Improved near surface wind speed predictions using Gaussian process regression combined with numerical weather predictions and observed meteorological data

With the large-scale development of wind energy, wind power forecasting plays a key role in power dispatching in the electric power grid, as well as in the operation and maintenance of wind farms. The most important technology for wind power forecasting is forecasting wind speed. The current mainstream methods for wind speed forecasting involve the combination of mesoscale numerical meteorological models with a post-processing system. Our work uses the WRF model to obtain the numerical weather forecast and the gradient boosting decision tree (GBDT) algorithm to improve the near-surface wind speed post-processing results of the numerical weather model. We calculate the feature importance of GBDT in order to find out which feature most affects the post-processing wind speed results. The results show that, after using about 300 features at different height and pressure layers, the GBDT algorithm can output more accurate wind speed forecasts than the original WRF results and other post-processing models like decision tree regression (DTR) and multi-layer perceptron regression (MLPR). Using GBDT, the root mean square error (RMSE) of wind speed can be reduced from 2.7–3.5 m/s in the original WRF result by 1–1.5 m/s, which is better than DTR and MLPR. While the index of agreement (IA) can be improved by 0.10–0.20, correlation coefficient be improved by 0.10–0.18, Nash–Sutcliffe efficiency coefficient (NSE) be improved by −0.06–0.6. It also can be found that the feature which most affects the GBDT results is the near-surface wind speed. Other variables, such as forecast month, forecast time, and temperature, also affect the GBDT results.

Download Full-text

Recurrence Characteristics Analysis of Near-Surface Wind Speed Signal with Different Sampling Frequencies

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.599-601.1605 ◽

2014 ◽

Vol 599-601 ◽

pp. 1605-1609 ◽

Cited By ~ 1

Author(s):

Ming Zeng ◽

Zhan Xie Wu ◽

Qing Hao Meng ◽

Jing Hai Li ◽

Shu Gen Ma

Keyword(s):

Wind Speed ◽

Surface Wind ◽

Surface Wind Speed ◽

Recurrence Plot ◽

Sampling Frequency ◽

Near Surface ◽

Gas Leakage ◽

Time Period ◽

Characteristics Analysis ◽

Quantification Analysis

The wind is the main factor to influence the propagation of gas in the atmosphere. Therefore, the wind signal obtained by anemometer will provide us valuable clues for searching gas leakage sources. In this paper, the Recurrence Plot (RP) and Recurrence Quantification Analysis (RQA) are applied to analyze the influence of recurrence characteristics of the wind speed time series under the condition of the same place, the same time period and with the sampling frequency of 1hz, 2hz, 4.2hz, 5hz, 8.3hz, 12.5hz and 16.7hz respectively. Research results show that when the sampling frequency is higher than 5hz, the trends of recurrence nature of different groups are basically unchanged. However, when the sampling frequency is set below 5hz, the original trend of recurrence nature is destroyed, because the recurrence characteristic curves obtained using different sampling frequencies appear cross or overlapping phenomena. The above results indicate that the anemometer will not be able to fully capture the detailed information in wind field when its sampling frequency is lower than 5hz. The recurrence characteristics analysis of the wind speed signals provides an important basis for the optimal selection of anemometer.

Download Full-text

From Numerical Weather Prediction Outputs to Accurate Local Surface Wind Speed: Statistical Modeling and Forecasts

Springer Proceedings in Mathematics & Statistics - Renewable Energy: Forecasting and Risk Management ◽

10.1007/978-3-319-99052-1_2 ◽

2018 ◽

pp. 23-44 ◽

Cited By ~ 2

Author(s):

Bastien Alonzo ◽

Riwal Plougonven ◽

Mathilde Mougeot ◽

Aurélie Fischer ◽

Aurore Dupré ◽

...

Keyword(s):

Wind Speed ◽

Statistical Modeling ◽

Numerical Weather Prediction ◽

Weather Prediction ◽

Surface Wind ◽

Surface Wind Speed ◽

Local Surface ◽

Numerical Weather

Download Full-text

A preliminary study on the influence of Beijing urban spatial morphology on near-surface wind speed

Urban Climate ◽

10.1016/j.uclim.2020.100703 ◽

2020 ◽

Vol 34 ◽

pp. 100703

Author(s):

Yonghong Liu ◽

Yongming Xu ◽

Fangmin Zhang ◽

Wenjun Shu

Keyword(s):

Wind Speed ◽

Surface Wind ◽

Surface Wind Speed ◽

Near Surface ◽

Spatial Morphology ◽

Preliminary Study

Download Full-text

Characteristics and Applications of the Ground-Based X Band Low Elevation Angle Brightness Temperatures under Low Sea State Based on Measured Data

Remote Sensing ◽

10.3390/rs12111736 ◽

2020 ◽

Vol 12 (11) ◽

pp. 1736

Author(s):

Zhongqing Cao ◽

Lixin Guo ◽

Shifeng Kang ◽

Xianhai Cheng ◽

Qingliang Li ◽

...

Keyword(s):

Wind Speed ◽

Meteorological Data ◽

Surface Wind ◽

Elevation Angle ◽

Surface Wind Speed ◽

Sea Surface ◽

Sea State ◽

Sea Surface Wind Speed ◽

Sea Surface Wind ◽

X Band

In ground-based microwave radiometer remote sensing, low-elevation-angle (−3°~3°) radiation data are often discarded because they are considered to be of little value and are often difficult to model due to the complicated mechanism. Based on the observed X-band horizontal polarization low elevation angle microwave radiation data and the meteorological data at the same time, this study investigated the generation mechanism of low elevation angle brightness temperature (LEATB) and its relationship with meteorological data, i.e., temperature, humidity, and wind speed, under low sea state. As a result, one could find that the LEATB was sensitive to the atmosphere at the elevation angle between 1° to 3°, and a diurnal variation of the LEATB reached up to 10 K. This study also found a linear relationship between the LEATB and sea surface wind speed under low sea state at an elevation range from −3° to 0°, i.e., the brightness temperature decreased as the wind speed increased, which was inconsistent with the observations at the elevation angle from −10° to −5°. The variation of the LEATB difference according to the change in the over-the-horizon detection capability (OTHDC) of the shipborne microwave radar was examined to identify the reason for this phenomenon theoretically. The results showed that the LEATB difference was significantly influenced by a change in the OTHDC. Further, this study examined a remote sensing method to extract the sea surface wind speed data from experimental LEATB data under low sea state. The results demonstrated that the X-band horizontal polarization LEATBs were useful to retrieve the sea surface wind speed data at a reasonable accuracy—the root mean square error of 0.02408 m/s. Overall, this study proved the promising potential of the LEATB data for retrieving temperature profiles, humidity profiles, sea surface winds, and the OTHDC.

Download Full-text