Nathan M. Urban

Climate Sensitivity Estimated from Temperature Reconstructions of the Last Glacial Maximum

Last Glacial Maximum temperature reconstructions and model simulations can constrain the equilibrium climate sensitivity. Assessing the impact of future anthropogenic carbon emissions is currently impeded by uncertainties in our knowledge of equilibrium climate sensitivity to atmospheric carbon dioxide doubling. Previous studies suggest 3 kelvin (K) as the best estimate, 2 to 4.5 K as the 66% probability range, and nonzero probabilities for much higher values, the latter implying a small chance of high-impact climate changes that would be difficult to avoid. Here, combining extensive sea and land surface temperature reconstructions from the Last Glacial Maximum with climate model simulations, we estimate a lower median (2.3 K) and reduced uncertainty (1.7 to 2.6 K as the 66% probability range, which can be widened using alternate assumptions or data subsets). Assuming that paleoclimatic constraints apply to the future, as predicted by our model, these results imply a lower probability of imminent extreme climatic change than previously thought.

Probabilistic hindcasts and projections of the coupled climate, carbon cycle, and Atlantic meridional overturning circulation system: A Bayesian fusion of century-scale observations with a simple model

How has the Atlantic Meridional Overturning Circulation (AMOC) varied over the past centuries and what is the risk of an anthropogenic AMOC collapse? We report probabilistic projections of the future climate which improve on previous AMOC projection studies by (i) greatly expanding the considered observational constraints and (ii) carefully sampling the tail areas of the parameter probability distribution function (pdf). We use a Bayesian inversion to constrain a simple model of the coupled climate, carbon cycle, and AMOC systems using observations to derive multi-century hindcasts and projections. Our hindcasts show considerable skill in representing the observational constraints. We show that robust AMOC risk estimates can require carefully sampling the parameter pdfs. We find a low probability of experiencing an AMOC collapse within the 21st century for a business-as-usual emissions scenario. The probability of experiencing an AMOC collapse within two centuries is 1/10. The probability of crossing a forcing threshold and triggering a future AMOC collapse (by 2300) is approximately 1/30 in the 21st century and over 1/3 in the 22nd. Given the simplicity of the model structure and uncertainty in the forcing assumptions, our analysis should be considered a proof of concept and the quantitative conclusions subject to severe caveats.

A climate sensitivity estimate using Bayesian fusion of instrumental observations and an Earth System model

[1] Current climate model projections are uncertain. This uncertainty is partly driven by the uncertainty in key model parameters such as climate sensitivity (CS), vertical ocean diffusivity (Kv), and strength of anthropogenic sulfate aerosol forcing. These parameters are commonly estimated using ensembles of model runs constrained by observations. Here we obtain a probability density function (pdf) of these parameters using the University of Victoria Earth System Climate Model (UVic ESCM) - an intermediate complexity model with a dynamic three-dimensional ocean. Specifically, we run an ensemble of UVic ESCM runs varying parameters that affect CS, ocean vertical diffusion, and the effects of anthropogenic sulfate aerosols. We use a statistical emulator that interpolates the UVic ESCM output to parameter settings where the model was not evaluated. We adopt a Bayesian approach to constrain the model output with instrumental surface temperature and ocean heat observations. Our approach accounts for the uncertainties in the properties of model-data residuals. We use a Markov chain Monte Carlo method to obtain a posterior pdf of these parameters. The mode of the climate sensitivity estimate is 2.8°C, with the corresponding 95% credible interval ranging from 1.8 to 4.9°C. These results are generally consistent with previous studies. The CS pdf is sensitive to the assumptions about the priors, to the effects of anthropogenic sulfate aerosols, and to the background vertical ocean diffusivity. Our method can be used with more complex climate models.

Multivariate Gaussian Process Emulators With Nonseparable Covariance Structures

The Gaussian process regression model is a popular type of “emulator” used as a fast surrogate for computationally expensive simulators (deterministic computer models). For simulators with multivariate output, common practice is to specify a separable covariance structure for the Gaussian process. Though computationally convenient, this can be too restrictive, leading to poor performance of the emulator, particularly when the different simulator outputs represent different physical quantities. Also, treating the simulator outputs as independent can lead to inappropriate representations of joint uncertainty. We develop nonseparable covariance structures for Gaussian process emulators, based on the linear model of coregionalization and convolution methods. Using two case studies, we compare the performance of these covariance structures both with standard separable covariance structures and with emulators that assume independence between the outputs. In each case study, we find that only emulators with nonseparable covariances structures have sufficient flexibility both to give good predictions and to represent joint uncertainty about the simulator outputs appropriately. This article has supplementary material online.

A comparison of Latin hypercube and grid ensemble designs for the multivariate emulation of an Earth system model

A statistical emulator is a fast proxy for a complex computer model which predicts model output at arbitrary parameter settings from a limited ensemble of training data. Regular grid designs for the training set are commonly used for their simplicity. However, Latin hypercube designs have well known theoretical advantages in the design of computer experiments, especially as the dimension of the parameter space grows. Here we use time series output from a simple Earth system model to compare the influence of these two design choices on the cross-validation prediction skill of a statistical emulator. We find that an emulator trained on a Latin hypercube design shows a small but clear improvement in prediction quality relative to an emulator trained on a grid design. We also find that the Latin hypercube emulator is more accurate than the grid emulator in single-parameter model sensitivity studies. We conclude with a discussion of ensemble design choices for emulator computer experiments.

Historical and future learning about climate sensitivity

Equilibrium climate sensitivity measures the long-term response of surface temperature to changes in atmospheric CO2. The range of climate sensitivities in the Intergovernmental Panel on Climate Change Fifth Assessment Report is unchanged from that published almost 30 years earlier in the Charney Report. We conduct perfect model experiments using an energy balance model to study the rate at which uncertainties might be reduced by observation of global temperature and ocean heat uptake. We find that a climate sensitivity of 1.5 ◦ C can be statistically distinguished from 3 ◦ C by 2030, 3 ◦ C from 4.5 ◦ C by 2040, and 4.5 ◦ Cf rom 6 ◦ C by 2065. Learning rates are slowest in the scenarios of greatest concern (high sensitivities), due to a longer ocean response time, which may have bearing on wait-and-see versus precautionary mitigation policies. Learning rates are optimistic in presuming the availability of whole ocean heat data but pessimistic by using simple aggregated metrics and model physics.

Probabilistic framework for assessing the ice sheet contribution to sea level change

Previous sea level rise (SLR) assessments have excluded the potential for dynamic ice loss over much of Greenland and Antarctica, and recently proposed “upper bounds” on Antarctica’s 21st-century SLR contribution are derived principally from regions where present-day mass loss is concentrated (basin 15, or B15, drained largely by Pine Island, Thwaites, and Smith glaciers). Here, we present a probabilistic framework for assessing the ice sheet contribution to sea level change that explicitly accounts for mass balance uncertainty over an entire ice sheet. Applying this framework to Antarctica, we find that ongoing mass imbalances in non-B15 basins give an SLR contribution by 2100 that: (i) is comparable to projected changes in B15 discharge and Antarctica’s surface mass balance, and (ii) varies widely depending on the subset of basins and observational dataset used in projections. Increases in discharge uncertainty, or decreases in the exceedance probability used to define an upper bound, increase the fractional contribution of non-B15 basins; even weak spatial correlations in future discharge growth rates markedly enhance this sensitivity. Although these projections rely on poorly constrained statistical parameters, they may be updated with observations and/or models at many spatial scales, facilitating a more comprehensive account of uncertainty that, if implemented, will improve future assessments.

CMIP5 temperature biases and 21st century warming around the Antarctic coast

ABSTRACT Projections of ice-sheet mass balance require regional ocean warming projections derived from atmosphere-ocean general circulation models (AOGCMs). However, the coarse resolution of AOGCMs: (1) may lead to systematic or AOGCM-specific biases and (2) makes it difficult to identify relevant water masses. Here, we employ a large-scale metric of Antarctic Shelf Bottom Water (ASBW) to investigate circum-Antarctic temperature biases and warming projections in 19 different Coupled Model Intercomparison Project Phase 5 (CMIP5) AOGCMs forced with two different ‘representative concentration pathways’ (RCPs). For high-emissions RCP 8.5, the ensemble mean 21st century ASBW warming is 0.66, 0.74 and 0.58°C for the Amundsen, Ross and Weddell Seas (AS, RS and WS), respectively. RCP 2.6 ensemble mean projections are substantially lower: 0.21, 0.26, and 0.19°C. All distributions of regional ASBW warming are positively skewed; for RCP 8.5, four AOGCMs project warming of greater than 1.8°C in the RS. Across the ensemble, there is a strong, RCP-independent, correlation between WS and RS warming. AS warming is more closely linked to warming in the Southern Ocean. We discuss possible physical mechanisms underlying the spatial patterns of warming and highlight implications of these results on strategies for forcing ice-sheet mass balance projections.

Thermodynamic properties and correlation functions of Ar films on the surface of a bundle of nanotubes

We employ canonical Monte Carlo simulations to explore the properties of an Ar film adsorbed on the external surface of a bundle of carbon nanotubes. The study is concerned primarily with three properties: specific heat

Global Sensitivity of Simulated Water Balance Indicators Under Future Climate Change in the Colorado Basin

c(T)