Transient Earth system responses to cumulative carbon dioxide emissions: linearities, uncertainties, and probabilities in an observation-constrained model ensemble

Abstract

Abstract. Information on the relationship between cumulative fossil CO 2 emissions and multiple climate targets is essential to design emission mitigation and climate adaptation strategies. In this study, the transient response of a climate or environmental variable per trillion tonnes of CO 2 emissions, termed TRE, is quantified for a set of impact-relevant climate variables and from a large set of multi-forcing scenarios extended to year 2300 towards stabilization. An ∼ 1000-member ensemble of the Bern3D-LPJ carbon-climate model is applied and model outcomes are constrained by 26 physical and biogeochemical observational data sets in a Bayesian, Monte Carlo-type framework. Uncertainties in TRE estimates include both scenario uncertainty and model response uncertainty. Cumulative fossil emissions of 1000 Gt C result in a global mean surface air temperature change of 1.9 • C (68 % confidence interval (c.i.): 1.3 to 2.7 • C), a decrease in surface ocean pH of 0.19 (0.18 to 0.22), and a steric sea level rise of 20 cm (13 to 27 cm until 2300). Linearity between cumulative emissions and transient response is high for pH and reasonably high for surface air and sea surface temperatures, but less pronounced for changes in Atlantic meridional overturning, Southern Ocean and tropical surface water saturation with respect to biogenic structures of calcium carbonate, and carbon stocks in soils. The constrained model ensemble is also applied to determine the response to a pulse-like emission and in idealized CO 2 -only simulations. The transient climate response is constrained, primarily by long-term ocean heat observations, to 1.7 • C (68 % c.i.: 1.3 to 2.2 • C) and the equilibrium climate sensitivity to 2.9 • C (2.0 to 4.2 • C). This is consistent with results by CMIP5 models but inconsistent with recent studies that relied on short-term air temperature data affected by natural climate variability.

Reviews and syntheses: parameter identification in marine planktonic ecosystem modelling

“…A full treatment of uncertainties in the projections of ESM is beyond the scope of our review and we can only address this topic here briefly. A comprehensive attempt to account for uncertainties in the models when determining likelihoods of reaching certain climate goals, like the politically widely accepted 2 • C warming goal, was presented by Steinacher et al (2013) and Steinacher and Joos (2016). Employing a somewhat simplified ESM of intermediate complexity, they ran perturbed parameter ensembles with some ad hoc assumptions about prior probability distributions of the model parameters.…”

A Bayesian Ensemble Data Assimilation to Constrain Model Parameters and Land Use Carbon Emissions

“…These may differ with respect to spatial and temporal resolution and quality and include observations from the local scale, such as data from individual biomass measurements or the seasonal CO 2 cycle at individual atmospheric sampling sites, up to global scale gridded data products such as satellite measurements of absorbed radiation by plants. A major advantage of this scheme compared to sequential assimilation techniques such as Ensemble Kalman Filters is that the influence of necessarily subjective framework is easily extendable to incorporate different or more observational constraints and to different mechanistic models including other DGVMs, ocean models (Battaglia et al, 2016) or Earth System Models Steinacher and Joos, 2016)). (v) The Bayesian, skill-score weighted ensemble is able to constrain the median and uncertainty ranges of unknown or uncertain quantities such as carbon emissions from anthropogenic land-use, marine nitrous oxide production (Battaglia and Joos, 2017), or climate sensitivity metrics ) (vi) Finally, the skill-score weighted 5 ensemble is suitable for probabilistic projections including both likely and less likely model configurations and assumptions.…”

“…The assessment of the performance of a given model version using observational benchmarks has also been actively discussed in the literature (Hoffman et al, 2017;Peng et al, 2014;Kelley et al, 2013;Luo 15 et al, 2012;Blyth et al, 2011;Randerson et al, 2009) and different frameworks have been proposed. Here we employ the Latin Hypercube Sampling (LHS) (McKay et al, 1979) approach, as used successfully in previous studies Battaglia et al, 2016;Steinacher and Joos, 2016;Battaglia and Joos, 2017;Zaehle et al, 2005). It allows simultaneous stratified sampling of a range of parameters, given an appropriate prior parameter distribution, while offering the opportunity to change evaluation metrics a posteriori, thus enabling a sensible incorporation of multiple observational constraints.…”

“…In addition to uncertain prescribed LULCC forcings and processes in DGVMs, the parametrization of those processes is subject to further uncertainties. We use a Monte-Carlo-like data assimilation approach (Steinacher 5 et al, 2013;Steinacher and Joos, 2016;Battaglia and Joos, 2017) to sample 15 key model parameters and construct a 1000-member model ensemble to investigate this parameter related uncertainty in the DGVM LPX-Bern. Furthermore, we establish a new reference version of the model.…”

The Bern Simple Climate Model (BernSCM) v1.0: an extensible and fully documented open-source re-implementation of the Bern reduced-form model for global carbon cycle–climate simulations

“…A sufficient coverage of structural uncertainty could allow the interpolation among alternative model components to represent uncertainty with scalable parameters (and removing the distinction between structural and parameter uncertainty). Such a parametrization of the uncertainty would enhance the possibilities for probabilistic applications of BernSCM, although more sophisticated models are avail-able for observation-constrained probabilistic quantification of climate targets (Holden et al, 2010;Steinacher and Joos, 2016;Steinacher et al, 2013).…”