A Novel Statistical Method to Temporally Downscale Wind Speed Weibull Distribution Using Scaling Property

Ju-Young Shin,Jun-Haeng Heo,Changsam Jeong

doi:10.3390/en11030633

Abstract

To improve our capacity to use available wind speed data, it is necessary to develop a new statistical temporal downscaling method that uses one or a few input variables of any temporal scale for mean wind speed data to obtain wind statistics at finer temporal resolution. In the present study, a novel statistical temporal downscaling method for wind speed statistics and probability distribution is proposed. The proposed method uses the temporal structure to downscale the wind speed statistics to a fine temporal scale without the use of additional variables. The Weibull distribution of the hourly and 10-min mean wind speed data is obtained by the downscaled wind speed statistics. The proposed method provides the downscaled Weibull distribution of fine temporal wind speed data using coarse temporal wind statistics. Particularly, the use of sub-daily mean wind speed data in the downscaling of the wind speed Weibull distribution leads to good estimation precision. The Weibull distribution downscaled by the proposed method successfully reproduces the wind power density based on the wind potential energy estimation.

Highlights

IntroductionMany regions that are expected to have a large wind power potential do not have weather stations that can measure wind speed
When planning wind farms, many regions that are expected to have a large wind power potential do not have weather stations that can measure wind speed
The ordinary least square (OLS) method method leads to the lower root mean square error (RMSE) than the weight least square (WLS) method

Summary

Introduction

Many regions that are expected to have a large wind power potential do not have weather stations that can measure wind speed. To assess the wind power potential in a region that does not have wind speed observations, the development of advanced climate models and remote sensing techniques allow for the use of various wind speed observations or estimates for investigations of the characteristics of wind speed data and wind power potential assessments in many regions [1,2,3]. The OSCAT wind speed data can be used for wind power potential assessments in this region, where there is a scarcity of in situ wind speed data These data improve our capacity to model wind speed and assess the wind power potential in these regions. The temporal and spatial scales of these wind observations and estimates are relatively coarse for investigating the detailed characteristics of the wind potential energy. Wind speed observations and estimates at a finer scale are required for an accurate wind power potential assessment

Objectives

Methods

Results

Conclusion