Predicting Global Temperature Anomaly: A Definitive Investigation Using an Ensemble of Twelve Competing Forecasting Models

In this paper we analyze whether (anthropometric) CO2 can forecast global temperature anomaly (GT) over an annual out-of-sample period of 1907-2012, using an in-sample of 1880- 1906. For our purpose, we use 12 parametric and non-parametric univariate (only comprising of GT) and multivariate (including both GT and CO2) models. Our results show that the Horizontal Multivariate Singular Spectral Analysis (HMSSA) models (both Recurrent (-R) and Vector (-V)) consistently outperform the other competing models. More importantly, from the performance of the HMSSA-R model, we find conclusive evidence that CO2 can forecast GT, and predict its direction of change. Our results highlight the superiority of the nonparametric approach of the SSA, which in turn, allows us to handle any statistical process, i.e., linear or nonlinear, stationary or non-stationary, Gaussian or non-Gaussian.