It is paramount to determine the critical forcing levels (and the associated uncertainties) beyond which the systems in question will abruptly change their state, with potentially devastating climatic, ecological, and societal impacts. For this purpose, we need to substantially enhance our understanding of the dynamics of the Tipping Elements and their interactions, on the basis of paleoclimatic evidence, present-day observations, and models spanning the entire hierarchy of complexity. Moreover, to be able to mitigate - or prepare for - potential future transitions, early warning signals have to be identified and monitored in both observations and models.
This multidisciplinary session invites contributions that address Tipping Points in the Earth system from the different perspectives of all relevant disciplines, including
- the mathematical theory of tipping points
- methods to anticipate critical transitions from data
- tipping points in climate models across the hierarchy, including comprehensive Earth system models
- climatic, ecological and socioeconomic impacts of tipping events
- decision theory in the presence of uncertain tipping point estimates and uncertain impacts
Tipping points have become a buzzword in earth system science. The more popular the term becomes, the less clear its definition seems. While for some a tipping point is simply a metaphor of something changing quickly, others mean a bifurcation threshold in a strict mathematical sense. Also the Intergovernmental Panel on Climate Change has struggled defining tipping, relying on challenging notions such as abruptness and irreversibility. However, agreeing on what tipping means, and whether a system tips or not, is important both for robust science as well as for communicating climate tipping risk to policymakers and the public.
Here we critically evaluate the problems with existing tipping definitions. Based on this, we propose a revised definition that characterizes tipping behavior as a nonlinear transition in forced systems. Our definition emphasizes both the phenomenology (observed time series) and cause (feedback mechanism) of a tipping event. While compatible with dynamical systems theory, our proposition avoids concepts such as bifurcations or equilibrium states, making the definition applicable also to transient dynamics in highly complex systems under time-varying forcing. We showcase its practical use in case studies of earth system model data, comparing slow tipping systems (e.g. ice sheets) with fast tipping systems (e.g. tropical rainforests).
How to cite: Börner, R. and Dijkstra, H. A.: To tip or not to tip, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18498, https://doi.org/10.5194/egusphere-egu26-18498, 2026.
The future behavioural fate of a forced nonlinear system may at times depend sensitively on the forcing profile as well as natural fluctuations within the system. This is especially the case for rate-induced tipping, where the forcing pushes the system to a basin boundary of a future behaviour and small changes in the forcing can lead to drastically different behaviours. This sensitivity may be present only for a limited time when the forcing is most rapidly changing and so we investigate a geometric early warning to evaluate whether we are in such a sensitive state. This involves computing an approximation of the R-tipping edge state which is a dynamic state that requires knowledge of the future behaviour of the forcing. We contrast this with early warnings of bifurcation-induced tipping, where an assumption of slow variation of forcing is needed. We provide an example of early prediction of future state for a 3-box model of the Atlantic Meridional Overturning Circulation (AMOC) with specified rapid forcing and show that the skill compares favourably with a simple threshold approach.
How to cite: Ritchie, P., Kachhara, S., and Ashwin, P.: Evaluating the skill of a geometric early warning for tipping in a rapidly forced nonlinear system, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10428, https://doi.org/10.5194/egusphere-egu26-10428, 2026.
Critical slowing down has been proposed as an early warning signal for critical transitions in ecosystems (tipping points) and for defining ecological resilience. However these concepts are difficult to define in ecological systems, which limits how they can be used operationally for advanced warning of changes in ecosystem function and composition. Here we use dense time-series satellite measurements of vegetation productivity in UK grasslands, combined with a dynamic linear model, to estimate ecosystem speed and construct stability landscapes: potential-like surfaces that quantify the geometry governing transitions in response to major droughts that provide empirical evidence for resilience concepts in real-world ecosystems. We anticipate the use of our approach to better understand and visualize ecosystem resilience and as a tool for identifying ecosystems in critical transition that can be targets for intervention, such as ecological restoration.
How to cite: Simms, D., Rust, W., Stojanovic, M., Bullock, J., Corstanje, R., and Harris, J.: Stability landscapes: evidencing critical slowing down and ecological resilience in grassland ecosystems, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17794, https://doi.org/10.5194/egusphere-egu26-17794, 2026.
We present TOAD v1.0 (Tipping and Other Abrupt events Detector), an open-source Python framework for the systematic detection and analysis of abrupt shifts in gridded Earth system data. TOAD provides a user-oriented workflow that combines grid-level shift detection with spatio-temporal clustering to identify domains where abrupt change co-occurs. An optional ensemble consensus step then identifies spatial patterns that are robust across ensemble members, models, variables, or method configurations, and quantifies associated statistics. The framework is method-agnostic, allowing different detection and clustering algorithms to be compared within a reproducible analysis pipeline. TOAD serves as an introspection tool for exploratory analysis of abrupt change across scales and addresses key questions in tipping-point research by identifying where such changes occur and providing first-order information on when they emerge along a time or forcing trajectory. The framework supports coordinated analyses in large model ensembles and intercomparison projects, such as TIPMIP and CMIP.
How to cite: Harteg, J., Röhrich, L., De Maeyer, K., Garbe, J., Sakschewski, B., Klose, A. K., Donges, J., Winkelmann, R., and Loriani, S.: Detecting Abrupt Shifts and Coherent Spatial Domains in Earth System Data with TOAD , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18199, https://doi.org/10.5194/egusphere-egu26-18199, 2026.
Classical critical slowing down early warning signals (observing increasing trends in the autocorrelation and variance) have been developed to try and detect when a system is approaching a tipping point, typically represented mathematically by a bifurcation. However, these signals often fail for strongly forced slow systems. Here we propose a new method that reconstructs the quasi-equilibrium state and therefore produces a robust indication of where the critical threshold may lie in a system.
We show that the variance of this reconstructed quasi-equilibrium state increases exponentially ahead of its critical threshold, at time tcrit, for both strongly forced fast and slow systems. Using both the reconstructed quasi-equilibrium state and the inverse of its variance allows us to diagnose the location of the critical threshold for a strongly forced, slow system which has passed the critical threshold but not yet tipped, and where classical critical slowing down indicators often fail.
How to cite: Parry, I., Ritchie, P., and Cox, P.: An early warning indicator for tipping in a strongly forced system, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-1228, https://doi.org/10.5194/egusphere-egu26-1228, 2026.
Early warning indicators (EWIs) of tipping points are typically derived from insights gained from simple dynamical systems. Whether these indicators provide robust and reliable warnings when applied to the climate system remains an open question. In this study, we use climate models with known tipping points to investigate the behavior of classical EWIs associated with increasing memory and variance. In addition, we explore an alternative EWI based on changes in spatial correlations.
A key challenge in applying EWIs is determining when a change is significant enough to constitute a warning. How large must a deviation be relative to background variability? How should this background variability be defined—using an early segment of the simulation or an unforced control experiment? How much temporal smoothing should be applied to the indicators? And what is the associated risk of false positives?
We address these questions for several tipping elements, including the Atlantic Meridional Overturning Circulation (AMOC), the subpolar gyre region, and sea ice. Our analysis is based on simulations from both the CMIP6 ensemble and the OptimESM/TipESM ensembles.
Our results indicate that EWIs are generally sensitive to methodological choices and, in some cases, exhibit significant changes only after the tipping point has occurred.
How to cite: Christiansen, B. and Yang, S.: On the robustness of Early Warning Indicators, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12010, https://doi.org/10.5194/egusphere-egu26-12010, 2026.
Crossing climate tipping points poses a rising risk under continued global warming.
Yet quantitative tipping risk assessments often rely on idealised system dynamics and do not take into account Earth system model (ESM) processes.
Here, we present a process-informed, updatable framework that links systematic stability assessments from comprehensive models to transparent low-order dynamical systems for three high-impact climate tipping elements (TEs): the Atlantic Meridional Overturning Circulation (AMOC), the Greenland Ice Sheet (GrIS), and the West Antarctic Ice Sheet (WAIS).
We assemble TE experiments from Earth system and Earth system component models, fit element-specific dynamical systems with saddle-node bifurcations that map external forcing to state transitions, and run idealised instantaneous-forcing experiments to show the application of our framework.
A simple, modular update protocol allows tipping thresholds and timescales to be revised as new simulations from ESMs become available without refitting the full framework.
Applied to current ESM simulations, our emulators reproduce multistability of the GrIS and WAIS and a freshwater-forced weakening of the AMOC, yielding decision-relevant transient and equilibrium behaviour consistent with the underlying ESMs.
Our approach provides a transparent bridge between comprehensive simulations and risk metrics, and can be extended to additional climate tipping elements as suitable experiments become available.
How to cite: Bochow, N., Krönke, J., Garbe, J., and Wunderling, N.: Emulating tipping elements: Linking Earth system models to low-order dynamics for tipping elements, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-6567, https://doi.org/10.5194/egusphere-egu26-6567, 2026.
The pivotal role in regulating the global climate system and the potential for irreversible collapse underscore the critical importance of the Atlantic Meridional Overturning Circulation (AMOC) and its stability. To address the inherent nonlinearity and stochastic nature of the AMOC, we develop a Convolutional Neural Network (CNN) model to project AMOC evolution using atmospheric and oceanic climate model inputs. Our CNN model successfully captures the stochastic AMOC bifurcation present in large-ensemble climate model simulations. Using explainable AI, we find that the salinity structure enables the CNN to predict future AMOC trajectories, suggesting that the salt-advection feedback amplifies subtle perturbations. Current climate models systematically misrepresent this salinity structure. We show that correcting these biases shifts the climate model projections towards a collapse-prone regime, implying that the AMOC’s stability in current climate models is likely overestimated. Our findings suggest that the risk of AMOC collapse cannot be ruled out simply based on model projections, calling for more thorough investigations of AMOC stability with focus on potential stability biases in climate models.
How to cite: Shin, Y., Boers, N., Huang, Y., Emirzade, B., Ko, J., Oh, J.-H., and Kug, J.-S.: AI emulator highlights underestimated risk of AMOC collapse in current climate models, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-16364, https://doi.org/10.5194/egusphere-egu26-16364, 2026.
The Atlantic Meridional Overturning Circulation (AMOC) is a major component of the global ocean circulation and plays a crucial role in regulating Earth’s climate. Similarly, the Amazon Rainforest (ARF), often referred to as the “lungs of the planet”, is a key regulator of the global carbon and hydrological cycles. Both systems are considered to be climate tipping elements, characterised by the existence of multiple equilibrium states separated by a critical threshold. Anthropogenic climate change is pushing these systems closer to their respective tipping points, potentially leading to an AMOC slowdown or collapse and a transition of the Amazon from a rainforest to a savanna-like state.
While the AMOC and the ARF have been widely studied separately, their potential interactions remain poorly understood. Rising global temperatures are associated with reduced precipitation over the Amazon and a weakening of the AMOC. These coupled changes suggest two-way interactions, as AMOC weakening can decrease rainfall over the Amazon, while changes in Amazon vegetation can affect AMOC strength, potentially stabilising the climate system or triggering a tipping cascade. In this study, we investigate whether changes in the Amazon Rainforest can influence the stability of the AMOC.
We use the Community Earth System Model (CESM) to perform a set of numerical experiments in which the Amazon region is prescribed with different land-cover states, representing rainforest and grassland conditions. By comparing these experiments, we investigate the climate response to prescribed Amazon vegetation changes and their influence on large-scale atmospheric and ocean circulation.
Our results show that replacing rainforest with grassland leads to a global increase in near-surface air temperature, with the strongest warming occurring over the Amazon region. This transition is associated with pronounced changes in precipitation patterns and atmospheric moisture transport. These findings indicate a potential coupling between Amazon land-cover changes and large-scale climate dynamics, suggesting that Amazon tipping may have implications for AMOC stability.
How to cite: Stanchieri, C., Dijkstra, H. A., Bastiaansen, R., De Maeyer, K., Rietkerk, M., and Staal, A.: The interaction between the Amazon rainforest and the AMOC, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8429, https://doi.org/10.5194/egusphere-egu26-8429, 2026.
Ocean and climate models of various complexity have shown that the Atlantic Meridional Overturning Circulation (AMOC) can undergo tipping as a function of freshwater forcing. Most of these model experiments compensate for the freshwater input to conserve global salinity, with salt being added either at the surface or throughout the ocean volume. However, these two different compensation methods have so far only been compared in a single, coarse-resolution climate model, and therefore little is known robustly about the effect of salinity compensation on the AMOC tipping point. Here, using an ocean model at 1° resolution and an intermediate-complexity coupled climate model, we systematically compare the effect of surface vs volume compensation on the tipping point of the AMOC as diagnosed from quasi-equilibrium experiments using a freshwater flux over the region 20°N–50°N.
Salinity compensation at the surface consistently delays AMOC tipping compared to volume compensation. This is mainly because the compensation salinity added over the subpolar North Atlantic counteracts the weakening salinity gradient from freshwater forcing. In contrast, the compensation method does not strongly impact AMOC recovery when tracing the full hysteresis loop. Our results indicate that the distance of present-day climate to the AMOC tipping point with respect to freshwater forcing might have been overestimated in recent modeling studies, compounding the effect of model biases.
How to cite: Mehling, O., Vanderborght, E., and Dijkstra, H. A.: Critical freshwater forcing for AMOC tipping in climate models – compensation matters, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4250, https://doi.org/10.5194/egusphere-egu26-4250, 2026.
The Atlantic Meridional Overturning Circulation (AMOC) has a strong influence on global and regional climate. In particular over Europe, previous works have shown that an AMOC tipping to a weak state would lead to a substantial cooling and drying, especially in the high latitudes. In this work, we aim at identifying the impacts of an AMOC collapse to the atmospheric circulation in the North Atlantic by looking at the jet stream and the storm track and link them with changes in the European climate variability and extremes.
We use dynamically downscaled regional climate simulations at 15 km resolution over a 20-year period run with the Weather Research and Forecasting model (WRF). These regional simulations are forced by fully-coupled global runs performed with multiple Earth System models under various prescribed conditions. The experiments consist of a simulation with stable global warming of 2 K and an imposed AMOC collapse on top of the stable global warming. Cyclones are tracked in the WRF simulations and their number, intensity and associated impacts (strong winds and/or heavy precipitation) in the global warming and AMOC collapse experiments are compared to a baseline simulation with pre-industrial forcing. The physical processes leading to the identified changes in the storm track are also studied. This analysis allows us to better understand the modifications in atmospheric dynamics caused by crossing a tipping point and to quantify the subsequent impacts.
How to cite: Hermoso, A. and Raible, C.: Impacts of an AMOC collapse on the North Atlantic jet stream and storm track, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9785, https://doi.org/10.5194/egusphere-egu26-9785, 2026.
The Atlantic surface ocean currents, connecting the atmosphere and the deep ocean currents like the Atlantic Meridional Overturning Circulation (AMOC), plays a central role in regulating Earth’s climate. Yet how large-scale surface currents respond to ongoing climate change remains poorly constrained. Here we identify a previously unrecognized phase of Atlantic surface circulation, termed the Atlantic Convergence–Divergence Mode (ACDM), characterized by a convergence–divergence structure in the North Atlantic and coherent meridional flows in the South Atlantic. We find that the ACDM has experienced a transition evidenced by a systematic weakening of vertical water exchange and meridional flows, with its interannual variability marking a regime shift in 2009, consistent with the RAPID-MOCHA AMOC observations. Our analysis indicates that this shift is driven by AMOC-modulated ocean-atmosphere interactions, including the North Atlantic Oscillation (NAO) and layered ocean heat transport. We therefore propose the ACDM’s interannual variability as a more sensitive proxy for AMOC at interannual timescales, revealing the coexistence of a gradual multidecadal decrease trend and an abrupt interannual shift in AMOC variability. Moreover, this step-like shift in AMOC also suggests the important role of atmospheric disturbances and reveal that AMOC may be more delicate and closer to tipping point than than previously anticipated. These findings confirm that AMOC variability can trigger rapid, large-scale transitions in surface circulation, pushing it into a new, weaker phase.
How to cite: Huang, H., Tao, N., Wang, H., Liu, T., Xie, F., Li, X., Zhang, Y., Boers, N., Fan, J., Chen, D., and Chen, X.: Phase Transition in the Atlantic Surface Currents, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4350, https://doi.org/10.5194/egusphere-egu26-4350, 2026.
During the last ice-age, temperatures in Greenland have frequently increased and decreased by 10°C. Detailed studies with climate models suggest that this is caused by collapses and recoveries of the Atlantic Meridional Overturning Circulation (AMOC). The causes of these AMOC transitions are still debated, though. Here we describe the development of a neural-network based surrogate model of the AMOC. It is trained on 32,000 years of climate model integrations to build a set of stochastic differential equations that emulate the climate models' AMOC behavior. In particular it reproduces the spectra and the asymmetry in the times it takes for the AMOC to recover and collapse, which makes it more realistic than previously published sets of coupled differential equations to study past AMOC transitions. Monte Carlo simulations with this model show that collapses are deterministic, but recoveries are stochastically forced, in partial support of the leading hypotheses surrounding the AMOC transitions.
How to cite: Wu, Q., Häfner, D., Nuterman, R., Vettoretti, G., and Jochum, M.: Deterministic Collapse and Stochastic Recovery in a Data-Driven Model of the Atlantic Meridional Overturning Circulation, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4473, https://doi.org/10.5194/egusphere-egu26-4473, 2026.
The permafrost soils in the northern high latitudes contain about twice as much carbon as the atmosphere. This organic soil matter has accumulated over many thousands of years and is now exposed to anthropogenic warming, which is amplified by a factor of three to four in the Arctic compared to global warming. The thawed organic matter is mineralized and released into the atmosphere as CO2 or CH4, which amplifies ongoing warming (permafrost carbon feedback). In addition, the thawing of permafrost soils leads to changes in land surface hydrology and potential drainage, which could also amplify global warming due to the decrease in summer cloud cover (permafrost cloud feedback). The permafrost changes impact other regions and Earth tipping elements, including tropical forests.
Are permafrost feedbacks nonlinear, is there a threshold for global warming above which the feedbacks lead to disproportional increase in carbon thaw? Future projections using Earth system and land surface models suggest a rather linear permafrost response to global warming, but they are mostly based on gradual thawing processes and do not take into account abrupt thawing and extreme events. Numerous processes that lead to abrupt thawing at the local level, such as thermokarst, lake formation and drainage, or surface subsidence, have been neglected in large-scale models to date. We will present the results of model experiments and discuss the potential impact of these missing processes on the nonlinear response, as well as indicators of multistability of carbon and hydrology at different scales. We will also discuss irreversibility of permafrost changes and their response timescales as supported by paleo evidence.
How to cite: Brovkin, V., Kleinen, T., de Vrese, P., and Bartsch, A.: Permafrost as a tipping element in the Earth System: scales matter, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4958, https://doi.org/10.5194/egusphere-egu26-4958, 2026.
The cryosphere is considered a potential tipping element of the Earth system, yet identifying critical thresholds and early warning signals remains challenging due to the computational cost of high-complexity ice-sheet models. Here, we investigate tipping behaviour in the glacier periphery of Greenland using the Open Global Glacier Model (OGGM), exploiting its efficiency to explore a wide range of climate forcing scenarios.
We perform a large ensemble of simulations driven by (i) standard future climate scenarios up to 2100, (ii) scenario-neutral climate trajectories, and (iii) additional physically realistic and idealized forcing experiments. This ensemble approach enables systematic exploration of glacier responses across a broad forcing space, including regimes not typically sampled by comprehensive ice-sheet models. We analyze the resulting time series of glacier volume, area and mass balance for early warning signals of critical transitions, focusing on indicators of critical slowing down such as increasing variance and autocorrelation.
By comparing the emergence and robustness of these signals across forcing types, we assess whether abrupt and potentially irreversible retreat of peripheral Greenland glaciers is preceded by detectable changes in system dynamics. We further identify climatic thresholds and boundary conditions under which tipping-like behaviour is most likely to occur.
Our results aim to provide physically grounded constraints on critical forcing levels relevant to Greenland’s glacier periphery and to inform ice-sheet modelling efforts by highlighting regions of parameter space where nonlinear responses are expected. This work demonstrates the value of intermediate-complexity glacier models for advancing the detection and interpretation of tipping points and early warning signals in the cryosphere.
How to cite: van der Laan, L. N., Chapman, R. R., Gu, P., Elvira, V., and Quintanilla, A.: Early warning signals and tipping behaviour in Greenland’s glacier periphery from large forcing ensembles, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20432, https://doi.org/10.5194/egusphere-egu26-20432, 2026.
Many parts of the Earth system are thought to have multiple stable equilibrium states, with the potential for catastrophic shifts between them. Common methods to assess system stability require stationary (trend- and seasonality-free) data, necessitating error-prone data pre-processing. Here, we use Floquet Multipliers to quantify the stability of periodically-forced systems of known periodicity (e.g., annual seasonality) using diverse data without pre-processing. We demonstrate our approach using synthetic time series and spatio-temporal vegetation models, and further investigate two real-world systems: mountain glaciers and the Amazon rainforest. We find that glacier surge onset can be predicted from surface velocity data and that we can recover spatially explicit destabilization patterns in the Amazon. Our method is robust to changing noise levels, such as those caused by merging data from different sensors, and can be applied to quantify the stability of a wide range of spatio-temporal systems, including climate subsystems, ecosystems, and transient landforms.
How to cite: Smith, T., Morr, A., Bookhagen, B., and Boers, N.: Predicting Instabilities in Transient Landforms and Interconnected Ecosystems, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-3854, https://doi.org/10.5194/egusphere-egu26-3854, 2026.
In view of ongoing climate and land use change, the resilience of tropical forests is crucial for maintaining ecosystem services and preventing potential forest loss and associated greenhouse gas emissions. Recent studies have attempted to quantify tropical forest resilience changes using different satellite vegetation products. However, it has not been assessed to what extent the data satisfies the theoretical assumptions of the employed resilience metrics. Here, we propose a framework to determine the most reliable combination of vegetation data and metrics to monitor resilience from space. We apply our framework to select the best combinations from 16 vegetation products and nine resilience metrics. Based on these, we consistently find that tropical forests in South America, Africa, and Asia have experienced a decline in resilience over recent decades. Our robust assessment of resilience changes in tropical forests has important implications for targeted actions to prevent further tropical forest loss.
How to cite: Blaschke, L., Bathiany, S., Hirota, M., and Boers, N.: Resilience Loss of Tropical Forests in Recent Decades, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5042, https://doi.org/10.5194/egusphere-egu26-5042, 2026.
The Amazon rainforest (ARF) is one of the most important tipping elements on the planet and has been under increasing stress due to global warming and deforestation over the last few decades. Several studies have investigated the stability of the ARF, and the best available estimates indicate that the forest could surpass a tipping point after approximately 25% deforestation or 2–4 °C of global warming. However, there are significant uncertainties surrounding these estimates, and there is still a lack of understanding of how these different forcings (global warming and deforestation) interact and how their combined effects could accelerate a critical transition of large parts of the ARF to a tropical savanna state. In this study, we performed idealized experiments to investigate the hydroclimatic response of the ARF to deforestation-only and composite (global warming + deforestation) forcings, with the aim of better understanding the tipping point potential, how it affects the hydroclimatic stability of the forest, and how the two forcings interact. We therefore performed experiments with 0%, 25%, 50%, 75%, and 100% deforestation under a stable global temperature 2 °C warmer than pre-industrial conditions and analysed the response of four hydroclimatic stress indicators: annual precipitation, dry season length (DSL), mean climatological water deficit (MCWD), and top 10 cm soil moisture. We calculated the deviation of the long-term average of each of these variables from stable (no-deforestation) scenarios and classified the deviations using relative anomalies, defined with respect to the standard deviation of the distribution of the stable scenarios. Using these metrics, an anomaly of 0.5 (i.e., a deviation of the mean by half of the standard deviation of the stable scenario) qualifies as a significant moderate anomaly, while anomalies exceeding 2.0 and 2.5 are classified as severe and extreme anomalies, respectively. We found that deforestation of 25% of the ARF can expose approximately 70% of the rainforest to significant hydroclimatic stress according to at least one of the four indicators analysed. When the combined effects of 25% deforestation and 2 °C of global warming are considered, this fraction increases to 89% of the ARF. This composite effect is larger than the deforestation-only stress resulting from a 50% removal of forest cover (82% of the ARF). Among the indicators, increasing dry season length is the most pronounced response, affecting a larger fraction of the forest and with greater intensity than the other variables, and with effects not limited to the vicinity of deforested areas. Whether tree mortality occurs in areas under significant stress is likely to depend strongly on the ability of the vegetation to rapidly adapt to substantial changes in hydroclimatic conditions. Overall, these results reveal quantitative aspects of the potential for deforestation and global warming to trigger cascading effects in the ARF that could ultimately lead to its transition to a tropical savanna state.
How to cite: Ferreira Correa, L., Bathiany, S., and Pongratz, J.: Amazon rainforest hydroclimatic stress under 2 °C global warming in response to progressive idealized deforestation simulated with MPI-ESM-HR, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12025, https://doi.org/10.5194/egusphere-egu26-12025, 2026.
Humanity is exerting unprecedented pressure on the Amazon forest through global warming, deforestation, land-use change, and large-scale infrastructure projects. As the Amazon may exhibit a tipping point beyond which detrimental changes become self-propelling, these pressures could trigger system-wide state shifts. We use a dynamical systems model to assess local transition risks and cascading transitions across the Amazon under different SSP-scenarios (SSP2-4.5, SSP3-7.0 and SSP5-8.5). For each scenario, atmospheric moisture transport is derived throughout the 21st century using an established moisture-tracking model.
In the absence of deforestation, we identify a critical global warming threshold of 3.7-4.0 °C, beyond which around one third of the Amazon loses stability. When deforestation is included, however, our simulations indicate a near system-wide transition (62-77% of the forest area) at global warming levels of 1.5-2.0°C combined with 20-30% deforestation across the basin. Most transitions are driven by drought-induced knock-on effects, causing long-range cascading impacts through the lack of atmospheric moisture recycling. Overall, our results highlight the need to limit warming to as close to 1.5 °C as possible and, halt deforestation at current levels (~17% across the basin), while ideally restoring degraded areas to reduce transition risks across the Amazon forest.
How to cite: Wunderling, N., Sakschewski, B., Rockström, J., Flores, B. M., Hirota, M., and Staal, A.: Pinpointing Amazon forest tipping in global warming and deforestation pathways, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4490, https://doi.org/10.5194/egusphere-egu26-4490, 2026.
The intrinsic variability of the Gulf Stream (GS), the abrupt transition that can occur at a certain tipping point and the early warning signals that precede it, are investigated through a process modeling study. The nonlinear reduced-gravity shallow water equations are solved with schematic but quite realistic geometric configuration and time-independent wind forcing. A reference simulation shows a GS with correct mean Florida Current (FC) transport, realistic separation at Cape Hatteras (CH) due to inertial overshooting, realistic northern recirculation gyre and a strong chaotic intrinsic variability.
To simulate the effect of anthropogenic forcing, a sensitivity analysis is performed by decreasing the forcing amplitude. The results show a gradual shift of the GS toward the coasts north of CH and, therefore, a connection between the decline of the FC and one of the most significant fingerprints of a weakened Atlantic Meridional Overturning Circulation (AMOC). Furthermore, at a tipping point there is an abrupt transition to a GS flowing along the coasts north of CH; this is preceded by an early warning characterized by the fluctuation-dissipation relation, which is revealed by an increase in the autocorrelation and variance of the signals. It is suggested that such critical behavior could impact the AMOC tipping element. As regards the intrinsic variability of the Mediterranean Sea circulation, preliminary results are presented based on ensemble simulations using the FESOM-C finite element model. This research was partially supported by the INVMED project funded by the Italian PRIN-2022.
How to cite: Pierini, S.: Intrinsic oceanic variability, tipping points and early warning signals: examples from the dynamics of the Gulf Stream and the Mediterranean Sea , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-2849, https://doi.org/10.5194/egusphere-egu26-2849, 2026.
The Tibetan Plateau (TP), known as the "Asian Water Tower," is currently undergoing a rapid wetting trend. While this moisture increase is commonly viewed as beneficial for water availability, it remains unclear whether the hydrological system itself is becoming more resilient, and whether continued warming could push it toward instability. Here, we apply an entropy-based framework to quantify the changing structural organization of the TP's soil moisture system. We show that from 2000 to 2024, regional wetting has driven a long-term decline in entropy, reflecting an increase in system order and stability due to enhanced hydrological buffering capacity. This stability is modulated by the El Niño-Southern Oscillation (ENSO), which regulates regional heterogeneity via a distinct spatial dipole. Crucially, however, CMIP6 climate projections reveal an alarming reversal: entropy increases under continued warming and regional contrasts intensify, with some models exhibiting an abrupt mid-century transition. Our findings suggest that while current wetting provides a stabilizing buffer, continued warming is projected to amplify spatial heterogeneity, thereby destabilizing the Asian Water Tower, with significant risks for downstream water security.
How to cite: Xie, Y., Liu, T., Ma, X., Lyu, Y., Wang, X., Qian, Y., Zhang, Y., Wang, M., and Chen, X.: Warming-driven rise in soil moisture entropy signals destabilization of the Asian Water Tower, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-3250, https://doi.org/10.5194/egusphere-egu26-3250, 2026.
The polar regions are critical components of complex Earth systems, housing potential tipping elements such as the West Antarctic Ice Sheet and sea ice systems. However, the climate trajectories of the two poles have diverged significantly over the past century. While the Arctic has exhibited rapid warming and dramatic sea ice loss—a phenomenon known as Arctic amplification—the Antarctic has shown a delayed and more heterogeneous response. Observations indicate that prior to the late 1980s, parts of Antarctica experienced warming and moistening while the Arctic remained relatively stable; subsequently, this pattern reversed, with the Arctic undergoing accelerated change while the Antarctic trend slowed or displayed spatial variability. Understanding the drivers of polar climate variability is paramount for anticipating potential abrupt transitions or tipping points in the regions.
Here, we identify a robust internal variability mode in atmospheric water vapor—termed the Interdecadal Bipolar Oscillation (IBO)—that provides a physical explanation for these historical asymmetries. Using the eigen microstate theory on ERA5 reanalysis and CMIP6 simulations (historical, piControl, and SSPs), we reveal that the IBO links the Arctic and Antarctic in a quasi-periodic (60–80 years) seesaw pattern. We demonstrate that the IBO has modulated interdecadal asymmetries in polar climate change over the past 80 years. Specifically, a phase shift in the late 1980s accelerated Arctic moistening while suppressing similar changes in the Antarctic.
Crucially, our projections under various Shared Socioeconomic Pathways (SSPs) indicate an imminent IBO phase reversal in the coming decades. This transition is expected to shift the IBO from a dampening to an amplifying phase for the Antarctic, coinciding with the background global warming signal. We suggest that this alignment could trigger a regime shift toward rapid Antarctic moistening and warming, potentially destabilizing the ice sheet–atmosphere interactions. The IBO thus acts as a critical internal regulator that may modulate the distance to tipping points in the polar climate system. By elucidating the interplay between this internal oscillation and external anthropogenic forcing, our study offers new insights into the mechanisms that could precipitate abrupt climate transitions in the Antarctic.
How to cite: Wang, H., Fan, J., Xie, F., Li, J., Shi, R., Xia, Y., Chen, D., and Chen, X.: The Interdecadal Bipolar Oscillation: A Potential Driver for Rapid Antarctic Climate Transitions, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-7606, https://doi.org/10.5194/egusphere-egu26-7606, 2026.
Terrestrial vegetation and carbon storage may exhibit irreversible responses under anthropogenic climate change and may shift from a carbon sink to a carbon source. Understanding the reversibility of this transition is critical for assessing future carbon-climate feedback. Idealized overshoot experiments provide a controlled framework to test the response of the terrestrial vegetation and its carbon pools to climate forcing. Non-linear responses and incomplete recovery indicate tipping-like behavior. Here, we quantify land carbon changes under idealized overshoot scenarios, following the Tipping Point Model Intercomparison Project (TIPMIP) protocol, employing the LPX Dynamic Global Vegetation Model. We find non-linear responses in land carbon during and after transient warming, as well as recovery behavior. We perform a transient experiment (T1A1 protocol) initialized from pre-industrial conditions (1850) and force LPX with temperature, precipitation, and cloud cover from TIPMIP simulations of the GFDL ESM2M. In the T1A1 protocol, surface air temperature increases linearly by 2 °C over 100 years, remains constant for 50 years, and then decreases by 2 °C over the next 100 years. To assess the long-term recovery in land carbon, we extend the experiment by 1,000 years beyond the T1A1. During the warming phase global total land carbon decreases by about 100 PgC (~5%), comprising losses from vegetation (35 PgC), soil (25 PgC), permafrost (25 PgC), and litter (15 PgC) carbon pools. During the cooling phase back to pre-industrial conditions, approximately 75% of the carbon loss (75 PgC) is restored. The remaining 25% of the deficit reflects a quasi-permanent loss of permafrost carbon associated with warming-induced thaw. Pronounced non-linear responses emerge in northern peatlands, that suggest tipping-like behavior. Ongoing analysis will further constrain the role of hydrology in shaping these responses and their limited reversibility.
How to cite: Kim, H., Pöppelmeier, F., Stocker, B. D., Hofmann Elizondo, U., and Frölicher, T. L.: Irreversible Land Carbon Losses under Idealized Overshoot Experiments, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9438, https://doi.org/10.5194/egusphere-egu26-9438, 2026.
Climate tipping points emerge from nonlinear feedbacks that can trigger abrupt and potentially irreversible transitions in the Earth system, with far-reaching societal and environmental consequences. Anticipating such critical transitions in complex, high-dimensional systems remains a central challenge. While traditional early-warning indicators rely on assumptions of stationarity, long time series, and simple bifurcation structures. Because these assumptions rarely hold in real data, machine learning approaches have emerged as an alternative, but typically require training data from the specific systems under study, limiting their generalizability.
Here, we present TipBox and TipPFN. TipBox is an open-source, JAX-based repository containing a collection of simple dynamical systems and box models designed for accelerated generation of synthetic data. It enables efficient simulation of deterministic and stochastic systems exhibiting a wide range of bifurcation behaviour such as fold, Hopf, rate- and noise-induced tipping. Since TipBox is differentiable out-of-the-box, it enables easy parameter sensitivity tests for tipping point studies especially when different box models are coupled together.
Building on this synthetic data foundation, we develop TipPFN, a Prior-Data Fitted Network (PFN) approach based on a transformer machine learning architecture that performs approximate Bayesian inference via in-context learning. Trained on carefully selected synthetic dynamical systems, during inference it conditions on a short context of noisy observed data and produces a probabilistic forecast in a single forward pass based on synthetic priors generated from TipBox. This enables fast and computationally cheap probabilistic prediction on systems not seen during training, including time-to-tip as well as the type of tipping point.
We validate our approach on three systems spanning different domains: an AMOC box model representing climate tipping elements, a predator-prey system from ecology, and a simplified power-grid model from infrastructure research. Preliminary results indicate that our PFN-based predictor generalizes to these complex test cases despite being trained exclusively on the simpler systems in TipBox. Benchmarking against state-of-the-art machine learning approaches shows promising results. We observe improved performance over traditional variance- and autocorrelation-based EWS, particularly under noisy conditions. Ongoing work evaluates conditional probabilistic predictions of the effects of changes in forcing on tipping dynamics.
Overall, we show that TipBox and TipPFN enable robust inference of tipping points on previously unseen systems with models trained purely on synthetic data without the need for additional retraining. This capability is especially powerful for the climate system where direct real-world observations of crucial tipping elements are unavailable but their prior proxies are.
How to cite: Herdeanu, B., Nathaniel, J., Ueltzhöffer, K., Roesch, C., Weber, T., Sevinchan, Y., Laschos, V., Ramien, G., Haux, J., and Gentine, P.: TipPFN and TipBox: Early tipping point detection using in-context learning, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12168, https://doi.org/10.5194/egusphere-egu26-12168, 2026.
In the study of nonlinear dynamics, tipping points have been a major focus of research. They describe a threshold where gradual changes in external forcing, e.g. increasing CO2 emissions can lead to an abrupt and persistent transition. This is a concern in ocean sciences due to the strong coupling of the ocean and climate. The capability to provide an early warning of a tipping element is desirable, providing time to mitigate and adapt.
Specific regions of the ocean are of more concern for potential tipping elements than others, a key region being the sub polar gyre. This is a basin-scale cyclonic gyre in the North Atlantic, driven by wind and buoyancy forcing. It is crucial in the formation of North Atlantic Deep Water, a main contributor to the lower branch of the Atlantic Meridional Overturning Circulation (AMOC). If this deep convection process collapses, then cascading changes in sea ice, atmospheric circulation, ocean circulation and sea level, and the terrestrial ecosystem are expected.
Machine Learning approaches suggest that a generalised deep learning (DL) model could potentially provide a robust and high confidence solution to predicting tipping points. We have applied an existing generalised DL model to a large ensemble of historical and future climate projections (1950-2100) based on the HadGEM3 Atmosphere-Ocean-Sea-Ice-Land model under the SSP370 future shared socioeconomic pathway scenario. Using change point analysis to identify tipping points in this ensemble, the DL was provided with timeseries of several parameters, leading up to but not including the identified tipping points. We then assessed the ability of the DL to predict the tipping points based on the chosen parameter timeseries across a number of specific geographic regions.
Mixed Layer Depth and Sea Surface Height were the most effective parameters and there was a large variation in the effectiveness across different regions, with some (Labrador Sea) being much better than others (Irminger Sea). While current DL models are not yet capable of robust tipping point detection there is clear promise in continuing to refine this method with new DL models specifically created for ocean surface and subsurface parameters, such as MLD and temperature and salinity depth profiles.
How to cite: Prime, T. and Sinha, B.: Early Warning of Ocean Tipping Points: A Deep Learning model approach, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17081, https://doi.org/10.5194/egusphere-egu26-17081, 2026.
Escalating climate extremes and environmental pressures are increasingly pushing ecosystems toward regime shifts across the biosphere. Before tipping, however, ecosystems typically lose resilience and exhibit slower recovery from perturbations. This “critical slowing down” behaviour can be quantified through statistical measures to detect early warning signals (EWSs) that arise in timeseries as rising variance, autocorrelation, and skewness. In the spatial domain, the emergence of regular vegetation patterns, known as Turing patterns, has been proposed as a spatial EWS that, along with increasing spatial variability, is interpreted as a hallmark of an approaching critical transition. Nonetheless, a broader gradual spatial reorganisation may instead reflect an ecosystems potential to avoid abrupt collapse by undergoing a Turing bifurcation, experiencing progressive changes in resilience while still drifting towards a less desirable alternative regime. These contradictory theories of what EWSs mean for ecosystems make it crucial to assess whether spatial variability in vegetation emerges as a distinctive sign of an ecosystems’ change in resilience following non-catastrophic shifts. We apply a theoretically grounded framework that links alternative tipping archetypes, such as fold and Turing bifurcations, to expected signatures in both temporal and spatial indicators. By comparing the observed multiscale spatiotemporal patterns with the expectations of different tipping archetypes, we aim to disentangle the scale dependence of tipping dynamics to identify the dominant forcing-response mechanisms operating in complex terrestrial ecosystems. Among these, the Canadian boreal forest stands as a crucial carbon stock exposed to rapid and pronounced warming that renders it vulnerable to structural and compositional transformation. Monitoring it is therefore essential to improve our understanding of the underlying physical mechanisms driving vegetation dynamics and addressing the potential impacts on the ecosystem services they provide if transitioning into degraded states. To achieve this, satellite derived vegetation indices were employed to quantify spatial patterns across multiple resolutions, using hexagonal discrete global grids to characterise vegetation changes in resilience and spatial structure through time. Within a case study region in northern Quebec, the boreal forest exhibited a greening trend from 1984 to 2022 evidenced by an increasing mean of the vegetation indices across scales, but, with an overall decreasing trend in spatial autocorrelation and contrasting trends in spatial variance throughout the region. The observed trends were predominantly prominent at the forest community scales, suggesting a potential scale dependence of the operating tipping archetype and its associated EWS metrics. Datasets on climatic drivers and catastrophic disturbances, including wildfires, droughts, pathogen, and insect outbreaks, are further incorporated to distinguish exogenous forcing from endogenous ecosystem responses and feedbacks. Identifying these could elucidate whether gradual spatial reorganisation or an impending critical transition in vegetation is occurring. The outcomes could further provide actionable insights to support and improve the management, restoration, and mitigation strategies for forested ecosystems under accelerating climate change.
How to cite: Murad, C., Dash, J., Dearing, J., Brummitt, N., and Eigenbrod, F.: Scale Dependence of Spatial Patterns and Tipping Dynamics in the Boreal Forest, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18335, https://doi.org/10.5194/egusphere-egu26-18335, 2026.
How to cite: Mascolo, V., Schiemann, R., and Dittus, A.: Impacts of the Subpolar Gyre weakening on European climate extremes , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-21078, https://doi.org/10.5194/egusphere-egu26-21078, 2026.
Escalating climate crises, characterized by rising sea levels, alongside excessive groundwater pumping, have severely exacerbated saltwater intrusion, posing a critical threat to coastal aquifers. These combined environmental stressors induce complex, non-linear dynamics in groundwater systems, making the exact prediction of regime shifts driven by tipping points increasingly challenging. To address these uncertainties, this study proposes a comprehensive data-driven approach designed to identify early warning signals (EWS) for approaching tipping points using Electrical Conductivity (EC) time-series data. The primary objective is to investigate the feasibility of utilizing complementary statistical indicators—Variance and Fisher Information (FI)—to assess system instability. We analyzed monitoring data from Incheon and Jeju, South Korea, to validate whether these metrics can effectively filter noise and detect genuine precursor signals. Empirical results demonstrate that our approach achieves significantly enhanced performance in distinguishing critical transitions compared to single-indicator methods. Ultimately, this study serves as a foundational step towards establishing an "Integrated Machine Learning" framework. By validating these statistical metrics as key features, we aim to incorporate them into advanced learning algorithms to further improve the robustness and predictive accuracy of coastal groundwater management systems against climate-induced risks.
Acknowledgement
This work was supported by National Research Foundation of Korea(NRF) grant funded by the Ministry of Science and Technology (RS-2024-00356786).
How to cite: Heo, E., Kim, S., and Park, J.: A Comprehensive Data-Driven Approach for Detecting Regime Shifts in Coastal Groundwater: Towards an Integrated Machine Learning Framework, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4647, https://doi.org/10.5194/egusphere-egu26-4647, 2026.
Predicting climate tipping points, such as the collapse of the Atlantic Meridional Overturning Circulation (AMOC), remains a formidable challenge due to the absence of direct observation records of such events and significant parametric and structural uncertainties in Earth System Models (ESMs). It is the grand challenge to explore whether climate tipping or its risk is foreseeable using ESMs and observation without any directs records of climate tipping in the past.
We performed Observation System Simulation Experiments (OSSE) using one of Earth system models of intermediate complexity, LOVECLIM. To assess the predictability of AMOC tipping under a freshwater hosing scenario, we employed a surrogate model-based uncertainty quantification approach to estimate five uncertain parameters related to atmospheric and oceanic physics. We introduced a new dimensionality reduction technique, Wasserstein GEV PCA, which maps the trends of mean climate and extreme events into a flattened statistical manifold using the Wasserstein metric. This allows for the quantification of observation errors and likelihoods even under transient climate conditions.
Our results demonstrate that while parameters related to precipitation adjustment are critical for accurate AMOC projections, they are difficult to constrain. Comparative analysis reveals that among single-variable observations, Sea Surface Salinity is the most effective constraint for reducing the parametric uncertainty including the precipitation adjustment parameters and narrowing the projection spread of the AMOC. Furthermore, while standard mean-field PCA methods exhibit significant estimation errors when applied to non-stationary data the proposed GEV-based method maintains high robustness and estimation accuracy even with the nonstationary observation. This study highlights that tracking the geometry of extreme value distributions provides a superior pathway for non-stationary climate data, thereby enabling more reliable risk assessments of climate tipping.
How to cite: Kubo, A. and Sawada, Y.: Uncertainty Quantification of an Earth System Model for risk assessment of Climate Tipping via Non-stationary Extreme Events, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9125, https://doi.org/10.5194/egusphere-egu26-9125, 2026.
Assessing the spatial patterns and drivers of terrestrial ecosystem resilience is essential for understanding ecosystem responses to climate change and other environmental pressures. In this study, we investigate global vegetation resilience using long-term solar-induced chlorophyll fluorescence (SIF) observations from two independent satellite products. Resilience is quantified using metrics derived from lag-one autocorrelation (AC1) and variance within the framework of critical slowing down theory (CSD). We first evaluate the reliability of SIF-based resilience metrics by comparing them with empirically estimated recovery rates and infer that the SIF datasets are suitable for CSD-based resilience estimates. We further examine how climatic conditions and vegetation structural properties regulate and shape spatial variations in ecosystem resilience. Specifically, we find that water availability and canopy structural complexity show a positive relationship with vegetation resilience, whereas temperature shows a negative relationship with vegetation resilience. In addition, alpha diversity is positively related to resilience across most vegetation types, although this relationship is weak or absent in grassland ecosystems. These findings confirm the importance of climatic controls while highlighting the combined roles of biodiversity and ecosystem structural complexity in shaping terrestrial vegetation resilience. The resilience spatial patterns and mechanisms identified here provide new insights into ecosystem stability under ongoing climate change.
How to cite: Diao, C., Bathiany, S., Blaschke, L. L., Behera, S., Liu, T., Wu, X., Wang, P., Smith, T., and Boers, N.: Vegetation resilience is linked to moisture availability, temperature, biodiversity and canopy complexity, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12309, https://doi.org/10.5194/egusphere-egu26-12309, 2026.
Tropical dry forests are important for biodiversity and people, yet also exposed to high pressure from agricultural expansion and climate change, raising concerns about declining forest resilience. Global analyses of forest resilience have revealed a widespread loss of resilience in tropical and arid forests, as well as declining and increasingly more variable rainfall in many tropical regions. However, tropical dry forests have so far not been explicitly focused on in assessments of resilience under climate and land-use change. A key limitation of existing studies, which have predominantly relied on MODIS-based proxies such as the Normalised Difference Vegetation Index or VODCA-based vegetation optical depth, is that the comparatively coarse spatial resolution of these sensors cannot adequately resolve and analyse small-scale resilience changes in tropical dry forests, which are characterised by high compositional and structural heterogeneity.
Here, we address these gaps and track changes in the resilience of South American tropical dry forests using the high-resolution Landsat archive since 1999. We derive vegetation resilience trends using robust regression based on temporal and spatial indicators of critical slowing down of kNDVI time series. Additionally, we explore how local resilience trends are associated with accessibility and land use gradients, such as distances to roads, cities, and agricultural areas, as well as the hydrological context, such as distances to water surfaces. Our results show distinct patterns of resilience change in South American tropical dry forests. Nonetheless, substantial uncertainties in resilience tracking using the Landsat archive remain due to sensor inconsistencies and missing data.
How to cite: Schulte-Fischedick, A., Liu, T., Blaschke, L., Smith, T., Bathiany, S., Boers, N., and Kuemmerle, T.: Satellite-based Tracking of Resilience Change in South American Tropical Dry Forests, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14948, https://doi.org/10.5194/egusphere-egu26-14948, 2026.
Terrestrial vegetation has continued to green in recent decades under warming and rising atmospheric CO2, yet ecosystem resilience has declined across large portions of the globe. The extent, spatial patterns, and drivers of this emerging decoupling between vegetation greening and resilience remain poorly understood. Here, we assess the dynamics of vegetation greenness and resilience, and disentangle the underlying mechanisms that drive their emerging decoupling using multiple satellite-derived and modelled data. We show that a pervasive pattern of vulnerable greening, characterized by increasing greenness but declining resilience, affects 41.5% of global vegetated land. This pattern is primarily driven by recent changes in temperature and water availability, which exert distinct impacts on vegetation greenness and resilience. Rising temperature generally enhances vegetation greening, but leads to a persistent decline in resilience, especially in tropical and boreal forests. Variability in water availability dominates resilience loss over 23.0-42.1% of vulnerable greening area across vegetation types, whereas its influence on greenness is negligible. Current Earth system models fail to capture the resilience dynamics, yielding systematically underestimated resilience but overly optimistic trends. Our findings reveal a pervasive hidden erosion of ecosystem stability beneath the apparent greening, highlighting growing risks to the terrestrial carbon sink, and the urgent need to better represent vegetation resilience in climate-change assessments.
How to cite: Zhao, M., Shangguan, Y., and Shi, Z.: Widespread vulnerable greening of terrestrial vegetation in a warming world, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-15745, https://doi.org/10.5194/egusphere-egu26-15745, 2026.
Understanding the potential future evolution of the Greenland Ice Sheet (GrIS) is of critical importance for anticipating the consequences of global climate change. GrIS melt contributes between 0.5-0.8 mm per year to total global sea level rise and the total ice volume has the potential to raise sea levels by 7 meters. Furthermore, freshwater delivery to the North Atlantic has important implications for the stability of the density-driven Atlantic Meridional Overturning Circulation. A critical challenge in anticipating GrIS sea level rise contribution and North Atlantic freshwater delivery arises from the so-called inertia of the ice sheet, wherein present-day and near-term future warming may trigger ice loss that unfolds on century to millennial timescales. The GrIS is considered one of the Earth system’s tipping elements, components of the Earth system wherein crossing a critical global warming threshold leads to large-scale and nonlinear change in system state. Though it is subject to a number of self-amplifying (and -dampening) feedbacks, the susceptibility of the GrIS to tipping is not well-constrained, with particular uncertainties arising over long (1-10 kyr) timescales, given different global warming rates, and under potential scenarios of “stabilised” or landing climate (i.e., the Paris-agreed 1.5 C).
The Tipping Points Modelling Intercomparison Project (TIPMIP) seeks to investigate the likelihood, impacts, and risk of crossing Earth system tipping points in so-called global tipping elements—components which, if tipped, have widespread consequences for the whole Earth system. Here we present initial explorations of the response of the GrIS to the transient, idealised warming scenarios following the TIPMIP-ESM and TIPMIP-ICESHEET domain protocols, with standalone, offline experiments undertaken using the Parallel Ice Sheet Model (PISM). These idealised warming scenarios have been designed to explore both the warming-induced triggers of tipping dynamics as well as their feedbacks over century to millennia timescales.
How to cite: Dennis, D. P., Albrecht, T., Ehrenfeucht, S., Klose, A. K., Reontgen, L., and Winkelmann, R.: Investigating Greenland Ice Sheet tipping in the Tipping Points Modeling Intercomparison Project (TIPMIP), EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20397, https://doi.org/10.5194/egusphere-egu26-20397, 2026.
The Atlantic Meridional Overturning Circulation (AMOC) and polar ice sheets are coupled tipping elements of the Earth system, allowing for potential cascading tipping events in which tipping is facilitated by their mutual interactions. However, while an AMOC destabilization driven by Greenland Ice Sheet (GIS) meltwater release is well-documented, the consequences of a West Antarctic Ice Sheet (WAIS) tipping on the AMOC remain unclear. In the Earth system Model of Intermediate Complexity CLIMBER-X, we perform experiments where meltwater fluxes representing plausible tipping trajectories of the GIS and WAIS are applied. We find that WAIS meltwater input can increase or decrease the AMOC resilience to GIS meltwater. In particular, for the first time in a comprehensive model, we show that WAIS meltwater can prevent an AMOC collapse. Moreover, we find this stabilzation to occur for ice sheet tipping trajectories that are relevant under high future greenhouse gas emission scenarios.
How to cite: Sinet, S., von der Heydt, A. S., and Dijkstra, H. A.: Meltwater from West Antarctic Ice Sheet Tipping Impacts AMOC Resilience, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-6440, https://doi.org/10.5194/egusphere-egu26-6440, 2026.
Satellite data below 100 m resolution can be of great benefit for prevention of geohazards. To utilise spatial and temporal data efficiently, it is necessary to develop data science techniques that are sensitive, computationally light, and capable of revealing signatures of critical events in bulky multivariate data. This emphasis on computationally light yet physically grounded detection aligns with recent climate emulation work that motivates efficient data-driven pipelines for extracting dynamical signatures from large observational datasets [1]. We apply tipping point analysis e.g., Early Warning Indicators (EWS), adapted to multivariate data flows, to demonstrate how this methodology can help complement and augment field work in the peatlands, thus optimising resources.
EWS are based on structural changes in trajectories of dynamical systems, which are described by autocorrelations and variability of the system potential [2-4]. Conventionally, peatlands are studied using expensive and slow ground surveys, but we show that equivalent information can be derived from the satellite Interferometric Synthetic Aperture Radar (InSAR) 6-12 day surface motion data using tipping point analysis. This includes processing the order of hundreds of thousands of time series potentially over 100’s of km, in combination with GIS data provided by stakeholders.
We demonstrate a case study using InSAR surface motion data over ~400km2area in Scotland with areas of critical changes in the soil surface. In a blind test, the area of a large fire (60km2) in Scottish peatlands was identified and its detection coincided with the area of actual fire damage in comparison with ground observations and existing fire detection tools based on MODIS data. This approach is promising and may be developed further to better understand peatland behaviour before and after such extreme events.
References
[1] Sudakow, I., Pokojovy, M., & Lyakhov, D., Statistical mechanics in climate emulation: Challenges and perspectives, Environmental Data Science 1, e16 (2022).
[2] Livina et al., Potential analysis reveals changing number of climate states during the last 60 kyr, Climate of the Past 6, 77-82 (2010)
[3] Livina et al., Forecasting the underlying potential governing the time series of a dynamical system, Physica A, 392 (18), 3891-3902 (2013)
[4] Prettyman et al., A novel scaling indicator of early warning signals helps anticipate tropical cyclones, Europhysics Letters 121, 10002 (2018).
How to cite: Sudakow, I., Andersen, R., Large, D., Bradley, A., and Livina, V.: Tipping point analysis of the Scottish peatlands, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8560, https://doi.org/10.5194/egusphere-egu26-8560, 2026.
As anthropogenic forcing increases, there is a rising concern about crossing tipping points. A key information is the critical thresholds at which the so-called tipping elements could undergo abrupt changes, and the associated early-warning signals. This is true, in particular, for boreal forests and the possible greening of the Sahara. In that perspective, testing climate models against paleoclimate allows for exploring the climate response over multi-millennial time series, while offering the possibility to compare it with past vegetation reconstructions.
We consider a transient simulation from the mid-Holocene (6,000 years before present) to 2100 obtained using the IPSL general circulation model including dynamical natural-only vegetation [1]. The simulation is forced with changes in orbital parameters and trace gases, transitioning from the paleo- to the historical period, and then to the scenario SSP4.5. We focus on two terrestrial ecosystems, both identified as tipping elements: the Sahara/Sahel and boreal forests in northern Europe.
We analyze the evolution of vegetation patterns and extents in the two regions along the Holocene with the objective to determine if their response to forcing conditions can be assimilated to the crossing of a tipping point. Thus, we isolate rapid shifts and investigate whether they correspond to tipping points or to centennial variability. We link them to changes in regional vegetation drivers, including vegetation feedbacks, or to AMOC variability. Then, we explore possible analogies between changes during the Holocene and in projections, allowing for the identification of sensitive regions, which may help to detect regional thresholds for tipping points.
References:
[1] Braconnot, P., Viovy, N., and Marti, O. (2025). Dynamic vegetation highlights first-order climate feedbacks and their dependence on climate mean state. Earth System Dynamics, 16, 2113–2136.
How to cite: Ragon, C., Braconnot, P., and Marti, O.: Sahara and boreal forests natural vegetation: from mid-Holocene to future, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14181, https://doi.org/10.5194/egusphere-egu26-14181, 2026.
Detecting tipping points in the Earth system is a significant challenge, particularly given the high dimensionality, observational sparsity and noisiness in real climatological data. Even in the most classical Bifurcation-induced tipping(B-tipping) with quasi-static forcing, conventional Early Warning Signals based on Critical Slowing Down often struggle with these complexities and can yield false positives in non-tipping scenarios. To address these limitations, we propose a tipping analysis indicator: High-dimensional Attractor’s Structural Complexity (HASC).
From reconstructed high-dimensional states from partial observations, we extract geometrical structures of trajectories using manifold learning method based graph approximation (Uniform Manifold Approximation and Projection) but without dimension reduction. We quantify the time-evolution of the system's structural complexity using the spectral property of the graph Laplacian, Von Neumann Entropy.
We show that HASC serves as a warning indicator of structural degeneracy of trajectory on 3box AMOC tipping model in B-tipping setting, and analyze further applications against N-and R-tipping. We also demonstrate its application to a realistic CESM AMOC collapse simulation (+10,000 dimensions). This approach offers a training-free, multivariate, and geometry-aware tool for monitoring regime shifts in complex systems.
How to cite: Hirose, T. and Sawada, Y.: Data-driven sequential analysis of tipping in high-dimensional complex systems, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-16548, https://doi.org/10.5194/egusphere-egu26-16548, 2026.
The resilience, or stability, of major Earth system components is increasingly threatened by anthropogenic pressures, demanding reliable early warning signals for abrupt and irreversible regime shifts. Widely used data-driven resilience indicators based on variance and autocorrelation detect 'critical slowing down', a signature of decreasing stability. However, the interpretation of these indicators is hampered by poorly understood interdependencies and their susceptibility to common data issues such as missing values and outliers. Here, we establish a rigorous mathematical analysis of the statistical dependency between variance- and autocorrelation-based resilience indicators, revealing that their agreement is fundamentally driven by the time series' initial data point. Using synthetic and empirical data, we demonstrate that missing values substantially weaken indicator agreement, while outliers introduce systematic biases that lead to overestimation of resilience based on temporal autocorrelation. Our results provide a necessary and rigorous foundation for preprocessing strategies and accuracy assessments across the growing number of disciplines that use real-world data to infer changes in system resilience.
How to cite: Liu, T., Morr, A., Bathiany, S., Blaschke, L. L., Qian, Z., Diao, C., Smith, T., and Boers, N.: The influence of data gaps and outliers on resilience indicators, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-3945, https://doi.org/10.5194/egusphere-egu26-3945, 2026.
The Amazon rainforest is the most species-rich region on Earth, being home to approximately ten percent of all species worldwide. However, human influences such as global warming, deforestation, and land-use change are placing unprecedented pressure on the rainforest, endangering this unique ecosystem. Recent findings suggest that the Amazon rainforest could cross a tipping point within this century when deforestation and climate change are considered together. In this study, we project the impact of such a potential tipping point on the biodiversity of the Amazon basin and compare it with a scenario without tipping. To do so, we use climate data generated by a dynamical tipping point model which simulates the Amazon forest system using a moisture-recycling network under different climate and deforestation scenarios. To assess the impact on species we compare changes in climatically suitable areas for nearly 2,000 species in the Amazon basin using an ensemble of species distribution models. Our results show that the combined effect of deforestation and climate change leads to a substantially stronger decline in climatically suitable areas than climate change alone. Including deforestation results in markedly intensified biodiversity losses already early in the century (2030-2044). Notably, the largest differences in species richness loss between scenarios do not occur in deforested area but several hundred kilometres away. These teleconnected losses are driven by deforestation-induced disruptions of atmospheric moisture transport, causing precipitation declines in distant regions and pushing species beyond their climatic niches. Overall, our results indicate that limiting global warming together with halting deforestation is critical to preventing severe and widespread biodiversity losses in the Amazon within the coming decades.
How to cite: Giesen, C., Kotz, M., Wunderling, N., Zurell, D., and Wenz, L.: Amazon tipping advances and amplifies biodiversity loss through teleconnected precipitation declines, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12673, https://doi.org/10.5194/egusphere-egu26-12673, 2026.
Posters virtual: Wed, 6 May, 14:00–18:00 | vPoster spot 4
EGU26-7797 | Posters virtual | VPS32
No critical slowing down in the Atlantic Overturning Circulation in historical CMIP6 simulationsWed, 06 May, 14:00–14:03 (CEST) vPoster spot 4
The Atlantic Meridional Overturning Circulation (AMOC) is a key component of the Earth’s climate system, and has been suggested to have multiple stable states. Critical slowing down (CSD) can detect stability changes in Earth system components, and has been found in sea-surface temperature (SST) based fingerprints of the AMOC. Here, we look for CSD in historical simulations from 27 models from the sixth Climate Model Intercomparison Project (CMIP6). We calculate three different CSD indicators for the AMOC streamfunction strengths at 26.5°N and 35°N, as well as for a previously suggested SST-based AMOC index (ASSTI) based on averaging SSTs in the subpolar gyre region. No model shows CSD in the ASSTI, which is in marked disagreement with the real-world. This lack of CSD is reflected in the AMOC streamfunctions in most models, although individual ensemble members in some models do show signs of CSD even under a conservative significance calculation. We thus conclude that: 1) The historical AMOC in CMIP6 models is not losing stability, 2) studies of AMOC stability must consider an ensemble of realisations, 3) no other physical process in the 1850-2014 period causes signs of CSD in North-Atlantic SSTs, and thus the CSD in the observed ASSTI is likely a sign of a change in the AMOC. This final result suggests that observed changes in the ASSTI could indicate a loss of stability in the real-world AMOC.
How to cite: Ben Yami, M., Blaschke, L., Bathiany, S., and Boers, N.: No critical slowing down in the Atlantic Overturning Circulation in historical CMIP6 simulations, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-7797, https://doi.org/10.5194/egusphere-egu26-7797, 2026.
EGU26-21355 | ECS | Posters virtual | VPS32
Sweden’s Food System Vulnerability to AMOC Collapse through Climate, Agricultural, and Social Work PerspectivesWed, 06 May, 14:03–14:06 (CEST) vPoster spot 4