Mikkel Bennedsen, Aarhus Universitet

"A New Statistical Reduced Complexity Climate Model"

Abstract

In this paper, we propose a new, fully statistical, reduced complexity climate model. The starting point for our model is a number of physical equations for the global climate system, which we show how to cast in non-linear state-space form. We propose to estimate the model using the method of maximum likelihood obtained from the extended Kalman filter. In an empirical exercise, we use a data set of historical observations from 1959–2019 to estimate the parameters of the model. A likelihood ratio test sheds light on the most appropriate equation for converting the atmospheric concentration of carbon dioxide (GtC) into radiative forcings (W/m2). We use the estimated model and assumptions on future greenhouse gas emissions to project global mean surface temperature out to the year 2100. As an application of the statistical model, we propose a simulation-based approach to construct uncertainty bands to the projections, and use these to quantify how much of the uncertainty is “aleatoric” (uncer- tainty arising from the internal variability of the climate system) and how much is “epistemic” (uncertainty arising from unknown model parameters).

(Co-authered med Eric Hillebrand og Siem Jan Koopman)

Contact person: Rasmus Søndergaard