History matching for exploring and reducing climate model parameter space using observations and a large perturbed physics ensemble

We apply an established statistical methodology called history matching to constrain the parameter space of a coupled non-flux-adjusted climate model (the third Hadley Centre Climate Model; HadCM3) by using a 10,000-member perturbed physics ensemble and observational metrics. History matching uses emulators (fast statistical representations of climate models that include a measure of uncertainty in the prediction of climate model output) to rule out regions of the parameter space of the climate model that are inconsistent with physical observations given the relevant uncertainties. Our methods rule out about half of the parameter space of the climate model even though we only use a small number of historical observations. We explore 2 dimensional projections of the remaining space and observe a region whose shape mainly depends on parameters controlling cloud processes and one ocean mixing parameter. We find that global mean surface air temperature (SAT) is the dominant constraint of those used, and that the others provide little further constraint after matching to SAT. The Atlantic meridional overturning circulation (AMOC) has a non linear relationship with SAT and is not a good proxy for the meridional heat transport in the unconstrained parameter space, but these relationships are linear in our reduced space. We find that the transient response of the AMOC to idealised CO₂ forcing at 1 and 2 % per year shows a greater average reduction in strength in the constrained parameter space than in the unconstrained space. We test extended ranges of a number of parameters of HadCM3 and discover that no part of the extended ranges can by ruled out using any of our constraints. Constraining parameter space using easy to emulate observational metrics prior to analysis of more complex processes is an important and powerful tool. It can remove complex and irrelevant behaviour in unrealistic parts of parameter space, allowing the processes in question to be more easily studied or emulated, perhaps as a precursor to the application of further relevant constraints.     

The seasonal footprinting mechanism in large ensemble simulations of the second generation Canadian earth system model: uncertainty due to internal climate variability

Climate Dynamics - Previous studies indicated that the wintertime North Pacific Oscillation (NPO) could exert marked impacts on the following winter El Niño-Southern Oscillation (ENSO) via the...     

An approximate energy cycle for inter-member variability in ensemble simulations of a regional climate model

The presence of internal variability (IV) in ensembles of nested regional climate model (RCM) simulations is now widely acknowledged in the community working on dynamical downscaling. IV is defined as the inter-member spread between members in an ensemble of simulations performed by a given RCM driven by identical lateral boundary conditions (LBC), where different members are being initialised at different times. The physical mechanisms responsible for the time variations and structure of such IV have only recently begun to receive attention. Recent studies have shown empirical evidence of a close parallel between the energy conversions associated with the time fluctuations of IV in ensemble simulations of RCM and the energy conversions taking place in weather systems. Inspired by the classical work on global energetics of weather systems, we sought a formulation of an energy cycle for IV that would be applicable for limited-area domain. We develop here a novel formalism based on local energetics that can be applied to further our understanding IV. Prognostic equations for ensemble-mean kinetic energy and available enthalpy are decomposed into contributions due to ensemble-mean variables (EM) and those due to deviations from the ensemble mean (IV). Together these equations constitute an energy cycle for IV in ensemble simulations of RCM. Although the energy cycle for IV was developed in a context entirely different from that of energetics of weather systems, the exchange terms between the various reservoirs have a rather similar mathematical form, which facilitates some interpretations of their physical meaning.     

Convective and stratiform precipitation characteristics in an ensemble of regional climate model simulations

We apply a recently proposed algorithm for disaggregating observed precipitation data into predominantly convective and stratiform, and evaluate biases in characteristics of parameterized convective (subgrid) and stratiform (large-scale) precipitation in an ensemble of 11 RCM simulations for recent climate in Central Europe. All RCMs have a resolution of 25 km and are driven by the ERA-40 reanalysis. We focus on mean annual cycle, proportion of convective precipitation, dependence on altitude, and extremes. The results show that characteristics of total precipitation are often better simulated than are those of convective and stratiform precipitation evaluated separately. While annual cycles of convective and stratiform precipitation are reproduced reasonably well in most RCMs, some of them consistently and substantially overestimate or underestimate the proportion of convective precipitation throughout the year. Intensity of convective precipitation is underestimated in all RCMs. Dependence on altitude is also simulated better for stratiform and total precipitation than for convective precipitation, for which several RCMs produce unrealistic slopes. Extremes are underestimated for convective precipitation while they tend to be slightly overestimated for stratiform precipitation, thus resulting in a relatively good reproduction of extremes in total precipitation amounts. The results suggest that the examined ensemble of RCMs suffers from substantial deficiencies in reproducing precipitation processes and support previous findings that climate models’ errors in precipitation characteristics are mainly related to deficiencies in the representation of convection.     

A perturbed parameter ensemble of HadGEM3-GC3.05 coupled model projections: part 2: global performance and future changes

This paper provides a quantitative assessment of large-scale features in a perturbed parameter ensemble (PPE) of Met Office Unified Model HadGEM-GC3.05 in coupled global historical and future simulations. The main motivation for the simulations is to provide a major component of the UK Climate Projections 2018 (UKCP18), but they will also be used to make worldwide projections and inform future model development. Initially, a 25-member PPE, with 25 different parameter combinations, was simulated. Five members were subsequently dropped because either their simulated climate was unrealistically cool by 1970 or they suffered from numerical instabilities. The remaining 20 members were evaluated after completing the historical phase (1900–2005) against 13 separately selected Climate Model Intercomparison Project Phase 5 (CMIP5) models, and five more members were dropped. The final product is a combined projection system of 15 PPE members and 13 CMIP5 models, which has a number of benefits. In particular, the range of outcomes available from the combined set of 28 is often larger than from either of the two constituent ensembles, thus providing users with a more complete picture of plausible impacts. Here we mainly describe the evaluation process of the 20 PPE members. We evaluate biases in a number of important properties of the global coupled system, including assessment of climatological averages, coupled modes of internal variability and historical and future changes. The parameter combinations yielded plausible yet diverse atmosphere and ocean model behaviours. The range of global temperature changes is narrow, largely driven by use of different CO₂ pathways. The range of global warming is seemingly not linked to range of feedbacks estimated from atmosphere-only runs, though we caution that the range of the latter is narrow relative to CMIP5, and therefore this result is not unexpected. This is the second of two papers describing the generation of the PPE for UKCP18 projections. Part 1 (Sexton et al. 2021) describes the selection of 25 parameter combinations of 47 atmosphere and land surface parameters, using a set of cheap atmosphere-only runs at a coarser resolution from nearly 3000 samples of parameter space.

     

A perturbed parameter ensemble of HadGEM3-GC3.05 coupled model projections: part 1: selecting the parameter combinations

This is the first of two papers that describe the generation of a 25-member perturbed parameter ensemble (PPE) of high-resolution, global coupled simulations for the period 1900–2100, using CMIP5 historical and RCP8.5 emissions. Fifteen of these 25 coupled simulations now form a subset of the global projections provided for the UK Climate Projections 2018 (UKCP18). This first paper describes the selection of 25 variants (combinations of 47 parameters) using a set of cheap, coarser-resolution atmosphere-only simulations from a large sample of nearly 3000 variants. Retrospective 5-day weather forecasts run at climate resolution, and simulations of 2004–2009 with prescribed SST and sea ice are evaluated to filter out poor performance. We opted for a single design choice and sensitivity tests were done after the PPE was generated to demonstrate the effect of design choices on the filtering. Given our choice, only 38 of the parameter combinations were found to have acceptable performance at this stage. Idealised atmosphere-only simulations were then used to select the subset of 25 members that were as diverse as possible in terms of their CO₂ and aerosol forcing, and their response to warmer SSTs. Using our parallel set of atmosphere-only and coupled PPEs (the latter from paper 2), we show that local biases in the atmosphere-only experiments are generally informative about the biases in the coupled PPE. Biases in radiative fluxes and cloud amounts are strongly informative for most regions, whereas this is only true for a smaller fraction of the globe for precipitation and dynamical variables. Therefore, the cheap experiments are an affordable way to search for promising parameter combinations but have limitations.

     

Climate sensitivity estimated from ensemble simulations of glacial climate

The concentration of greenhouse gases (GHGs) in the atmosphere continues to rise, hence estimating the climate system’s sensitivity to changes in GHG concentration is of vital importance. Uncertainty in climate sensitivity is a main source of uncertainty in projections of future climate change. Here we present a new approach for constraining this key uncertainty by combining ensemble simulations of the last glacial maximum (LGM) with paleo-data. For this purpose we used a climate model of intermediate complexity to perform a large set of equilibrium runs for (1) pre-industrial boundary conditions, (2) doubled CO₂ concentrations, and (3) a complete set of glacial forcings (including dust and vegetation changes). Using proxy-data from the LGM at low and high latitudes we constrain the set of realistic model versions and thus climate sensitivity. We show that irrespective of uncertainties in model parameters and feedback strengths, in our model a close link exists between the simulated warming due to a doubling of CO₂, and the cooling obtained for the LGM. Our results agree with recent studies that annual mean data-constraints from present day climate prove to not rule out climate sensitivities above the widely assumed sensitivity range of 1.5–4.5°C (Houghton et al. 2001). Based on our inferred close relationship between past and future temperature evolution, our study suggests that paleo-climatic data can help to reduce uncertainty in future climate projections. Our inferred uncertainty range for climate sensitivity, constrained by paleo-data, is 1.2–4.3°C and thus almost identical to the IPCC estimate. When additionally accounting for potential structural uncertainties inferred from other models the upper limit increases by about 1°C.     

Rain-on-Snow events in Japan as projected by a large ensemble of regional climate simulations

Climate Dynamics - Rain-on-Snow (ROS) events can cause severe snowmelt hazards such as river flooding, avalanches, and landslides that have significant impacts on various sectors. The influence of...     

ENSO forcing of the Northern Hemisphere climate in a large ensemble of model simulations based on a very long SST record

The January–March (JFM) climate response of the Northern Hemisphere atmosphere to observed sea surface temperature (SST) anomalies for the period 1855–2002 is analysed from a 35-member ensemble made with SPEEDY, an atmospheric general circulation model (AGCM) of intermediate complexity. The model was run at the T30-L8 resolution, and initial conditions and the early stage of model runs differ among ensemble members in the definition of tropical diabatic heating. SST anomalies in the Niño3.4 region were categorised into five classes extending from strong cold to strong warm. Composites based on such a categorisation enabled an analysis of the influence of the tropical Pacific SST on the Northern Hemisphere atmospheric circulation with an emphasis on the Pacific-North America (PNA) and the North Atlantic-Europe (NAE) regions. As expected, the strongest signal was detected over the PNA region. An “asymmetry” in the model response was found for the opposite polarity of the Niño3.4 index; however, this asymmetry stems mainly from the difference in the amplitude of model response rather than from the phase shift between responses to warm and cold El Niño-Southern Oscillation (ENSO) events. The extratropical signal associated with warm ENSO events was found to be stronger than that related to cold events. The results also reveal that, for the PNA region, the amplitude of the response is positively correlated with the strength of ENSO, irrespective of the sign of ENSO. With almost no phase shift between model responses to El Niño and La Niña, the linear component of the response is much stronger than the non-linear component. Although the model climate response over the NAE region is much weaker than that over the PNA region, some striking similarities with the PNA are found. Both sea level pressure and precipitation responses are positively correlated with the strength of ENSO. This is not true for the 200-hPa geopotential heights, and no plausible explanation for such a result could be offered. An appreciable linear component in model response over the NAE was also found. The model results over the NAE region agree reasonably well with observational studies. An additional analysis of the remote atmospheric response to very weak ENSO forcing (defined from the interval between 0.5σ and 1.0σ of the interannual variance) was also carried out. A discernible model response in the Northern Hemisphere to such a weak SST forcing was found.     

Frequency distributions of transient regional climate change from perturbed physics ensembles of general circulation model simulations

Large ensembles of coupled atmosphere–ocean general circulation model (AOGCM) simulations are required to explore modelling uncertainty and make probabilistic predictions of future transient climate change at regional scales. These are not yet computationally feasible so we have developed a technique to emulate the response of such an ensemble by scaling equilibrium patterns of climate change derived from much cheaper “slab” model ensembles in which the atmospheric component of an AOGCM is coupled to a mixed-layer ocean. Climate feedback parameters are diagnosed for each member of a slab model ensemble and used to drive an energy balance model (EBM) to predict the time-dependent response of global surface temperature expected for different combinations of uncertain AOGCM parameters affecting atmospheric, land and sea-ice processes. The EBM projections are then used to scale normalised patterns of change derived for each slab member, and hence emulate the response of the relevant atmospheric model version when coupled to a dynamic ocean, in response to a 1% per annum increase in CO₂. The emulated responses are validated by comparison with predictions from a 17 member ensemble of AOGCM simulations, constructed from variants of HadCM3 using the same parameter combinations as 17 members of the slab model ensemble. Cross-validation permits estimation of the spatial and temporal dependence of emulation error, and also allows estimation of a correction field to correct discrepancies between the scaled equilibrium patterns and the transient response, reducing the emulation error. Emulated transient responses and their associated errors are obtained from the slab ensemble for 129 pseudo-HadCM3 versions containing multiple atmospheric parameter perturbations. These are combined to produce regional frequency distributions for the transient response of annual surface temperature change and boreal winter precipitation change. The technique can be extended to any surface climate variable demonstrating a scaleable, approximately linear response to forcing.     

Analysis of precipitation extremes in an ensemble of transient regional climate model simulations for the Rhine basin

A non-stationary index-flood model was used to analyse the 1-day summer and 5-day winter precipitation maxima in the Rhine basin in an ensemble of 15 transient regional climate model (RCM) simulations. It is assumed that the seasonal precipitation maxima follow a generalized extreme value (GEV) distribution with time varying parameters. The index-flood assumption implies that the dispersion coefficient (the ratio of the scale and the location parameters) and the shape parameter are constant over predefined regions, while the location parameter varies within these regions. A comparison with the estimates from gridded observations shows that these GEV parameters are too large in the summer season, while there is a large overestimation of the location parameter and underestimation of the dispersion coefficient in winter. However, a large part of the biases in the summer season might be due to the low number of stations used for gridding the observations. Though there is considerable variation in the changes of the extreme value distributions among the RCM simulations, common tendencies can be identified. In summer, large quantiles increase as a consequence of an increase of the dispersion coefficient, while there is almost no change of low quantiles. In winter, low quantiles increase because of an increase of the location parameter. This effect is, however, counterbalanced by a decrease of the shape parameter in most RCM simulations, resulting in only a slight increase of large quantiles. Departures from the assumed index-flood model were observed in the Alpine region in the south of the basin. This is due to the strong spatial heterogeneity in the dispersion coefficient in a number of RCM simulations and a significant altitude dependence of the trend in the location parameter in winter in five RCM simulations.     

An overview of decadal climate predictability in a multi-model ensemble by climate model MIROC

Decadal climate predictability is examined in hindcast experiments by a multi-model ensemble using three versions of the coupled atmosphere-ocean model MIROC. In these hindcast experiments, initial conditions are obtained from an anomaly assimilation procedure using the observed oceanic temperature and salinity with prescribed natural and anthropogenic forcings on the basis of the historical data and future emission scenarios in the Intergovernmental Panel of Climate Change. Results of the multi-model ensemble in our hindcast experiments show that predictability of surface air temperature (SAT) anomalies on decadal timescales mostly originates from externally forced variability. Although the predictable component of internally generated variability has considerably smaller SAT variance than that of externally forced variability, ocean subsurface temperature variability has predictive skills over almost a decade, particularly in the North Pacific and the North Atlantic where dominant signals associated with Pacific decadal oscillation (PDO) and the Atlantic multidecadal oscillation (AMO) are observed. Initialization enhances the predictive skills of AMO and PDO indices and slightly improves those of global mean temperature anomalies. Improvement of these predictive skills in the multi-model ensemble is higher than that in a single-model ensemble.     

Future projections of Indian Summer Monsoon under multiple RCPs using a high resolution global climate model multiforcing ensemble simulations

Climate Dynamics - Present-day simulations (1983–2003) of a global climate model of 60-km resolution with three deep convection schemes are analysed to find the best scheme for simulation of...     

Comparison of four ensemble methods combining regional climate simulations over Asia

A number of uncertainties exist in climate simulation because the results of climate models are influenced by factors such as their dynamic framework, physical processes, initial and driving fields, and horizontal and vertical resolution. The uncertainties of the model results may be reduced, and the credibility can be improved by employing multi-model ensembles. In this paper, multi-model ensemble results using 10-year simulations of five regional climate models (RCMs) from December 1988 to November 1998 over Asia are presented and compared. The simulation results are derived from phase II of the Regional Climate Model Inter-comparison Project (RMIP) for Asia. Using the methods of the arithmetic mean, the weighted mean, multivariate linear regression, and singular value decomposition, the ensembles for temperature, precipitation, and sea level pressure are carried out. The results show that the multi-RCM ensembles outperform the single RCMs in many aspects. Among the four ensemble methods used, the multivariate linear regression, based on the minimization of the root mean square errors, significantly improved the ensemble results. With regard to the spatial distribution of the mean climate, the ensemble result for temperature was better than that for precipitation. With an increasing number of models used in the ensembles, the ensemble results were more accurate. Therefore, a multi-model ensemble is an efficient approach to improve the results of regional climate simulations.     

Carbon cycle–climate feedback sensitivity to parameter changes of a zero-dimensional terrestrial carbon cycle scheme in a climate model of intermediate complexity

Summary A series of sensitivity runs have been performed with a coupled climate–carbon cycle model. The climatic component consists of the climate model of intermediate complexity IAP RAS CM. The carbon cycle component is formulated as a simple zero-dimensional model. Its terrestrial part includes gross photosynthesis, and plant and soil respirations, depending on temperature via Q ₁₀-relationships (Lenton, 2000). Oceanic uptake of anthropogenic carbon is formulated is a bi-linear function of tendencies of atmospheric concentration of CO₂ and globally averaged annual mean sea surface temperature. The model is forced by the historical industrial and land use emissions of carbon dioxide for the second half of the 19th and the whole of the 20th centuries, and by the emission scenario SRES A2 for the 21st century. For the standard set of the governing parameters, the model realistically captures the main features of the Earth’s observed carbon cycle. A large number of simulations have been performed, perturbing the governing parameters of the terrestrial carbon cycle model. In addition, the climate part is perturbed, either by zeroing or artificially increasing the climate model sensitivity to the doubling of the atmospheric CO₂ concentration. Performing the above mentioned perturbations, it is possible to mimic most of the range found in the C4MIP simulations. In this way, a wide range of the climate–carbon cycle feedback strengths is obtained, differing even in the sign of the feedback. If the performed simulations are subjected to the constraints of a maximum allowed deviation of the simulated atmospheric CO₂ concentration (pCO_2(a)) from the observed values and correspondence between simulated and observed terrestrial uptakes, it is possible to narrow the corresponding uncertainty range. Among these constraints, considering pCO_2(a) and uptakes are both important. However, the terrestrial uptakes constrain the simulations more effectively than the oceanic ones. These constraints, while useful, are still unable to rule out both extremely strong positive and modest negative climate–carbon cycle feedback.     

On the linear response of tropical African climate to SST changes deduced from regional climate model simulations

Summary Previous studies have highlighted the crucial role of sea surface temperature (SST) anomalies in the tropical Atlantic region in forcing the summer monsoon rainfall over subsaharan West Africa. Understanding the physical processes, relating SST variations to changes in the amount and distribution of African rainfall, is a key factor in improving weather and climate forecasts in this highly vulnerable region. Here, we present sensitivity experiments from a regional climate model with prescribed warmer tropical SSTs, according to enhanced greenhouse conditions at the end of the 21st century. This dynamical downscaling approach provides information about the nonlinear response of the atmosphere to oceanic heating. It has been suggested that the response is at least partly accounted for by the linear theory of tropical dynamics, involving a Kelvin and Rossby wave response to a tropical heat source. We compute the major modes of the linear Matsuno-Gill model for geopotential height and horizontal wind components and project the simulated response patterns onto these linear modes, in order to evaluate to which extent the simple linear theory may explain the SST-induced climate anomalies over Africa. A multivariate Hotelling T² test is used to evaluate whether these anomalies are statistically significant. Forcing the regional climate model by warmer SSTs leads to substantial climate anomalies over tropical Africa: Rainfall is increases over the Guinea Coast region (GCR) and tropical East Africa, but decreases over the Congo Basin and the Sahel Zone (SHZ). At the 850 hPa level, a trough develops over southern West Africa and the Gulf of Guinea, and is associated with stronger surface wind convergence over the GCR. These changes in the atmospheric dynamics strongly project onto the leading modes of the linear Matsuno-Gill model at various zonal wave numbers. The corresponding atmospheric heating pattern is highly reminiscent of the simulated nonlinear model reponse. The T² test statistics reveal that the SST forcing induces a statistically significant climate anomaly over tropical Africa if the climate state vector is reduced by projecting the simulated data onto the leading 10 linear modes. It is also shown that the linear response prevails in a long-term simulation with more realistic lower and lateral boundary conditions. Thus, linear tropical dynamics are assumed to be a major physical process on the ground of the prominent SST-African rainfall relationship.     

A multi-physics ensemble of present-day climate regional simulations over the Iberian Peninsula

This work assesses the influence of the model physics in present-day regional climate simulations. It is based on a multi-phyiscs ensemble of 30-year long MM5 hindcasted simulations performed over a complex and climatically heterogeneous domain as the Iberian Peninsula. The ensemble consists of eight members that results from combining different parametrization schemes for modeling the Planetary Boundary Layer, the cumulus and the microphysics processes. The analysis is made at the seasonal time scale and focuses on mean values and interannual variability of temperature and precipitation. The objectives are (1) to evaluate and characterize differences among the simulations attributable to changes in the physical options of the regional model, and (2) to identify the most suitable parametrization schemes and understand the underlying mechanisms causing that some schemes perform better than others. The results confirm the paramount importance of the model physics, showing that the spread among the various simulations is of comparable magnitude to the spread obtained in similar multi-model ensembles. This suggests that most of the spread obtained in multi-model ensembles could be attributable to the different physical configurations employed in the various models. Second, we obtain that no single ensemble member outperforms the others in every situation. Nevertheless, some particular schemes display a better performance. On the one hand, the non-local MRF PBL scheme reduces the cold bias of the simulations throughout the year compared to the local Eta model. The reason is that the former simulates deeper mixing layers. On the other hand, the Grell parametrization scheme for cumulus produces smaller amount of precipitation in the summer season compared to the more complex Kain-Fritsch scheme by reducing the overestimation in the simulated frequency of the convective precipitation events. Consequently, the interannual variability of precipitation (temperature) diminishes (increases), which implies a better agreement with the observations in both cases. Although these features improve in general the accuracy of the simulations, controversial nuances are also highlighted.     

Transient responses of North-American grasslands to changes in climate

Our objective was to evaluate the transient responses of grasslands in the central grassland region of North America to changes in climate. We used an individual plant-based gap dynamics simulation model (STEPPE-GP) linked with a soil water model (SOILWAT) to evaluate the effects of changes in climate on the composition and structure of grassland vegetation. Five functional types of plants were simulated based upon lifeform, physiology, and rooting distribution with depth. C₃ and C₄ perennial grasses with either a shallow or deep rooting distribution, and deeply rooted C₃ shrubs were simulated under current climatic conditions and under a GFDL climate change scenario for nine sites representative of the temperature and precipitation regimes in the grassland region.Although vegetation at the sites responded differently to climate change, shifts in functional types occurred within 40 years of the start of the climate change. C₄ grasses increased in dominance or importance at all sites with a change in climate, primarily as a result of increases in temperature in all months at all sites. The coolest sites that arc currently dominated by C₃ grasses were predicted to shift to a dominance by C₄ grasses, whereas sites that are currently dominated by C₄ grasses had an increase in importance of this functional type with a change in climate. Current annual temperature was the best predictor of changes in C₃ biomass, and C₃ and C₄ biomass combined; current annual precipitation was the best predictor of changes in C₄ biomass. These predicted shifts in dominance and importance of C₃ versus C₄ grasses would have important implications for the management of natural grasslands as well as the cultivation of crops in the central grassland region.