Large Ensemble Simulations Research Articles

High-order moment closure models with random batch method for efficient computation of multiscale turbulent systems.

We propose a high-order stochastic-statistical moment closure model for efficient ensemble prediction of leading-order statistical moments and probability density functions in multiscale complex turbulent systems. The statistical moment equations are closed by a precise calibration of the high-order feedbacks using ensemble solutions of the consistent stochastic equations, suitable for modeling complex phenomena including non-Gaussian statistics and extreme events. To address challenges associated with closely coupled spatiotemporal scales in turbulent states and expensive large ensemble simulation for high-dimensional systems, we introduce efficient computational strategies using the random batch method (RBM). This approach significantly reduces the required ensemble size while accurately capturing essential high-order structures. Only a small batch of small-scale fluctuation modes is used for each time update of the samples, and exact convergence to the full model statistics is ensured through frequent resampling of the batches during time evolution. Furthermore, we develop a reduced-order model to handle systems with really high dimensions by linking the large number of small-scale fluctuation modes to ensemble samples of dominant leading modes. The effectiveness of the proposed models is validated by numerical experiments on the one-layer and two-layer Lorenz '96 systems, which exhibit representative chaotic features and various statistical regimes. The full and reduced-order RBM models demonstrate uniformly high skill in capturing the time evolution of crucial leading-order statistics, non-Gaussian probability distributions, while achieving significantly lower computational cost compared to direct Monte-Carlo approaches. The models provide effective tools for a wide range of real-world applications in prediction, uncertainty quantification, and data assimilation.

The warm Arctic-cold north american pattern in CanESM5 large ensemble simulations: Eurasian influence and uncertainty due to internal variability

This study examines the warm Arctic-cold North American pattern (WACNA) and its connection with the warm Arctic-cold Eurasia pattern (WACE) using ERA5 reanalysis and a 50-member ensemble of historical climate simulations produced by CanESM5, the Canadian model participated in CMIP6. The results indicate that a negative WACE-like pattern typically precedes a positive WACNA pattern by one month, and the presence of a negative Asian-Bering-North American (ABNA)-like circulation pattern connects Eurasia and North America, along with the Pacific-North American (PNA)-like pattern. The negative ABNA-like pattern can be attributed to anomalous heating in southern Siberia, which is associated with the negative WACE pattern and its featured Eurasian warming. The negative PNA-like pattern is influenced by negative SST anomalies in the tropical Pacific, resembling tropical ENSO variability. Anomalous temperature advection in the lower troposphere follows the circulation anomaly, which supports the formation of WACNA. Conversely, processes with circulation anomalies of opposite sign result in a negative WACNA pattern. The tropical ENSO variability does not significantly impact the WACNA pattern and its linkage with WACE. CanESM5 simulates the WACNA pattern and WACE-WACNA connection well, with some discrepancies in the magnitude of anomalies compared to ERA5 reanalysis. The uncertainty in the simulated WACNA pattern due to internal climate variability is dominated by two modes of inter-member variability: a southeast-northwest phase shift and a local variation in amplitude.

Understanding the recent increase in multiyear La Niñas

Five out of six La Niña events since 1998 have lasted two to three years. Why so many long-lasting multiyear La Niña events have emerged recently and whether they will become more common remains unknown. Here we show that ten multiyear La Niña events over the past century had an accelerated trend, with eight of these occurring after 1970. The two types of multiyear La Niña events over this time period followed either a super El Niño or a central Pacific El Niño. We find that multiyear La Niña events differ from single-year La Niñas by a prominent onset rate, which is rooted in the western Pacific warming-enhanced zonal advective feedback for the central Pacific multiyear La Niña events type and thermocline feedback for the super El Niño multiyear La Niña events type. The results from large ensemble climate simulations support the observed multiyear La Niña events–western Pacific warming link. More multiyear La Niña events will exacerbate adverse socioeconomic impacts if the western Pacific continues to warm relative to the central Pacific.

The Effect of Greenhouse Gas–Induced Warming on the Impact of El Niño and La Niña Events on Daily Precipitation Extremes in the Boreal Cold Season

Abstract El Niño–Southern Oscillation (ENSO) has a profound influence on the occurrence of extreme precipitation events at local and regional scales in the present-day climate, and thus it is important to understand how that influence may change under future global warming. We consider this question using the large-ensemble simulations of CESM2, which simulates ENSO well historically. CESM2 projects that the influence of ENSO on extreme precipitation will strengthen further under the SSP3–7.0 scenario in most regions whose extreme precipitation regimes are strongly affected by ENSO in the boreal cold season. Extreme precipitation in the boreal cold season that exceeds historical thresholds is projected to become more common throughout the ENSO cycle. The difference in the intensity of extreme precipitation events that occur under El Niño and La Niña conditions will increase, resulting in “more extreme and more variable hydroclimate extremes.” We also consider the processes that affect the future intensity of extreme precipitation and how it varies with the ENSO cycle by partitioning changes into thermodynamic and dynamic components. The thermodynamic component, which reflects increases in atmospheric moisture content, results in a relatively uniform intensification of ENSO-driven extreme precipitation variation. In contrast, the dynamic component, which reflects changes in vertical motion, produces a strong regional difference in the response to forcing. In some regions, this component amplifies the thermodynamic-induced changes, while in others, it offsets them or even results in reduction in extreme precipitation variation.

Assessing the impact of climate change on sediment discharge using a large ensemble rainfall dataset in Pekerebetsu River basin, Hokkaido

Increased rainfall associated with climate change can increase sediment discharge. The supply of fine sediment from slope failures inhibits bed armoring of mountain rivers and increases sediment discharge to the downstream reaches. Floods without slope failures lead to bed erosion and armoring and may ultimately decrease sediment discharge. Thus, it is important to consider sediment discharge from slope failure and bed erosion as factors affecting sediment production. Climate change affects not only the rainfall amount, but also the temporal rainfall pattern; consequently, the pattern affects the sediment production factors and the amount of sediment discharge. However, changes in sediment discharge due to climate change based on sediment production sources have not yet been clarified. In this study, we statistically analyzed 1200 results simulated using a physics-based sediment runoff model to assess the impact of changes in temporal rainfall patterns on sediment discharge and sediment production sources in the Pekerebetsu River Basin. In the simulations, we used the rainfall predicted in d4PDF (Database for policy decision-making for future climate change), a large ensemble climate simulation database at 5 km and 20 km resolutions. Our results showed that the climate-driven increase in sediment discharge was considerably larger than that of rainfall. An increase in short-term heavy rainfall increased the supply of fine sediments from slope failure. This resulted in the suppression of bed armoring and a large increase in sediment discharge. Thus, the increase in sediment discharge is not only caused by an increase in rainfall but also by changes in temporal rainfall patterns and sediment production factors. The sediment discharge calculated for the 20 km resolution climate projection was nearly one order of magnitude smaller than that for the 5 km resolution. This suggests that the 20 km resolution climate projections do not adequately represent orographic rainfall in the mountains and thus, do not adequately reproduce extreme sediment discharge events. An increased sediment supply causes bed aggradation and decreases the river conveyance capacity of the downstream channel. The model developed in this study will contribute to flood risk analysis and flood control planning for increased rainfall due to climate change.

ENSO skewness hysteresis and associated changes in strong El Niño under a CO2 removal scenario

El Niño-Southern Oscillation (ENSO) sea surface temperature (SST) anomaly skewness encapsulates the nonlinear processes of strong ENSO events and affects future climate projections. Yet, its response to CO2 forcing remains not well understood. Here, we find ENSO skewness hysteresis in a large ensemble CO2 removal simulation. The positive SST skewness in the central-to-eastern tropical Pacific gradually weakens (most pronounced near the dateline) in response to increasing CO2, but weakens even further once CO2 is ramped down. Further analyses reveal that hysteresis of the Intertropical Convergence Zone migration leads to more active and farther eastward-located strong eastern Pacific El Niño events, thus decreasing central Pacific ENSO skewness by reducing the amplitude of the central Pacific positive SST anomalies and increasing the scaling effect of the eastern Pacific skewness denominator, i.e., ENSO intensity, respectively. The reduction of eastern Pacific El Niño maximum intensity, which is constrained by the SST zonal gradient of the projected background El Niño-like warming pattern, also contributes to a reduction of eastern Pacific SST skewness around the CO2 peak phase. This study highlights the divergent responses of different strong El Niño regimes in response to climate change.

Dynamical Drivers of the Exceptional Warmth over Siberia during the Spring of 2020

Abstract Much of Siberia experienced exceptional warmth during the spring of 2020, which followed an unusually warm winter over the same region. Here, we investigate the drivers of the spring warmth from the perspective of atmospheric dynamics and remote influences, focusing on monthly-time-scale features of the event. We find that the warm anomalies were associated with separate quasi-stationary Rossby wave trains emanating from the North Atlantic in April and May. The wave trains are shown to be extreme manifestations of the dominant modes of spring subseasonal meridional wind variability over the Northern Hemisphere. Using a large ensemble of simulations from NASA’s GEOS atmospheric model, in which the model is constrained to remain close to observations over selected regions, we further elucidate the remote drivers of the unusual spring temperatures in Siberia. In both April and May, the wave trains were likely forced from an upstream region including eastern North America and the western North Atlantic. Analysis with a stationary wave model shows that transient vorticity flux forcing over and downwind of the North Atlantic, which is strongly related to storm activity caused by internal variability, is key to generating the wave trains, suggesting limited subseasonal predictability of the Rossby waves and hence the exceptional Siberian warmth. Our observational and model analyses also suggest that anomalous tropical atmospheric heating contributed to the unusual warmth in Siberia through a teleconnection involving upper-troposphere dynamics and the mean meridional circulation. This tropical–extratropical teleconnection offers a possible physical mechanism by which anthropogenic climate change influenced the extreme Siberian warmth.

Winter Precipitation Responses to Projected Arctic Sea Ice Loss and Global Ocean Warming and Their Opposing Influences over the Northeast Atlantic Region

Abstract Using a large ensemble of simulations from the Polar Amplification Model Intercomparison Project (PAMIP) and phase 6 of the Coupled Model Intercomparison Project (CMIP6), we compare the response of winter-mean precipitation and daily extremes across the Northern Hemisphere to future Arctic sea ice loss and global ocean warming. The North Atlantic is simulated to become drier in response to future Arctic sea ice loss, with reduced precipitation intensity and more dry days. A wetting response to sea ice loss is simulated over the midlatitude Atlantic Ocean. These responses are robust across the eight models analyzed, albeit with differences in their magnitude and spatial pattern. The precipitation response to global ocean warming is broadly opposite in sign, but larger in magnitude, compared to the response to sea ice loss, over these regions. The precipitation responses to both sea ice loss and ocean warming are strongly related to coincident changes in storm density and intensity. More specifically, there is an equatorward shift of the North Atlantic storm track in response to sea ice loss, and a northeastern extension of the North Atlantic storm track and a weakening of the Mediterranean storm track in response to ocean warming. The linear combination of the responses of future Arctic sea ice loss and global ocean warming explain well the spatial pattern of the precipitation change at 2°C global warming projected in CMIP6. Our results suggest that the projected future precipitation change over the North Atlantic reflects a “tug of war” between Arctic sea ice loss and global ocean warming, but the latter dominates over the former.

Quantifying the statistical dependence of mid-latitude heatwave intensity and likelihood on prevalent physical drivers and climate change

Abstract. Recent heatwaves such as the 2021 Pacific Northwest heatwave have shattered temperature records across the globe. The likelihood of experiencing extreme temperature events today is already strongly increased by anthropogenic climate change, but it remains challenging to determine to what degree prevalent atmospheric and land surface conditions aggravated the intensity of a specific heatwave event. Quantifying the respective contributions is therefore paramount for process understanding but also for attribution and future projection statements conditional on the state of atmospheric circulation or land surface conditions. We here propose and evaluate a statistical framework based on extreme value theory, which enables us to learn the respective statistical relationship between extreme temperature and process variables in initial-condition large ensemble climate model simulations. Elements of statistical learning theory are implemented in order to integrate the effect of the governing regional circulation pattern. The learned statistical models can be applied to reanalysis data to quantify the relevance of physical process variables in observed heatwave events. The method also allows us to make conditional attribution statements and answer “what if” questions. For instance, how much would a heatwave intensify given the same dynamic conditions but at a different warming level? How much additional warming is needed for the same heatwave intensity to occur under average circulation conditions? Changes in the exceedance probability under varying large- and regional-scale conditions can also be assessed. We show that each additional degree of global warming increases the 7 d maximum temperature for the Pacific Northwest area by almost 2 ∘C, and likewise, we quantify the direct effect of anti-cyclonic conditions on heatwave intensity. Based on this, we find that the combined global warming and circulation effect of at least 2.9 ∘C accounts for 60 %–80 % of the 2021 excess event intensity relative to average pre-industrial heatwave conditions.

Floods and droughts: a multivariate perspective

Abstract. Multivariate or compound hydrological-extreme events such as successive floods, large-scale droughts, or consecutive drought-to-flood events challenge water management and can be particularly impactful. Still, the multivariate nature of floods and droughts is often ignored by studying individual characteristics only, which can lead to the under- or overestimation of risk. Studying multivariate extremes is challenging because of variable dependencies and because they are even less abundant in observational records than univariate extremes. In this review, I discuss different types of multivariate hydrological extremes and their dependencies, including regional extremes affecting multiple locations, such as spatially connected flood events; consecutive extremes occurring in close temporal succession, such as successive droughts; extremes characterized by multiple characteristics, such as floods with jointly high peak discharge and flood volume; and transitions between different types of extremes, such as drought-to-flood transitions. I present different strategies to describe and model multivariate extremes and to assess their hazard potential, including descriptors of multivariate extremes, multivariate distributions and return periods, and stochastic and large-ensemble simulation approaches. The strategies discussed enable a multivariate perspective on hydrological extremes, which allows us to derive risk estimates for extreme events described by more than one variable.

Robust projection of East Asian summer monsoon rainfall based on dynamical modes of variability

The Asian monsoon provides the freshwater that a large population in Asia depends on, but how anthropogenic climate warming may alter this key water source remains unclear. This is partly due to the prevailing point-wise assessment of climate projections, even though climate change patterns are inherently organized by dynamics intrinsic to the climate system. Here, we assess the future changes in the East Asian summer monsoon precipitation by projecting the precipitation from several large ensemble simulations and CMIP6 simulations onto the two leading dynamical modes of internal variability. The result shows a remarkable agreement among the ensembles on the increasing trends and the increasing daily variability in both dynamical modes, with the projection pattern emerging as early as the late 2030 s. The increase of the daily variability of the modes heralds more monsoon-related hydrological extremes over some identifiable East Asian regions in the coming decades.

Role of the Maritime Continent in the remote influence of Atlantic Niño on the Pacific

Atlantic Niño, the dominant climate mode in the equatorial Atlantic, is known to remotely force a La Niña-like response in the Pacific, potentially affecting seasonal climate predictions. Here, we use both observations and large-ensemble simulations to explore the physical mechanisms linking the Atlantic to the Pacific. Results indicate that an eastward propagating atmospheric Kelvin wave from the Atlantic, through the Indian Ocean, to the Pacific is the primary pathway. Interaction of this Kelvin wave with the orography of the Maritime Continent induces orographic moisture convergence, contributing to the generation of a local Walker Cell over the Maritime Continent-Western Pacific area. Moreover, land friction over the Maritime Continent dissipates Kelvin wave energy, affecting the strength of the Bjerknes feedback and thus the development of the La Niña-like response. Therefore, improving the representation of land–atmosphere–ocean interactions over the Maritime Continent may be fundamental to realistically simulate Atlantic Niño’s impact on El Niño-Southern Oscillation.

Fire weather index data under historical and shared socioeconomic pathway projections in the 6th phase of the Coupled Model Intercomparison Project from 1850 to 2100

Abstract. Human-induced climate change is increasing the incidence of fire events and associated impacts on livelihood, biodiversity, and nature across the world. Understanding current and projected fire activity together with its impacts on ecosystems is crucial for evaluating future risks and taking actions to prevent such devastating events. Here we focus on fire weather as a key driver of fire activity. Fire weather products that have a global homogenous distribution in time and space provide many advantages to advance fire science and evaluate future risks. Therefore, in this study we calculate and provide for the first time the Canadian Fire Weather Index (FWI) with all available simulations of the 6th phase of the Coupled Model Intercomparison Project (CMIP6). Furthermore, we expand its regional applicability by combining improvements to the original algorithm for the FWI from several packages. A sensitivity analysis of the default version versus our improved version shows significant differences in the final FWI. With the improved version, we calculate the FWI using average relative humidity in one case and minimum relative humidity in another case. We provide the data for both cases while recommending the one with minimum relative humidity for studies focused on actual FWI values and the one with average relative humidity for studies requiring larger ensembles. The following four annual indicators, (i) maximum value of the FWI (fwixx), (ii) number of days with extreme fire weather (fwixd), (iii) length of the fire season (fwils), and (iv) seasonal average of the FWI (fwisa), are made available and are illustrated here. We find that, at a global warming level of 3 ∘C, the mean fire weather would increase on average by at least 66 % in duration and frequency, while associated 1-in-10-year events would approximately triple in duration and increase by at least 31 % in intensity. Ultimately, this new fire weather dataset provides a large ensemble of simulations to understand the potential impacts of climate change spanning a range of shared socioeconomic narratives with their radiative forcing trajectories over 1850–2100 at annual and 2.5∘ × 2.5∘ resolutions. The produced full global dataset is a freely available resource at https://doi.org/10.3929/ethz-b-000583391 (Quilcaille and Batibeniz, 2022) for fire danger studies and beyond, which highlights the need to reduce greenhouse gas emissions for reducing fire impacts.

Anthropogenic forcing and Pacific internal variability-determined decadal increase in summer precipitation over the Asian water tower

The increased precipitation in the Asian water tower has prompted the abrupt lake expansion and increased runoff, significantly reshaping the water resource redistribution in the Inner Tibetan Plateau (ITP). However, the dynamic attribution behind this decadal increment remains unclear. Here, analysis of observations, large ensemble simulations, and pacemaker experiments indicates that this decadal increase was mainly attributed to the synergistic effects of the external forcing (anthropogenic greenhouse and aerosol emissions) and the Pacific internal variability, while the Atlantic and the Indian Ocean play a secondary role. Observations and simulations show that thermodynamic and dynamic effects work collaboratively to this increase. Remarkably, the upper-level dynamic convergence over the ITP would be enhanced through teleconnection and atmospheric dynamic feedback when involving the Pacific internal variability, resulting in more precipitation occurrence. Further analyses show that the enhanced stationary Rossby wave propagation over Eurasia and strengthened transient eddy activity over North Pacific could contribute to the anomalous cyclone over the ITP and weakened East Asian westerly jet, which built a pathway for the external forcing and Pacific internal variability collaboratively impacting the decadal increase in precipitation in the ITP. These results can improve our understanding of ITP summer precipitation attribution and can be applied to emergent constraints on future decadal precipitation prediction.

PARC: Physics-aware recurrent convolutional neural networks to assimilate meso scale reactive mechanics of energetic materials.

The thermo-mechanical response of shock-initiated energetic materials (EMs) is highly influenced by their microstructures, presenting an opportunity to engineer EM microstructures in a "materials-by-design" framework. However, the current design practice is limited, as a large ensemble of simulations is required to construct the complex EM structure-property-performance linkages. We present the physics-aware recurrent convolutional (PARC) neural network, a deep learning algorithm capable of learning the mesoscale thermo-mechanics of EM from a modest number of high-resolution direct numerical simulations (DNS). Validation results demonstrated that PARC could predict the themo-mechanical response of shocked EMs with comparable accuracy to DNS but with notably less computation time. The physics-awareness of PARC enhances its modeling capabilities and generalizability, especially when challenged in unseen prediction scenarios. We also demonstrate that visualizing the artificial neurons at PARC can shed light on important aspects of EM thermos-mechanics and provide an additional lens for conceptualizing EM.

Risk-based versus storyline approaches for global warming impact assessment on basin-averaged extreme rainfall: a case study for Typhoon Hagibis in eastern Japan

Two methods exist to address the degree to which past extreme events and associated disasters will be intensified due to climate change: storyline approaches and risk-based approaches. However, the risk-based approach applied to weather similar to the target event (typhoons, a stationary weather front,…etc) becomes theoretically similar to the storyline approach. We examine this theory for the climate change impact of a real event, Typhoon Hagibis, which caused devastating flood damage to eastern Japan in 2019, while focusing on basin-averaged accumulated rainfall (BAAR) in major eastern river basins. A risk-based approach was conducted to determine the future change of BAAR by calculating the quantile change corresponding to Hagibis from the probability distribution of typhoon-induced events in a large ensemble climate simulation dataset database for Policy Decision-making for Future climate change (past, +2K and +4K future climates). A storyline approach for Typhoon Hagibis was realized using a pseudo global warming (PGW) experiment with a 5 km non-hydrostatic model. The projected BAAR in the two approaches were consistent for all target basins, supporting the robustness of the calculated changes in extreme catchment precipitation. This presents an important practical benefit: one can assess future climate change impact on a past symbolic event using either PGW experiments or large ensemble climate projections for the target weather.

Advancing research on compound weather and climate events via large ensemble model simulations

Societally relevant weather impacts typically result from compound events, which are rare combinations of weather and climate drivers. Focussing on four event types arising from different combinations of climate variables across space and time, here we illustrate that robust analyses of compound events — such as frequency and uncertainty analysis under present-day and future conditions, event attribution to climate change, and exploration of low-probability-high-impact events — require data with very large sample size. In particular, the required sample is much larger than that needed for analyses of univariate extremes. We demonstrate that Single Model Initial-condition Large Ensemble (SMILE) simulations from multiple climate models, which provide hundreds to thousands of years of weather conditions, are crucial for advancing our assessments of compound events and constructing robust model projections. Combining SMILEs with an improved physical understanding of compound events will ultimately provide practitioners and stakeholders with the best available information on climate risks.

Future Indian Ocean warming patterns

Most future projections conducted with coupled general circulation models simulate a non-uniform Indian Ocean warming, with warming hotspots occurring in the Arabian Sea (AS) and the southeastern Indian Ocean (SEIO). But little is known about the underlying physical drivers. Here, we are using a suite of large ensemble simulations of the Community Earth System Model 2 to elucidate the causes of non-uniform Indian Ocean warming. Strong negative air-sea interactions in the Eastern Indian Ocean are responsible for a future weakening of the zonal sea surface temperature gradient, resulting in a slowdown of the Indian Ocean Walker circulation and the generation of southeasterly wind anomalies over the AS. These contribute to anomalous northward ocean heat transport, reduced evaporative cooling, a weakening in upper ocean vertical mixing and an enhanced AS future warming. In contrast, the projected warming in the SEIO is related to a reduction of low-cloud cover and an associated increase in shortwave radiation. Therefore, the regional character of air-sea interactions plays a key role in promoting future large-scale tropical atmospheric circulation anomalies with implications for society and ecosystems far outside the Indian Ocean realm.

Projected Atlantic overturning slow-down is to be compensated by a strengthened South Atlantic subtropical gyre

The upper limb of the Atlantic meridional overturning circulation brings shallow interocean contributions to replenish the North Atlantic export of deepwaters. It is primarily formed in the southern South Atlantic where the converging entrainment of Pacific and Indian Ocean waters meet and incorporate into the South Atlantic subtropical gyre. Here, we use Community Earth System Model 1 Large Ensemble simulation results along 1920–2100 to investigate the response of the Atlantic meridional overturning circulation upper limb and the South Atlantic subtropical gyre to future human-induced climate warming under business-as-usual greenhouse gas emissions. In terms of flow redistribution, we find that the Atlantic meridional overturning circulation upper limb weakens not because less waters are being imported from the Pacific and Indian basins — but because waters are being mostly directed to recirculate in the southwestern portion of a distorted South Atlantic subtropical gyre, turning back southward after reaching the South Atlantic western boundary.

Regime-oriented causal model evaluation of Atlantic–Pacific teleconnections in CMIP6

Abstract. The climate system and its spatio-temporal changes are strongly affected by modes of long-term internal variability, like the Pacific decadal variability (PDV) and the Atlantic multidecadal variability (AMV). As they alternate between warm and cold phases, the interplay between PDV and AMV varies over decadal to multidecadal timescales. Here, we use a causal discovery method to derive fingerprints in the Atlantic–Pacific interactions and to investigate their phase-dependent changes. Dependent on the phases of PDV and AMV, different regimes with characteristic causal fingerprints are identified in reanalyses in a first step. In a second step, a regime-oriented causal model evaluation is performed to evaluate the ability of models participating in the Coupled Model Intercomparison Project Phase 6 (CMIP6) in representing the observed changing interactions between PDV, AMV and their extra-tropical teleconnections. The causal graphs obtained from reanalyses detect a direct opposite-sign response from AMV to PDV when analyzing the complete 1900–2014 period and during several defined regimes within that period, for example, when AMV is going through its negative (cold) phase. Reanalyses also demonstrate a same-sign response from PDV to AMV during the cold phase of PDV. Historical CMIP6 simulations exhibit varying skill in simulating the observed causal patterns. Generally, large-ensemble (LE) simulations showed better network similarity when PDV and AMV were out of phase compared to other regimes. Also, the two largest ensembles (in terms of number of members) were found to contain realizations with similar causal fingerprints to observations. For most regimes, these same models showed higher network similarity when compared to each other. This work shows how causal discovery on LEs complements the available diagnostics and statistical metrics of climate variability to provide a powerful tool for climate model evaluation.