Forecasting Chinese carbon emissions from fossil energy consumption using non-linear grey multivariable models

Abstract Much theoretical and empirical research has verified the non-linear and uncertain relationships between carbon emissions and economic growth. To forecast the carbon emissions from fossil energy consumption, this paper introduces the power exponential term of the relevant variables as exogenous variables into a multivariable grey model. Under the target of minimisation of the mean absolute percentage error, two non-linear programming models are constructed to solve the unknown parameters of the non-linear grey multivariable model. In addition, to enhance the adaptability of the grey model to large sample sizes, we divide the data of Chinese gross domestic product and carbon emissions from fossil energy consumption of 1953–2013 into 15 stages. The empirical results show that the non-linear grey multivariable model can reflect the mechanism of the non-linear effects of gross domestic product on carbon emissions from fossil energy consumption, and has higher forecast accuracy than the traditional grey model and the autoregressive integrated moving average models. In three schemes – economic growth at low, medium, and high, speeds – we use the non-linear grey model to quantify future Chinese carbon emissions from fossil energy consumption from 2014 to 2020, and the predicted results can provide the basis for energy planning and the formulation of environmental policy.