Global temperature projections from a statistical energy balance model using multiple sources of historical data

Abstract This paper estimates a two-component energy balance model as linear state space system (EBM-SS model) using historical data. It is joint for the temperature in mixed layer, deep ocean and radiative forcing. The EBM-SS allows modeling of non-stationarity forcing incorporation multiple data sources unobserved processes. We estimate datasets at global level period 1955 to 2020 by maximum likelihood. show empirical estimation simulations that processes reduces parameter uncertainty. When fitting six observational mean surface (GMST) anomaly series, GMST projections under Representative Concentration Pathway scenarios are comparable those from Coupled Model Intercomparison Project models. results simple statistical climate estimated on can produce compatible with output large-scale Earth System Models.