Modeling the impacts of climate change on monthly commercial sector electricity consumption: A case of Suzhou, China

Abstract

Multiple regression equations are the most common method for modeling the influences of climate variables on electricity demand. However, the multiple regression equations based on time series data often suffer from the problem of serial correlation or heteroskedasticity, resulting in biased estimation of the coefficients of parameters. Given that previous studies have paid less attention to such issues, in quantifying the effects of climate change on monthly commercial electricity consumption, in Suzhou, China, we paid more attention to the detection of the autocorrelation or heteroskedasticity that exists in a regression model. The results confirm that if a regression equation suffers from autocorrelation or heteroskedasticity, taking the necessary remedial measures can significantly improve the goodness of fit for the model. Again, it also found that if the temperature is less than the lower limit of the comfort zone, the relationship between climate change and electricity consumption is linear; if the temperature exceeds the upper limit of the comfort zone, this relationship is nonlinear, and a quadratic curve is more appropriate for modeling this relationship.