High-Resolution Coastal Wind Energy Assessments Based on Copula Models and Meteorological Data

Abstract Accurate wind energy assessment (WEA) is essential to reduce the risk of wind project investment. In this paper, the overall framework and detailed process of wind resource assessment are presented for the first time, and the effectiveness of this method is validated by combining meteorological data with copula models. The average hourly wind speed and direction in Wenchang, Hainan Province, from 2016 to 2017 are used to study the joint probability distribution function (JPDF). However, the marginal probability density function (MPDF) of wind direction and wind speed is established and compared by parametric and nonparametric methods. The comparison results reveal that nonparametric kernel density estimation (KDE) performs better than parametric distributions in fitting wind speed empirical distribution but worse in wind direction empirical distribution. The copula function is subsequently utilized to construct JPDF. Six individual and comprehensive goodness-of-fit (GoF) metrics are employed to discern the ideal model out of all the potential models. The evaluation results show that the optimal model varies from year to year, and the Gumbel and Ali–Mikhail–Haq (AMH) copula models can best reproduce the 2-yr empirical distribution of wind speed and wind direction, respectively. Compared with the WEA method based on wind speed, the efficacy of the copula method is confirmed. The study simplifies the WEA procedure and realizes the systematic evaluation of available wind resources, which is of great significance for the further study of direction-related WEA in other regions.