Global warming results from anthropogenic greenhouse gas emissions which upset the delicate balance between the incoming sunlight, and the reflected and emitted radiation from Earth. The imbalance leads to energy accumulation in the atmosphere, oceans and land, and melting of the cryosphere, resulting in increasing temperatures, rising sea levels, and more extreme weather around the globe. Despite the fundamental role of the energy imbalance in regulating the climate system, as known to humanity for more than two centuries, our capacity to observe it is rapidly deteriorating as satellites are being decommissioned.

Climate feedbacks over the historical period have been investigated in a 47 member ensemble of atmosphere-ocean general circulation model (AOGCM) simulations. Here, the model response to historical forcing, as well as individual forcing constituents such as aerosol and greenhouse gases separately, has been analysed. The analysis addresses the cause of differing feedbacks across the ensemble, the disparity between feedbacks seen in these AOGCM simulations and atmosphere-only GCM (AGCM) experiments prescribed with observed SSTs, and the different forcing efficacies of the respective forcing agents. It is found that much of the spread in feedbacks across ensemble members and different experiments can be explained through varying SST patterns. The level of polar amplification is shown to strongly control the amount of sea ice melt per degree of global warming, a mechanism responsible for the spread in shortwave clear sky feedback and a large contributor to the different forcing efficacies seen across the different forcing agents. The spread in feedbacks across the historical ensemble is also shown to be caused by both the level of tropical surface temperature warming, due to its influence on longwave clear sky feedback, and the response of cloud feedbacks to local surface temperatures and large scale changes in tropospheric temperature. It is also shown that each of these processes discussed are partly responsible for the disparity in feedbacks seen between AOGCM simulations and AGCM experiments prescribed with observed SSTs.

Climate feedbacks over the historical period (1850–2014) have been investigated in large ensembles of historical, hist-ghg, hist-aer, and hist-nat experiments, with 47 members for each experiment. Across the historical ensemble with all forcings, a range in estimated Effective Climate Sensitivity (EffCS) between approximately 3–6 K is found, a considerable spread stemming solely from initial condition uncertainty. The spread in EffCS is associated with varying Sea Surface Temperature (SST) patterns seen across the ensemble due to their influence on different feedback processes. For example, the level of polar amplification is shown to strongly control the amount of sea ice melt per degree of global warming. This mechanism is responsible for the large spread in shortwave clear-sky feedbacks and is the main contributor to the different forcing efficacies seen across the different forcing agents, although in HadGEM3-GC3.1-LL these differences in forcing efficacy are shown to be small. The spread in other feedbacks is also investigated, with the level of tropical SST warming shown to strongly control the longwave clear-sky feedbacks, and the local surface-air-temperatures and large scale tropospheric temperatures shown to influence cloud feedbacks. The metrics used to understand the spread in feedbacks can also help to explain the disparity between feedbacks seen in the historical experiment simulations and a more accurate modeled estimate of the feedbacks seen in the real world derived from an atmosphere-only experiment prescribed with observed SSTs (termed amip-piForcing).

The Last Interglacial (LIG), a warmer period 130,000–116,000 years before present, is a potential analogue for future climate change. Stronger LIG summertime insolation at high northern latitudes drove Arctic land summer temperatures 4–5 °C higher than in the pre-industrial era. Climate model simulations have previously failed to capture these elevated temperatures, possibly because they were unable to correctly capture LIG sea-ice changes. Here, we show that the latest version of the fully coupled UK Hadley Center climate model (HadGEM3) simulates a more accurate Arctic LIG climate, including elevated temperatures. Improved model physics, including a sophisticated sea-ice melt-pond scheme, result in a complete simulated loss of Arctic sea ice in summer during the LIG, which has yet to be simulated in past generations of models. This ice-free Arctic yields a compelling solution to the long-standing puzzle of what drove LIG Arctic warmth and supports a fast retreat of future Arctic summer sea ice. Arctic climate in the Last Interglacial (LIG)—a warm period 130,000–116,000 years ago—is poorly simulated by modern climate models. A model with improved sea-ice melt-pond physics reproduces LIG Arctic temperatures, suggests an ice-free Arctic during this period and predicts the same by 2035.

We compare the radiative feedbacks resulting from a uniform warming and cooling of sea surface temperatures by 4 K in an ensemble of global climate models. The global-mean net feedback is less stabilising in response to warming in all nine models. This is primarily due to a stronger tropical water vapour feedback, with a smaller contribution from the shortwave cloud feedback. The zonal-mean feedbacks are similarly robust across the ensemble. In the extra-tropics, more positive shortwave cloud feedback under warming is associated with further poleward migration of the mean Southern Hemisphere jet latitude in some models. However, additional experiments with an aquaplanet version of the HadGEM3 model suggest that the asymmetry of the jet shift is not driving that in the cloud feedbacks at these latitudes. In the tropics, stronger water vapour feedback under warming is offset by a weaker shortwave cloud feedback. The result is that the ensemble spread in the differences between the global feedbacks under warming and cooling is mainly determined by their differences in the tropics. The spatial distribution of the feedbacks largely reflects the zonal mean behaviour, although there is considerable intermodel variation in the regional cloud feedbacks, particularly in the tropical shortwave cloud feedback. Comparison with CO2- and solar-forced coupled experiments suggests that the global-mean longwave cloud feedback is nearly invariant to warming and cooling, regardless of the nature of the forcing. The shortwave cloud feedback is generally more positive under warming in the coupled models, consistent with the uniform SST perturbation experiments.

Abstract Water vapor in the Arctic can be attributed to local evaporation and transport from mid‐latitudes. Here we investigate the response of Arctic water vapor to declining sea ice in a climate model, using an idealized warming scenario in which the Arctic becomes ice free year‐round. We find distinct monthly patterns in water vapor change with warming, with a strong influence of sea ice loss on changes to evaporation, except in summer when transports dominate. Before the Arctic becomes ice‐free, total column water increases at a rate slower than that indicated by the Clausius‐Clapeyron relationship; once ice‐free, total column water increases at a rate equal to Clausius‐Clapeyron. The exception is summer when the increase is initially faster then becomes slower when ice‐free. The boundary layer structure changes, under global warming, from stratified to well mixed in all seasons except summer, which remains largely unchanged throughout the ice‐free transition.

The emerging signal of climate change is now clearly evident in Global Navigation Satellite System (GNSS) radio occultation (RO) data, matching predictions made by climate models 15 years ago. The observed RO trends represent well-understood responses to global warming, in particular the widespread cooling of the lower stratosphere and warming of the troposphere. This demonstrates the value of RO measurements for climate monitoring, consistent with their information content and their use in both weather forecasting and atmospheric reanalyses.

The effective climate sensitivity (EffCS) of models in the Coupled Model Intercomparison Project 6 (CMIP6) has increased relative to CMIP5. Consequently, using CMIP6 models tends to lead to larger projections of global mean surface air temperature (GSAT) increase for a given emissions scenario. The effect of increased EffCS on projections of global mean sea-level (GMSL) change however, has so far only been studied using a reduced complexity model. Here, we explore the implications of increased EffCS in CMIP6 models for GMSL change projections in 2100 for three emissions scenarios: SSP5-8.5, SSP2-4.5 and SSP1-2.6. Whereas CMIP6 projections of GSAT change are substantially higher than in CMIP5, projections of global mean thermal expansion (GTE) are only slightly higher. We use these projections as input to construct projections of GMSL change, using the Monte Carlo approach of IPCC AR5. Isolating the impact of the CMIP6 simulations using consistent methods is an important step to ensure traceability to past IPCC projections of global and regional sea-level change. The resulting 95th percentile of projected GMSL change at 2100 is only 3-7 cm higher for CMIP6 than for CMIP5, depending on the emissions scenario. Projected rates of GMSL rise around 2100 increase more strongly from CMIP5 to CMIP6, though, implying more pronounced differences beyond 2100 and greater committed sea-level rise. GMSL change in 2100 is accurately predicted by time-integrated temperature change and therefore mitigation requires early reduction of emissions. We also find that the 95th percentile projections based on individual CMIP6 models can differ as much as 51 cm and that the 5-95% range of projected GMSL change for individual CMIP6 models can be substantially outside of the 5-95% range of the CMIP6 multi-model ensemble. Thus, through subsetting the CMIP6 ensemble using EffCS, a choice can be made between characterizing the central part of the probability distribution and more comprehensively sampling the high end of the GMSL projection space, which is relevant to risk-averse stakeholders. Our results show this may substantially alter ensemble projections, underlining the need to constrain EffCS in global climate models in order to reduce uncertainty in sea-level projections.

Aerosol forcing remains the most uncertain component of the total climate forcing on the Earth system. RFMIP and AerChemMIP contained experiments that allow us to determine time-slice present day (2014 minus 1850) from 17 CMIP6 models, and transient (1850 to 2014, or 2100) aerosol forcing from 11 models. In CMIP6, aerosol present-day aerosol forcing is -1.01 (full range -1.37 to -0.63) W m-2, a range considerably narrower than comprehensive assessments of aerosol forcing from multiple lines of evidence such as AR5 (-1.9 to -0.1 W m-2) or Bellouin et al. 2020 (-2.0 to -0.35 W m-2). The transient experiments also show a diversity in time histories, with most models showing a peak negative aerosol forcing at some time between 1975 and 2010, and recent trends varying from strongly recovering to slightly strengthening aerosol forcing. Models that were run to 2100 under SSP2-4.5 all show a projected weakening aerosol forcing. By fitting a simple relationship of how globally integrated emissions of black carbon, organic carbon and SO2 relate to effective radiative forcing from aerosol-radiation interactions (ERFari) and aerosol-cloud interactions (ERFaci), an emissions to forcing relationship can be determined for these 11 RFMIP and AerChemMIP models. Using a 100,000 member Monte Carlo ensemble of historical aerosol time series, where coefficients are drawn from these model-derived distributions, and total 1850 to 2014 aerosol forcing is taken from the wider distributions of Bellouin et al. (2020), we create a best estimate historical time series for aerosol forcing (with uncertainty) that is constrained to historical warming and observed ocean heat uptake using a simple climate model. This method can also be used to predict aerosol forcing from future emissions scenarios, such as the SSPs and those derived from integrated assessment models, and provides estimates of the likely ranges for equilibrium climate sensitivity and transient climate response based on the historical aerosol forcing.

An observationally constrained time series of historical aerosol effective radiative forcing (ERF) from 1750 to 2019 is developed in this study. We find that the time history of aerosol ERFs diagnosed in CMIP6 models exhibits considerable variation and explore how the time history of aerosol forcing influences the probability distributions of present‐day aerosol forcing and emergent metrics such as climate sensitivity. Using a simple energy balance model, trained on CMIP6 climate models and constrained by observed near‐surface warming and ocean heat uptake, we derive estimates for the historical aerosol forcing. We find 2005–2014 mean aerosol ERF to be −1.1 (−1.8 to −0.5) W m −2 relative to 1750. Assuming recently published historical emissions from fossil fuel and industrial sectors and biomass burning emissions from SSP2‐4.5, aerosol ERF in 2019 is −0.9 (−1.5 to −0.4) W m −2 . There is a modest recovery in aerosol forcing (+0.025 W m −2 decade −1 ) between 1980 and 2014. This analysis also gives a 5%–95% range of equilibrium climate sensitivity of 1.8°C –5.1°C (best estimate 3.1°C) with a transient climate response of 1.2°C –2.6°C (best estimate 1.8°C).

A large parameter-perturbed ensemble (PPE) of the Met Office climate model is used to explore the relationship between radiative feedbacks and the present-day simulation of the associated physical processes. We highlight three tropical regimes (deep convection over ocean and land, and marine stratocumulus) in which the same set of processes drives the present-day simulation of clouds and their feedbacks. In each case, the amount of the dominant cloud types reduces in response to warming and the reduction is approximately proportional to the amount simulated in the present day. In deep convective regions, convective process parameters lead the spread among multiple contributing processes, with vegetation processes contributing as well for the land regions. Multiple parameters, such as boundary layer processes, drive stratocumulus regions. However, the low-thick clouds are systematically overestimated, suggesting a structural error in their process representations which would limit the efficacy of the constraint. The influence of convection is largely confined to the tropical deep convective regions in the present day but extends to mid-latitudes under warming. Because of this, contributing processes to the spread in the present-day and the response are different in the extra-tropics, making it much more difficult to establish links between the present-day and the feedback within the region. This suggests that identifying a constraint on convective processes in the tropics for the present-day simulations could constrain both the tropical feedbacks and feedbacks in the extra-tropics. A parameter representing deep-convective entrainment links the present-day tropical mean high cloud and clear-sky longwave flux to their feedbacks in our model, suggesting a potential process constraint from the observations. Understanding and improving the detailed processes controlling feedbacks is ultimately only possible in individual models. Different process-based constraints might be inferred for different models. The approach described here could usefully extended to other single model ensembles and the collective understandings could be valuable for improving model process and feedbacks more generally.

Abstract We describe and evaluate historical simulations which use the third Hadley Centre Global Environment Model in the Global Coupled configuration 3.1 (HadGEM3‐GC3.1) and which form part of the UK's contribution to the sixth Coupled Model Intercomparison Project, CMIP6. These simulations, run at two resolutions, respond to historically evolving forcings such as greenhouse gases, aerosols, solar irradiance, volcanic aerosols, land use, and ozone concentrations. We assess the response of the simulations to these historical forcings and compare against the observational record. This includes the evolution of global mean surface temperature, ocean heat content, sea ice extent, ice sheet mass balance, permafrost extent, snow cover, North Atlantic sea surface temperature and circulation, and decadal precipitation. We find that the simulated time evolution of global mean surface temperature broadly follows the observed record but with important quantitative differences which we find are most likely attributable to strong effective radiative forcing from anthropogenic aerosols and a weak pattern of sea surface temperature response in the low to middle latitudes to volcanic eruptions. We also find evidence that anthropogenic aerosol forcings play a role in driving the Atlantic Multidecadal Variability and the Atlantic Meridional Overturning Circulation, which are key features of the North Atlantic ocean. Overall, the model historical simulations show many features in common with the observed record over the period 1850–2014 and so provide a basis for future in‐depth study of recent climate change.

The Atlantic Meridional Overturning Circulation (AMOC) has been, and will continue to be, a key factor in the modulation of climate change both locally and globally. However, there remains considerable uncertainty in recent AMOC evolution. Here, we show that the multimodel mean AMOC strengthened by approximately 10% from 1850-1985 in new simulations from the 6th Coupled Model Intercomparison Project (CMIP6), a larger change than was seen in CMIP5. Across the models, the strength of the AMOC trend up to 1985 is related to a proxy for the strength of the aerosol forcing. Therefore, the multimodel difference is a result of stronger anthropogenic aerosol forcing on average in CMIP6 than CMIP5, which is primarily due to more models including aerosol-cloud interactions. However, observational constraints-including a historical sea surface temperature fingerprint and shortwave radiative forcing in recent decades-suggest that anthropogenic forcing and/or the AMOC response may be overestimated.

A key contribution to the latest generation of climate projections for the UK (UKCP18) was a perturbed parameter ensemble (PPE) of global coupled models based on HadGEM3-GC3.05. Together with 13 CMIP5 simulations, this PPE provides users with a dataset that samples modelling uncertainty and is ideal for use in impacts studies. Evaluations of global mean surface temperatures for this PPE have shown twenty-first century warming rates consistently at the top end of the CMIP5 range. Here we investigate one potential contributory factor to this lack of spread: that the methodology to select plausible members from a larger, related PPE of atmosphere-only experiments preferentially ruled out those predicted to have more negative climate feedbacks (i.e. lower climate sensitivities). We confirm that this is indeed the case. We show that performance in extratropical long-wave cloud forcing played a key role in this by constraining ice cloud parameters, which in turn constrained the feedback distribution (though causal links are not established). The relatively weak relationship driving this constraint is shown to arise from stronger relationships for the long-wave and short-wave cloud feedback components, which largely cancel out due to changes in tropical high clouds. Moreover, we show that the strength of these constraints is due to a structural bias in extratropical long-wave cloud forcing across the PPE. We discuss how choices made in the methodology to pick the plausible PPE members may result in an overly strong constraint when there is a structural bias and possible improvements to this methodology for the future.

We analyze the atmospheric processes that explain the large changes in radiative feedbacks between the two latest climate configurations of the Hadley Centre Global Environmental model. We use a large set of atmosphere-only climate change simulations (amip and amip-p4K) to separate the contributions to the differences in feedback parameter from all the atmospheric model developments between the two latest model configurations. We show that the differences are mostly driven by changes in the shortwave cloud radiative feedback in the midlatitudes, mainly over the Southern Ocean. Two new schemes explain most of the differences: the introduction of a new aerosol scheme and the development of a new mixed-phase cloud scheme. Both schemes reduce the strength of the preexisting shortwave negative cloud feedback in the midlatitudes. The new aerosol scheme dampens a strong aerosol-cloud interaction, and it also suppresses a negative clear-sky shortwave feedback. The mixed-phase scheme increases the amount of cloud liquid water path (LWP) in the present day and reduces the increase in LWP with warming. Both changes contribute to reducing the negative radiative feedback of the increase of LWP in the warmer climate. The mixed-phase scheme also enhances a strong, preexisting, positive cloud fraction feedback. We assess the realism of the changes by comparing present-day simulations against observations and discuss avenues that could help constrain the relevant processes.

Abstract Climate forcing, sensitivity, and feedback metrics are evaluated in both the United Kingdom's physical climate model HadGEM3‐GC3.1 at low (‐LL) and medium (‐MM) resolution and the United Kingdom's Earth System Model UKESM1. The effective climate sensitivity (EffCS) to a doubling of CO2 is 5.5 K for HadGEM3.1‐GC3.1‐LL and 5.4 K for UKESM1. The transient climate response is 2.5 and 2.8 K, respectively. While the EffCS is larger than that seen in the previous generation of models, none of the model's forcing or feedback processes are found to be atypical of models, though the cloud feedback is at the high end. The relatively large EffCS results from an unusual combination of a typical CO2 forcing with a relatively small feedback parameter. Compared to the previous U.K. climate model, HadGEM3‐GC2.0, the EffCS has increased from 3.2 to 5.5 K due to an increase in CO2 forcing, surface albedo feedback, and midlatitude cloud feedback. All changes are well understood and due to physical improvements in the model. At higher atmospheric and ocean resolution (HadGEM3‐GC3.1‐MM), there is a compensation between increased marine stratocumulus cloud feedback and reduced Antarctic sea‐ice feedback. In UKESM1, a CO2 fertilization effect induces a land surface vegetation change and albedo radiative effect. Historical aerosol forcing in HadGEM3‐GC3.1‐LL is −1.1 W m−2. In HadGEM3‐GC3.1‐LL historical simulations, cloud feedback is found to be less positive than in abrupt‐4xCO2, in agreement with atmosphere‐only experiments forced with observed historical sea surface temperature and sea‐ice variations. However, variability in the coupled model's historical sea‐ice trends hampers accurate diagnosis of the model's total historical feedback.

The Coupled Model Intercomparison Project 6 protocol suggests prescribing preindustrial ozone concentrations in abrupt‐4xCO2 simulations. This leads to a mismatch between the thermal tropopause, which rises due to climate change, and the ozone tropopause, which remains fixed. The result is unphysically high ozone concentrations in the upper troposphere, leading to a warm bias in cold point temperature and increased stratospheric water vapor. In the U.K. physical climate model HadGEM3‐GC3.1 this increases the surface climate sensitivity. In the future, other climate models without interactive ozone schemes may face similar problems. We describe a method to interactively redistribute ozone in climate simulations, which removes the inconsistency between the thermal and ozone tropopause heights while retaining the prescribed ozone distribution as closely as possible. This removes unphysical consequences of the tropopause mismatch, while still allowing a fair comparison against other Coupled Model Intercomparison Project 6 model simulations. After each model year, the monthly mean, zonal mean, thermal tropopause is formed based on the previous two model years. The ozone tropopause is defined at 1 km below the thermal tropopause by setting ozone concentrations there to 80 ppbv, and smoothing appropriately. The mass of ozone removed from the troposphere is added to the stratosphere thus conserving the total mass of ozone. This redistribution is then applied proportionally to the 3‐D monthly mean ozone concentrations. The climate model is run for the following year, using this redistributed ozone, and then the whole process is repeated. Results are presented from preindustrial and abrupt‐4xCO2 simulations, but this method can be used for any climate simulation.

The Earth’s solar reflectance is reduced through rapid climate adjustments to increasing CO2, via a decrease in total cloud cover over ocean. Perturbations to marine boundary-layer clouds are essentially important for the global radiative balance at the top of the atmosphere. However, the physical robustness of low cloud adjustments to increasing CO2 has not been assessed systematically. Here we show that low cloud adjustment is distinct from that in total cloud and is seasonally variant. Among multiple climate models, marine boundary-layer clouds over the subtropics and extratropics (especially over the Northern Hemisphere) are consistently increased in the rapid adjustment, while middle and high clouds are greatly reduced. The increase in low cloud cover is only found during summer, associated with a summertime enhancement of lower tropospheric stability. We further examine mechanisms behind the rapid adjustments of low cloud and inversion strength of the boundary layer, using land surface temperature prescribing experiments in an atmospheric general circulation model (AGCM). Summertime increases in low cloud and enhanced inversion strength over the ocean simulated in this AGCM are attributed to (1) CO2-induced land warming; and (2) reduced radiative cooling in the lower troposphere due to increased CO2. The seasonality in the cloud adjustment implies an importance of seasonal variations in background cloud and atmospheric circulation related to the Hadley and monsoon circulations for radiative forcing, feedback and climate sensitivity.

High-quality observations are powerful tools for the evaluation of climate models towards improvement and reduction of uncertainty. Particularly at low latitudes, the most uncertain aspect lies in the representation of moist convection and interaction with dynamics, where rising motion is tied to deep convection and sinking motion to dry regimes. Since humidity is closely coupled with temperature feedbacks in the tropical troposphere, a proper representation of this region is essential. Here we demonstrate the evaluation of atmospheric climate models with satellite-based observations from Global Positioning System (GPS) radio occultation (RO), which feature high vertical resolution and accuracy in the troposphere to lower stratosphere. We focus on the representation of the vertical atmospheric structure in tropical convection regimes, defined by high updraft velocity over warm surfaces, and investigate atmospheric temperature and humidity profiles. Results reveal that some models do not fully capture convection regions, particularly over land, and only partly represent strong vertical wind classes. Models show large biases in tropical mean temperature of more than 4 K in the tropopause region and the lower stratosphere. Reasonable agreement with observations is given in mean specific humidity in the lower to mid-troposphere. In moist convection regions, models tend to underestimate moisture by 10 to 40 % over oceans, whereas in dry downdraft regions they overestimate moisture by 100 %. Our findings provide evidence that RO observations are a unique source of information, with a range of further atmospheric variables to be exploited, for the evaluation and advancement of next-generation climate models.

Future responses of cloud regimes are analyzed for five CMIP5 models forced with observed SSTs and subject to a patterned SST perturbation. Correlations between cloud properties in the control climate and changes in the warmer climate are investigated for each of a set of cloud regimes defined using a clustering methodology. The only significant (negative) correlation found is in the in-regime net cloud radiative effect for the stratocumulus regime. All models overestimate the in-regime albedo of the stratocumulus regime. Reasons for this bias and its relevance to the future response are investigated. A detailed evaluation of the models' daily-mean contributions to the albedo from stratocumulus clouds with different cloud cover fractions reveals that all models systematically underestimate the relative occurrence of overcast cases but overestimate those of broken clouds. In the warmer climate the relative occurrence of overcast cases tends to decrease while that of broken clouds increases. This suggests a decrease in the climatological in-regime albedo with increasing temperature (a positive feedback); this is opposite to the feedback suggested by the analysis of the bulk in-regime albedo. Furthermore we find that the inter-model difference in the sign of the in-cloud albedo feedback is consistent with the difference in sign of the in-cloud liquid water path response, and there is a strong positive correlation between the in-regime liquid water path in the control climate and its response to warming. We therefore conclude that further breakdown of the in-regime properties into cloud cover and in-cloud properties is necessary to better understand the behavior of the stratocumulus regime. Since cloud water is a physical property and is independent of a model's radiative assumptions, it could potentially provide a useful emergent constraint on cloud feedback.