Abstract. Cloud processes constitute one of the key uncertainties for climate change projections. The fourth iteration of the Cloud Feedback Model Intercomparison Project, CFMIP4, contributes to the Coupled Model Intercomparison Project phase 7 (CMIP7), by providing a set of global climate model experiments aiming to enhance our understanding of clouds, circulation and climate sensitivity, thereby informing improved projections of future climate change. CFMIP4 targets four knowledge gaps: (1) Physical mechanisms of cloud feedback and adjustment; (2) Dependence of cloud feedback and adjustment on climate base state and on the nature of the forcing; (3) Coupled mechanisms of the sea-surface temperature "pattern effect"; and (4) Coupling of clouds with circulation and precipitation. CFMIP4 contributes four CMIP7 Assessment Fast Track experiments that are central to the quantification of climate feedback and sensitivity in past, present and future climates, essential for process understanding and model evaluation. Furthermore, CFMIP4 supports the joint analysis of models and observations through a data request that includes process and satellite simulator output.

We report the first results of an intercomparison of nine atmospheric models with horizontal resolutions between 150 m and 2.5 km: ICON, DALES, MesoNH, MicroHH, MetOffice UM, HARMONIE-AROME, AROME, WRF, and COSMO. These models all simulated shallow cumulus convection over the subtropical Atlantic Ocean during the period of the EUREC4A field campaign (Jan-Feb 2020), most of them with open boundary conditions and on large (>500 km) domains. Our "EUREC4A-MIP" seeks to answer how consistently such models reproduce i) statistics of the observed atmospheric state, clouds and energy/water budgets over large (200 km) areas, ii) the horizontal cloud organization that set the regions' contribution to the planetary albedo and iii) the coupling between clouds and circulations at mesoscales. We extend an open invitation for further exploration of the simulation data, which we publish through the EUREC4A intake catalog.

Climate models indicate that dry extremes will be exacerbated in many regions of the world1,2. However, confidence in the magnitude and timing of these projected changes remains low3,4, leaving societies largely unprepared5,6. Here we show that constraining model projections with observations using a newly proposed emergent constraint (EC) reduces the uncertainty in predictions of a core drought indicator, the longest annual dry spell (LAD), by 10-26% globally. Our EC-corrected projections reveal that the increase in LAD will be 42-44% greater, on average, than 'mid-range' or 'high-end' future forcing scenarios currently indicate. These results imply that by the end of this century, the global mean land-only LAD could be 10 days longer than currently expected. Using two generations of climate models, we further uncover global regions for which historical LAD biases affect the magnitude of projected LAD increases, and we explore the role of land-atmosphere feedbacks therein. Our findings reveal regions with potentially higher- and earlier-than-expected drought risks for societies and ecosystems, and they point to possible mechanisms underlying the biases in the current generation of climate models.

Au début des années 1990, le rôle des nuages dans le changement climatique devient une question centrale dans les sciences du climat. Cet article discute du rôle du LMD dans l’avancée de cette thématique au niveau international : dans la mise en évidence du rôle des nuages bas dans les incertitudes sur la sensibilité climatique, dans la compréhension des mécanismes de rétroaction associés aux nuages bas et aux nuages d’enclume, puis dans la réduction de ces incertitudes à l’aide d’observations spatiales et de campagnes de terrain. Le rôle du LMD dans cette thématique est caractéristique de sa culture, à savoir comprendre le climat par l’étude des processus, et vice-versa.

Coherent structures are characterized in high-resolution simulations of three atmospheric boundary layers: dry convection, marine cumulus, and stratocumulus. Based on radioactive-decaying tracers emitted at different altitudes (surface, top of well-mixed layer, and cloud top), a object-oriented methodology allows individual characterization of coherent tridimensional plumes within the flow. Each boundary layer shows updraft structures surrounded by subsiding shells that have similar thermodynamical characteristics. Well-mixed downdrafts are located relatively close to updrafts and entrain dry, warm air from the free troposphere. Identified in all boundary layers, these subsiding structures are triggered by air mass convergence linked to updrafts' divergence and are thus part of an overturning circulation in well-mixed layers. Close to the surface, downdrafts' divergence constrain updrafts' locations and thus shape a mesoscale cellular organisation with cell sizes scaling with the boundary-layer height (aspect ratio of around 2).Active cumulus formation does not strongly perturb the spatial organisation of the sub-cloud well-mixed layer. The stratocumulus-topped boundary layer also shares similarities with the overturning circulation despite having condensation and cloud-radiation diabatic effects within the mixed layer. However, the visible mesoscale organisation of stratocumulus shows larger cells than the boundary-layer depth (aspect ratio > 10) that suggest deviations from the clear-sky conceptual view. The boundary-layer decoupling influences mass fluxes of coherent structures and thus potentially play a role in shaping the spatial organisation.Since well-mixed downdrafts contribute to a significant part of resolved flux of heat and moisture, our results suggest that downdraft properties in well mixed layers should be represented at the subgrid scale in climate models through non-local mass-flux parameterizations.

The increasing risk of dry extremes and droughts and their further projected exacerbation due to climate change urges the development of reliable risk assessments and mitigation pathways on a regional and global scale. This foremost requires accurate and unambiguous model predictions of dry extremes, as this underpins the effectiveness of the proposed strategies. At present, however, the confidence in regional drought projections is defined as ‘medium to low' by the Intergovernmental Panel on Climate Change (IPCC) sixth assessment report (AR6), and reducing this uncertainty remains one of the main goals in coming years. In this study, the bias in future projected changes in annual meteorological drought duration (hereafter, longest annual drought, LAD) is assessed in the ensemble of CMIP5 and CMIP6 models. The analyses show that it is the present-day inter-model spread in LAD climatology that largely determines the inter-model uncertainty in future predicted LAD changes. Hereby, both CMIP5 and CMIP6 model ensembles indicate a robust “dry-model-gets-drier” relationship in future LAD projections on a global and regional scale. Correcting for this bias using emerging constraint principles and past observational LAD information, we find that nearly half of the world's land area with projected increases in drought duration is underestimating the predicted model ensemble mean change, imposing higher-than-expected risks to the societies and ecosystems. Analysis of physical mechanisms that could underlie this emergent “present-future relationship” points to differences in the responses of “dry models” and “wet models” to CO2 forcing. Dry and wet models show differences in climate states, which support the role of land–atmosphere feedbacks and convective scheme sensitivity to atmospheric moisture in the spread of future LAD change projections. In conclusion, the study reveals world regions where climate change may cause stronger drought duration aggravation than expected, and emphasizes the importance of reducing systematic model errors, which are presently largely owed to rainfall biases. Correcting these biases will increase the confidence of future dry extremes predictions, a prerequisite for the effective drought risk reduction in the near future with direct benefits for human and natural systems.

Abstract. Studies of emergent constraints have frequently proposed that a single metric alone can constrain future responses of the Earth system to anthropogenic emissions. The prevalence of this thinking has led to literature and messaging which is sometimes confusing to policymakers, with a series of studies over the last decade making confident, yet contradictory, claims on the probability bounds of key climate variables. Here, we illustrate that emergent constraints are more likely to occur where the variance across an ensemble of climate models of both observable and future climate arises from common structural assumptions and few degrees of freedom. Such cases are likely to occur when processes are represented in a common, oversimplified fashion throughout the ensemble, about which we have the least confidence in performance out of sample. We consider these issues in the context of a number of published constraints, and argue that the application of emergent constraints alone to estimate uncertainties in unknown climate responses can potentially lead to bias and overconfidence in constrained projections. Together with statistical robustness and plausibility of mechanism, assessments of climate responses must include multiple lines of evidence to identify biases that arise from common oversimplified modeling assumptions which impact both present and future climate simulations in order to mitigate against the influence of common structural biases.

How tropical low clouds change with climate remains the dominant source of uncertainty in global warming projections. An analysis of an ensemble of CMIP5 climate models reveals that a significant part of the spread in the models’ climate sensitivity can be accounted by differences in the climatological shallowness of tropical low clouds in weak-subsidence regimes: models with shallower low clouds in weak-subsidence regimes tend to have a higher climate sensitivity than models with deeper low clouds. The dynamical mechanisms responsible for the model differences are analyzed. Competing effects of parameterized boundary-layer turbulence and shallow convection are found to be essential. Boundary-layer turbulence and shallow convection are typically represented by distinct parameterization schemes in current models—parameterization schemes that often produce opposing effects on low clouds. Convective drying of the boundary layer tends to deepen low clouds and reduce the cloud fraction at the lowest levels; turbulent moistening tends to make low clouds more shallow but affects the low-cloud fraction less. The relative importance different models assign to these opposing mechanisms contributes to the spread of the climatological shallowness of low clouds and thus to the spread of low-cloud changes under global warming.

Abstract The equilibrium climate sensitivity (ECS) in the latest version of CNRM climate model, CNRM‐CM6‐1, and in its high‐resolution counterpart, CNRM‐CM6‐1‐HR, is significantly larger than in the previous version (CNRM‐CM5.1). The traceability of this climate sensitivity change is investigated using coupled ocean‐atmosphere model climate change simulations. These simulations show that the increase in ECS is the result of changes in the atmospheric component. A particular attention is paid to the method used to decompose the equilibrium temperature response difference, by using a linearized decomposition of the individual radiative agents diagnosed by a radiative kernel technique. The climate sensitivity increase is primarily due to the cloud radiative responses, with a predominant contribution of the tropical longwave response (including both feedback and forcing adjustment) and a significant contribution of the extratropical and tropical shortwave feedback changes. A series of stand‐alone atmosphere experiments is carried out to quantify the contributions of each atmospheric development to this difference between CNRM‐CM5.1 and CNRM‐CM6‐1. The change of the convection scheme appears to play an important role in driving the cloud changes, with a large effect on the tropical longwave cloud feedback change.

Abstract. Studies of emergent constraints have frequently proposed that a single metric can constrain future responses of the Earth system to anthropogenic emissions. Here, we illustrate that strong relationships between observables and future climate across an ensemble can arise from common structural model assumptions with few degrees of freedom. Such cases have the potential to produce strong yet overconfident constraints when processes are represented in a common, oversimplified fashion throughout the ensemble. We consider these issues in the context of a collection of published constraints and argue that although emergent constraints are potentially powerful tools for understanding ensemble response variation and relevant observables, their naïve application to reduce uncertainties in unknown climate responses could lead to bias and overconfidence in constrained projections. The prevalence of this thinking has led to literature in which statements are made on the probability bounds of key climate variables that were confident yet inconsistent between studies. Together with statistical robustness and a mechanism, assessments of climate responses must include multiple lines of evidence to identify biases that can arise from shared, oversimplified modelling assumptions that impact both present and future climate simulations in order to mitigate against the influence of shared structural biases.

Abstract The development of parameterizations is a major task in the development of weather and climate models. Model improvement has been slow in the past decades, due to the difficulty of encompassing key physical processes into parameterizations, but also of calibrating or “tuning” the many free parameters involved in their formulation. Machine learning techniques have been recently used for speeding up the development process. While some studies propose to replace parameterizations by data‐driven neural networks, we rather advocate that keeping physical parameterizations is key for the reliability of climate projections. In this paper we propose to harness machine learning to improve physical parameterizations. In particular, we use Gaussian process‐based methods from uncertainty quantification to calibrate the model free parameters at a process level. To achieve this, we focus on the comparison of single‐column simulations and reference large‐eddy simulations over multiple boundary‐layer cases. Our method returns all values of the free parameters consistent with the references and any structural uncertainties, allowing a reduced domain of acceptable values to be considered when tuning the three‐dimensional (3D) global model. This tool allows to disentangle deficiencies due to poor parameter calibration from intrinsic limits rooted in the parameterization formulations. This paper describes the tool and the philosophy of tuning in single‐column mode. Part 2 shows how the results from our process‐based tuning can help in the 3D global model tuning.

Droughts are defined as one of the most devastating natural disasters of modern times and a key challenge faced under climate change. The complexity of interacting physical processes that underlie the shortage of rainfall in climate models hampers accurate representation of present-day droughts, and leads to differences in their responses to increased greenhouse gas (GHG) concentrations in the future. As a result, the confidence in drought projections is currently defined as ‘medium to low' by the Intergovernmental Panel on Climate Change (IPCC), and reducing this uncertainty remains one of the main goals in coming years, with significant benefits for human and natural systems.In this study we explore a relationship between biases in simulated present-day values of longest annual drought (LAD) and future projections of LAD in an ensemble of CMIP5 and CMIP6 models. We find that present-day model bias explains almost 95 % of the future uncertainty in LAD by the end of the 21st century, attributed to the well-known precipitation simulation errors: “drier” models with longer annual droughts at present tend to predict larger LAD values worldwide in the future, as well as a stronger response to GHG forcing in LAD, which is significant in more than 40 % of the global land area.Substituting observational LAD estimates from satellite data into this model-revealed “present–future relationlarship” suggests that the 21st century global mean increase in duration of annual meteorological droughts could be significantly larger than predicted by the CMIP5 and CMIP6 model ensembles. This emergent constraint reduces global mean uncertainty range in future LAD estimates from 45–100 to 75–90 days, a level more typical of the prediction range of “drier” models. The findings reveal world regions where climate change may cause stronger meteorological drought aggravation than expected, and emphasise the importance of reducing model errors, which are presently largely owed to rain biases, to increase confidence in future predictions.

Models disagree on a significant number of responses to climate change, such as climate feedback, regional changes, or the strength of equilibrium climate sensitivity. Emergent constraints aim to reduce these uncertainties by finding links between the inter-model spread in an observable predictor and climate projections. In this paper, the concepts underlying this framework are recalled with an emphasis on the statistical inference used for narrowing uncertainties, and a review of emergent constraints found in the last two decades. Potential links between highlighted predictors are explored, especially those targeting uncertainty reductions in climate sensitivity, cloud feedback, and changes of the hydrological cycle. Yet the disagreement across emergent constraints suggests that the spread in climate sensitivity can not be significantly narrowed. This calls for weighting the realism of emergent constraints by quantifying the level of physical understanding explaining the relationship. This would also permit more efficient model evaluation and better targeted model development. In the context of the upcoming CMIP6 model intercomparison a growing number of new predictors and uncertainty reductions is expected, which call for robust statistical inferences that allow cross-validation of more likely estimates.

Abstract The representation of stratocumulus clouds, and of the stratocumulus to cumulus transitions which are ubiquitous features of marine boundary layer clouds, remains a challenge for climate models. We show how a mass flux representation of boundary layer convective structures combined with an eddy diffusivity scheme, the “thermal plume model,” first developed to represent cumulus clouds, can also adequately simulate stratocumulus and the stratocumulus to cumulus transition in a climate model. To achieve this, the detrainment formulation, in which detrainment increases for increasing negative buoyancy, has to be slightly modified: the buoyancy of a thermal plume parcel of air is computed by comparing the virtual potential temperature θv,th of the parcel with that of the surrounding environment θv,env at a given distance above instead of at the same level. This is consistent with the picture of detrained air parcels that experience some overshoot and reach a final destination at a level lower than the one at which they effectively leave the cloud or organized convective plume. The impacts of this modification are documented both for selected cases of stratocumulus, in comparison with large‐eddy simulations, and in full 3‐D climate simulations, in comparison with satellite observations of cloud cover. The modified scheme provides a uniform treatment of the dry convective boundary layer, of cumulus clouds, of stratocumulus, and of the transition from stratocumulus to cumulus. It is included in the most recent version of the LMDZ atmospheric general circulation model.

Understanding what controls the water vapor isotopic composition of the sub-cloud layer (SCL) over tropical oceans (δD0) is a first step towards understanding the water vapor isotopic composition everywhere in the troposphere. We propose an analytical model to predict δD0 motivated by the hypothesis that the altitude from which the free tropospheric air originates (zorig) is an important factor: when the air mixing into the SCL is lower in altitude, it is generally moister, and thus it depletes the SCL more efficiently. We extend previous simple box models of the SCL by prescribing the shape of δD vertical profiles as a function of humidity profiles and by accounting for rain evaporation and horizontal advection effects. The model relies on the assumption that δD profiles are steeper than mixing lines, and that the SCL is at steady state, restricting its applications to timescales longer than daily. In the model, δD0 is expressed as a function of zorig, humidity and temperature profiles, surface conditions, a parameter describing the steepness of the δD vertical gradient, and a few parameters describing rain evaporation and horizontal advection effects. We show that δD0 does not depend on the intensity of entrainment, in contrast to several previous studies that had hoped that δD0 measurements could help estimate this quantity. Based on an isotope-enabled general circulation model simulation, we show that δD0 variations are mainly controlled by mid-tropospheric depletion and rain evaporation in ascending regions and by sea surface temperature and zorig in subsiding regions. In turn, could δD0 measurements help estimate zorig and thus discriminate between different mixing processes? For such isotope-based estimates of zorig to be useful, we would need a precision of a few hundred meters in deep convective regions and smaller than 20 m in stratocumulus regions. To reach this target, we would need daily measurements of δD in the mid-troposphere and accurate measurements of δD0 (accuracy down to 0.1 ‰ in the case of stratocumulus clouds, which is currently difficult to obtain). We would also need information on the horizontal distribution of δD to account for horizontal advection effects, and full δD profiles to quantify the uncertainty associated with the assumed shape for δD profiles. Finally, rain evaporation is an issue in all regimes, even in stratocumulus clouds. Innovative techniques would need to be developed to quantify this effect from observations.

A novel methodology is proposed to characterize coherent structures in large eddy simulations. Based on two passive tracers emitted respectively at the surface and at cloud top, the object‐oriented framework allows individual characterization of coherent tridimensional plumes within the flow. Applying this method in a simulation of the diurnal cycle of a marine stratocumulus‐topped boundary layer shows that coherent updraft and downdraft structures contribute to most of the total transport of heat and moisture, although covering a small part of the domain volume. On average, downdrafts contribute equally compared to updrafts for moisture fluxes and more than updrafts for heat fluxes. The relative contribution of updraft and downdraft objects to heat transport exhibits a large diurnal cycle, which suggests cloud‐turbulence‐radiation interaction. Our results suggest that subgrid downdraft properties within stratocumulus‐topped boundary layers should be represented through nonlocal mass‐flux parameterization in climate models.

12 Models disagree on a significant number of responses to climate change, such as climate feedback, 13 regional changes, or the strength of equilibrium climate sensitivity. Emergent constraints aim to 14 reduce these uncertainties by finding links between the inter-model spread in an observable predictor 15 and climate projections. In this paper, the concepts underlying this framework are recalled with 16 an emphasis on the statistical inference used for narrowing uncertainties, and a review of emergent 17 constraints found in the last two decades. Potential links between highlighted predictors are explored, 18 especially those targeting uncertainty reductions in climate sensitivity, cloud feedback, and changes 19 of the hydrological cycle. Yet the disagreement across emergent constraints suggests that the spread in 20 climate sensitivity can not be significantly narrowed. This calls for weighting the realism of emergent 21 constraints by quantifying the level of physical understanding explaining the relationship. This would 22 also permit more efficient model evaluation and better targeted model development. In the context of 23 the upcoming CMIP6 model intercomparison a growing number of new predictors and uncertainty 24 reductions is expected, which call for robust statistical inferences that allow cross-validation of more 25 likely estimates. 26

Current climate models represent the zonal- and annual-mean intertropical convergence zone (ITCZ) position in a biased way, with an unrealistic double precipitation peak straddling the equator in the ensemble mean over the models. This bias is seasonally and regionally localized. It results primarily from two regions: the eastern Pacific and Atlantic (EPA), where the ITCZ in boreal winter and spring is displaced farther south than is observed; and the western Pacific (WP), where a more pronounced and wider than observed double ITCZ straddles the equator year-round. Additionally, the precipitation associated with the ascending branches of the zonal overturning circulations (e.g., Walker circulation) in the Pacific and Atlantic sectors is shifted westward. We interpret these biases in light of recent theories that relate the ITCZ position to the atmospheric energy budget. WP biases are associated with the well known Pacific cold tongue bias, which, in turn, is linked to atmospheric net energy input biases near the equator. In contrast, EPA biases are shown to be associated with a positive bias in the cross-equatorial divergent atmospheric energy transport during boreal winter and spring, with two potential sources: tropical biases associated with equatorial sea surface temperatures (SSTs) and tropical low clouds, and extratropical biases associated with Southern Ocean clouds and north Atlantic SST. The distinct seasonal and regional characteristics of WP and EPA biases and the differences in their associated energy budget biases suggest that the biases in the two sectors involve different mechanisms and potentially different sources.