Development of an integrated model on the basis of GCMs-RF-FA for predicting wind energy resources under climate change impact: A case study of Jing-Jin-Ji region in China

Abstract

Climate change mitigation and renewable resources utilization are becoming particularly urgent for socioeconomic sustainability and human survival. This study develops an integrated model (named as GCMs-RF-FA) for long-term wind energy resources prediction through incorporating multiple global climate models (GCMs), random forest (RF) and factorial analysis (FA) techniques within a general framework. GCMs-RF-FA is capable of (i) dealing with the heterogeneity of structures and parameters in multiple GCMs, (ii) addressing the nonlinear/discrete relationships and uncertainties between the variables, and (iii) identifying the individual and interactive effects of main factors on wind resources prediction. GCMs-RF-FA is then applied to Jing-Jin-Ji region for wind energy prediction under two representative emission scenarios (RCP4.5 & RCP8.5) for 80 years. Multiple validation coefficients prove that the proposed model is effective and feasible. Results reveal that the main factors affecting wind power density (WPD) prediction and their effects at different sites. Results also show that (i) the predicted wind energy value under RCP4.5 is lower than that under RCP8.5; (ii) the seasonal difference of wind energy is obvious, high in winter and spring, low in autumn and summer; (iii) in 2060, the shares of wind energy resources in Beijing, Tianjin and Hebei account for 17.8%, 11.6% and 46.1% of total energy supply, respectively.