint Modeling of Wind Speed and Power via a Nonparametric Approach

Deterministic and Interval Wind Speed Prediction Method in Offshore Wind Farm Considering the Randomness of Wind

Sequential Aggregation of Probabilistic Forecasts -- Applicaton to Wind Speed Ensemble Forecasts

Spatiotemporal wind speed forecasting using conditional local convolution and multidimensional meteorology features

Support Vector Regression Method for Wind Speed Prediction Incorporating Probability Prior Knowledge

Enhancing Regional Wind Power Forecasting through Advanced Machine-Learning and Feature-Selection Techniques

Improving sub-seasonal wind-speed forecasts in Europe with a non-linear model