Study of wind speed and relative humidity using stochastic technique in a semi-arid climate region

Abstract

This paper deals with the stochastic analysis of wind speed based on relative humidity data. We propose a stochastic regression technique to estimate the time-varying parameters of wind speed in a semi-arid climate region. The modeling of stochastic parameters of atmospheric data with consistent properties facilitates prediction with higher precision. In order to compare the estimation, we used simulated atmospheric time series and observational time series. The atmospheric time series was generated by the Weather Research and Forecasting (WRF) model, whereas the observational time series was obtained from the surface weather stations. The time-varying parameters of the model used are estimated by Maximum Likelihood process. The results obtained suggest that relative humidity exhibits a stochastic effect to predict stationary wind speed data. This type of analysis helps to characterize some key meteorological variables, which would be useful in forecasting irregular wind speed.

Keywords

• wind speed
• relative humidity
• maximum likelihood estimation
• stochastic regression
• bootstrapping
• wrf model