Wind speed monitoring using entropy theory and a copula-based approach

Abstract

In this study, wind speeds in the Lut desert (Iran) was monitored at the Bam, Tabas and Birjand stations during the period 1973–2020 using entropy theory. The conventional entropy method was improved by considering the interaction between stations by a copula-based approach. Two different methods were examined for this purpose: 1) A trivariate vine copula was used to evaluate the interaction among the stations. For this purpose, a d-vine copula was considered to simulate the wind speed at each station as a function of the wind speed at the other stations. 2) A bivariate copula was used to model the simulated and observed wind speed at each station. In particular, a Frank copula was used to perform the joint probability analysis. The results indicate that implementing the proposed method for wind speed monitoring increases the Information Transfer Index (ITI) by approximately 24 %, 10 %, and 33 % at the Bam, Birjand, and Tabas stations, respectively, compared to using the conventional entropy method.