With the world already experiencing dangerous climate change, decarbonization and climate adaptation are urgent challenges. The potential for even more rapid and irreversible changes augments this urgency. To emphasize the need for rapid climate action, some researchers have labeled such potential changes as "tipping points." But the increasingly broad application of this term across biogeophysical and social phenomena muddles its meaning, and its use can conceal the multidimensional complexity of processes that can drive rapid change. The term is better viewed as a metaphor than as describing a technically defined class of processes. This metaphor is not risk-free, however, and should be employed with caution. For example, many people confuse the 1.5 degrees Celsius milestone for global warming established by the Paris Agreement as a physical tipping point, which it is not. That confusion poses risk to the credibility of the scientific community, and it can harm climate action by encouraging both risky "emergency" techno-fixes and despondent doomism.

Understanding extreme storm surge events that threaten low-lying coastal communities is key to effective flood mitigation/adaptation measures. However, observational estimates are sparse and highly uncertain along most coastal regions with a lack of observational evidence about long-term underlying trends and their contribution to overall extreme sea-level changes. Here, using a spatiotemporal Bayesian hierarchical framework, we analyse US tide gauge record for 1950-2020 and find that observational estimates have underestimated likelihoods of storm surge extremes at 85% of tide gauge sites nationwide. Additionally, and contrary to prevailing beliefs, storm surge extremes show spatially coherent trends along many widespread coastal areas, providing evidence of changing coastal storm intensity in the historical monitoring period. Several hotspots exist with regionally significant storm surge trends that are comparable to trends in mean sea-level rise and its key components. Our findings challenge traditional coastal design/planning practices that rely on estimates from discrete observations and assume stationarity in surge extremes.

Conventional computational models of climate adaptation frameworks inadequately consider decision-makers' capacity to learn, update, and improve decisions. Here, we investigate the potential of reinforcement learning (RL), a machine learning technique that efficaciously acquires knowledge from the environment and systematically optimizes dynamic decisions, in modeling and informing adaptive climate decision-making. We consider coastal flood risk mitigations for Manhattan, New York City, USA (NYC), illustrating the benefit of continuously incorporating observations of sea-level rise into systematic designs of adaptive strategies. We find that when designing adaptive seawalls to protect NYC, the RL-derived strategy significantly reduces the expected net cost by 6 to 36% under the moderate emissions scenario SSP2-4.5 (9 to 77% under the high emissions scenario SSP5-8.5), compared to conventional methods. When considering multiple adaptive policies, including accomodation and retreat as well as protection, the RL approach leads to a further 5% (15%) cost reduction, showing RL's flexibility in coordinatively addressing complex policy design problems. RL also outperforms conventional methods in controlling tail risk (i.e., low probability, high impact outcomes) and in avoiding losses induced by misinformation about the climate state (e.g., deep uncertainty), demonstrating the importance of systematic learning and updating in addressing extremes and uncertainties related to climate adaptation.

Climate change threatens global food systems1, but the extent to which adaptation will reduce losses remains unknown and controversial2. Even within the well-studied context of US agriculture, some analyses argue that adaptation will be widespread and climate damages small3,4, whereas others conclude that adaptation will be limited and losses severe5,6. Scenario-based analyses indicate that adaptation should have notable consequences on global agricultural productivity7-9, but there has been no systematic study of how extensively real-world producers actually adapt at the global scale. Here we empirically estimate the impact of global producer adaptations using longitudinal data on six staple crops spanning 12,658 regions, capturing two-thirds of global crop calories. We estimate that global production declines 5.5 × 1014 kcal annually per 1 °C global mean surface temperature (GMST) rise (120 kcal per person per day or 4.4% of recommended consumption per 1 °C; P < 0.001). We project that adaptation and income growth alleviate 23% of global losses in 2050 and 34% at the end of the century (6% and 12%, respectively; moderate-emissions scenario), but substantial residual losses remain for all staples except rice. In contrast to analyses of other outcomes that project the greatest damages to the global poor10,11, we find that global impacts are dominated by losses to modern-day breadbaskets with favourable climates and limited present adaptation, although losses in low-income regions losses are also substantial. These results indicate a scale of innovation, cropland expansion or further adaptation that might be necessary to ensure food security in a changing climate.

Geological records of past environmental change provide crucial insights into long-term climate variability, trends, non-stationarity, and nonlinear feedback mechanisms. However, reconstructing spatiotemporal fields from these records is statistically challenging due to their sparse, indirect, and noisy nature. Here, we present PaleoSTeHM, a scalable and modern framework for spatiotemporal hierarchical modeling of paleo-environmental data. This framework enables the implementation of flexible statistical models that rigorously quantify spatial and temporal variability from geological data while clearly distinguishing measurement and inferential uncertainty from process variability. We illustrate its application by reconstructing temporal and spatiotemporal paleo-sea-level changes across multiple locations. Using various modeling and analysis choices, PaleoSTeHM demonstrates the impact of different methods on inference results and computational efficiency. Our results highlight the critical role of model selection in addressing specific paleo-environmental questions, showcasing the PaleoSTeHM framework's potential to enhance the robustness and transparency of paleo-environmental reconstructions.

Reliable estimates of storm surge and sea level extremes with proper uncertainty quantification are key for cost-effective risk/adaptation planning. However, observational estimates are often unavailable or uncertain along most coastlines owing to data scarcity. Here, we provide a fully observational-driven probabilistic dataset (US-CoastEX) of storm surge and sea level extremes for the U.S. coast (1950-2020). Non-stationary extreme storm surge distributions are generated for gauged and ungauged sites by applying Bayesian methods to the U.S. tide gauge network, complemented with additional storm data unavailable in commonly used tide gauge data. The distributions are combined with tidal peak data to estimate return periods and levels of extreme sea levels and their uncertainty. Ou results show that traditional site-by-site estimates based on existing model data, as well as regionally-aggregated analysis of standard tide gauge data, have underestimated 100-year extreme sea levels by 50% (on average) along much of the U.S. coast, especially in regions exposed to extreme storms. The data supports coastal managers to make decisions, especially in vulnerable areas where in-situ sea-level monitoring is limited.

Coastal planning, mitigation and adaptation efforts rely on credible sea-level projections generated by physical models. However, the large uncertainties in these projections pose a challenge for policymakers. Here we provide an overview of the main sources of uncertainty in model projections of sea-level change on multi-decadal to centennial timescales and we offer perspectives on the use of observations to narrow uncertainties. We propose several directions for future research, including improvements in emerging emulation techniques, more systematic quantification of uncertainty structure within both observations and models, lengthening observational records of processes, and expanding collaborations across physical and social sciences. Advancements in these areas are urgently needed, as the next phase of the Intergovernmental Panel on Climate Change assessment cycle gets underway.

Quantifying physical mechanisms driving sea-level change-including global mean sea level (GMSL) and regional-to-local components (that is, sea-level budget)-is essential for reliable future projections and effective coastal management1,2. Although previous research has attempted to resolve China's sea-level budget from the 1950s3,4, these studies capture short timescales and lack the long-term context necessary to fully assess modern sea-level rise in southeastern China5-one of the world's most densely populated regions with immense socioeconomic importance6. Here we show that GMSL followed three distinct stages from 11,700 years before present (BP) to the modern day: (1) rapid early Holocene rise driven by the deglacial melt of land ice; (2) 4,000 years of stability from around 4200 BP to the mid-nineteenth century when regional processes dominated sea-level change; and (3) accelerating rise from the mid-nineteenth century. Our results arise from spatiotemporal hierarchical modelling of geological sea-level proxies and tide gauge data to produce site-specific sea-level budget estimates with uncertainty quantification. It is extremely likely (P ≥ 0.95) that the GMSL rise rate since 1900 (1.51 ± 0.16 mm year-1, 1σ) has exceeded any century over at least the past four millennia. Moreover, our analysis indicates that at least 94% of rapid modern urban subsidence is attributable to anthropogenic activities, with localized subsidence rates often exceeding GMSL rise. Such concurrent acceleration of global sea-level rise and rapid localized subsidence has not been observed in our Holocene geological record.

On behalf of the journal, AGU, and the scientific community, we, the editors of Earth's Future, are delighted to publish the names of the 1,061 peer reviewers who provided 1,642 reviews for our journal in 2024 (italicized names have contributed three or more reviews). Your diligent efforts to provide timely comments on our submissions have significantly improved the manuscripts and elevated the scientific rigor of future research. As a unique transdisciplinary journal, Earth's Future delves into the state of the planet and its inhabitants, sustainable and resilient societies, the science of the Anthropocene, and predictions of our shared future through research articles, reviews, and commentaries. In the face of observed and anticipated global environmental and climatic changes, the need for high‐quality scientific theories, assessments, and projections about the future of our planet has never been more pressing. To safeguard research integrity in this crucial area, we rely on our reviewers' expertise and selfless cooperation. We extend our heartfelt thanks to each of the individuals listed below for their contributions to our journal and the broader scientific discourse. Your dedication is immensely appreciated.

Sea-level science has seen many recent developments in observations and modelling of the different contributions and the total mean sea-level change. Here, we focus on sea-level projections in the recent IPCC reports, and discuss (1) the evolution in IPCC projections (“how we got here”), (2) how the projections compare to observations (“where we are”) and (3) the outlook for further improving projections (“where we’re going”). We start by discussing how the model projections of 21st century sea-level change have changed from the IPCC AR5 report (2013) to SROCC (2019) and AR6 (2021), highlighting similarities and differences in the methodologies and comparing the global mean and regional projections. This shows that there is good agreement in the median values, but also highlights some differences. In addition, we discuss how the different reports included high-end projections. We then show how the AR5 projections (from 2007 onwards) compare against the observations, and find that they are highly consistent with each other. Finally, we discuss how to further improve sea-level projections in future studies.

The Intergovernmental Panel on Climate Change (IPCC) exists to provide policy-relevant assessments of the science related to climate change. As such, the IPCC has long grappled with characterizing and communicating uncertainty in its assessments. Decision Making under Deep Uncertainty (DMDU) is a set of concepts, methods, and tools to inform decisions when there exist substantial and significant limitations on what is and can be known about policy-relevant questions. Over the last twenty-five years, the IPCC has drawn increasingly on DMDU concepts to more effectively include policy-relevant, but lower-confidence scientific information in its assessments. This paper traces the history of the IPCC’s use of DMDU and explains the intersection with key IPCC concepts such as risk, scenarios, treatment of uncertainty, storylines and high-impact, low-likelihood outcomes, and both adaptation and climate resilient development pathways. The paper suggests how the IPCC might benefit from enhanced use of DMDU in its current (7th) assessment cycle.

ABSTRACTGeologic reconstructions of overwash events can extend storm records beyond the brief instrumental record. However, the return periods of storms calculated from geologic records alone may underestimate the frequency of events given the preservation bias of geologic records. Here, we compare a geologic reconstruction of storm activity from a salt marsh in New Jersey to two neighboring instrumental records at the Sandy Hook and Battery tide gauges. Eight overwash deposits were identified within the marsh's stratigraphy by their fan‐shaped morphology and coarser mean grain size (3.6 ± 0.7 φ) compared to autochthonous sediments they were embedded in (5.6 ± 0.8 φ). We used an age–depth model based on modern chronohorizons and three radiocarbon dates to provide age constraints for the overwash deposits. Seven of the overwash deposits were attributed to historical storms, including the youngest overwash deposit from Hurricane Sandy in 2012. The four youngest overwash deposits overlap with instrumental records. In contrast, the Sandy Hook and Battery tide gauges recorded eight and 11 extreme water levels above the 10% annual expected probability (AEP) of exceedance level, respectively, between 1932/1920 and the present. The geologic record in northern New Jersey, therefore, has a 36–50% preservation potential of capturing extreme water levels above the 10% AEP level.

Anticipating and managing the impacts of sea‐level rise for nations astride active tectonic margins requires understanding of rates of sea surface elevation change in relation to coastal land elevation. Vertical land motion (VLM) can either exacerbate or reduce sea‐level changes with impacts varying significantly along a coastline. Determining rate, pattern, and variability of VLM near coasts leads to a direct improvement of location‐specific relative sea level (RSL) estimates for coastal hazard risk assessment. Here, we utilize vertical velocity field from interferometric synthetic aperture radar (InSAR) data, calibrated with campaign and continuous Global Navigation Satellite System data, to determine the VLM for the entire coastline of New Zealand. Guided by available knowledge of the seismic cycle, the VLM data infer secular, interseismic rates of land surface deformation. Using the Framework for Assessing Changes to Sea‐level (FACTS), we build probabilistic RSL projections using the same emissions scenarios employed in IPCC Assessment Report 6 and local VLM data at 8,179 sites, thereby enhancing spatial coverage that was previously limited to four tide gauges. We present ensembles of probability distributions of RSL for each scenario to 2150, and for low confidence sea‐level processes to 2300. Where land subsidence is occurring at rates >2 mm/y VLM makes a significant contribution to RSL projections for all scenarios out to 2150. Our approach can be applied to similar locations across the world and has significant implications for adaptation planning, as timing of threshold exceedance for coastal inundation can be brought forward (or delayed) by decades.

AbstractGlobal mean sea-level (GMSL) change can shed light on how the Earth system responds to warming. Glaciological evidence indicates that Earth’s ice sheets retreated inland of early industrial (1850 CE) extents during the Holocene (11.7-0 ka), yet previous work suggests that Holocene GMSL never surpassed early industrial levels. We merge sea-level data with a glacial isostatic adjustment model ensemble and reconstructions of postglacial thermosteric sea-level and mountain glacier evolution to estimate Holocene GMSL and ice volume. We show it is likely (probability P = 0.75) GMSL exceeded early industrial levels after 7.5ka, reaching 0.24 m (−3.3 to 1.0 m, 90% credible interval) above present by 3.2ka; Antarctica was likely (P = 0.78) smaller than present after 7ka; GMSL rise by 2150 will very likely (P = 0.9) be the fastest in the last 5000 years; and by 2060, GMSL will as likely than not (P = 0.5) be the highest in 115,000 years.

Estimates of extreme sea-level return periods guide flood hazard mitigation. Return period estimates calculated from tide gauge records, which are relatively short (typically less than 100 years), can fail to capture the rarest and most potentially impactful extreme events. Here, we employ a two-dimensional Poisson point process model to fuse water-level data from tide gauges with data from multi-century geologic records of extreme overwash events. Experiments with synthetic data show that including geologic data reduces the uncertainty of 1% and 0.1% average annual chance water levels by about half, relative to using tide gauge data alone. Similar uncertainty reductions occur with two case studies of geologic data (Mattapoisett Marsh, Massachusetts and Cheesequake, New Jersey) and their neighboring tide gauges (Woods Hole, Massachusetts and the Battery, New York). The analysis also reveals non-stationarity at Cheesequake and The Battery, arising from either climatic changes or changes in the fidelity of the geological record, with substantially higher 1–10% average annual chance water levels since 1900 compared to prior centuries.

Climate change is posing unprecedented challenges, necessitating the development of effective climate adaptation. Conventional computational models of climate adaptation frameworks inadequately account for our capacity to learn, update, and enhance decisions as exogenous information is collected. Here we investigate the potential of reinforcement learning (RL), a machine learning technique that exhibits efficacy in acquiring knowledge from the environment and systematically optimizing dynamic decisions, to model and inform adaptive climate decision-making. To illustrate, we derive adaptive stratigies for coastal flood protections for Manhattan, New York City, considering continuous observations of sea-level rise throughout the 21 century. We find that, when designing adaptive seawalls to protect Manhattan, the RL-derived strategy leads to a significant reduction in the expected cost, 6% to 36% under the moderate emissions scenario SSP2-4.5 (9% to 77% under the high emissions scenario SSP5-8.5), compared to previous methods. When considering multiple adaptive policies (buyout, accommodate, and dike), the RL approach leads to a further 5% (15%) reduction in cost, showcasing RL’s flexibility in addressing complex policy design problems when multiple policies interact. RL also outperforms conventional methods in controlling tail risk (i.e., low probability, high impacts) and avoiding losses induced by misinformation (e.g., biased sea-level projections), demonstrating the importance of systematic learning and updating in addressing extremes and uncertainties related to climate adaptation. The analysis also reveals that, given the large uncertainty and potential misjudgment about climate projection, “preparing for the worst” is economically more beneficial when adaptive strategies, such as those supported by the RL approach, are applied.

Global climate models (GCMs) are important tools for understanding the climate system and how it is projected to evolve under scenario-driven emissions pathways. Their output is widely used in climate impacts research for modeling the current and future effects of climate change. However, climate model output remains coarse in relation to the high-resolution climate data needed for climate impacts studies, and it also exhibits biases relative to observational data. Treatment of the distribution tails is a key challenge in existing bias-adjusted and downscaled climate datasets available at a global scale; many of these datasets used quantile mapping techniques that were known to dampen or amplify trends in the tails. In this study, we apply the Quantile Delta Mapping (QDM) method (Cannon et al., 2015) for bias adjustment. After bias adjustment, we apply a new spatial downscaling method called Quantile-Preserving Localized-Analog Downscaling (QPLAD), which is designed to preserve trends in the distribution tails. Both methods are integrated into a transparent and reproducible software pipeline, which we apply to global, daily GCM surface variable outputs (maximum and minimum temperature and total precipitation) from the Coupled Model Intercomparison Project Phase 6 (CMIP6) experiments (O'Neill et al., 2016) for the historical experiment and four future emissions scenarios ranging from aggressive mitigation to no mitigation, namely SSP1–2.6, SSP2–4.5, SSP3–7.0, and SSP5–8.5 (Riahi et al., 2017). We use the European Centre for Medium-Range Weather Forecasts (ECMWF) ERA5 (Hersbach et al., 2020) temperature and precipitation reanalysis as the reference dataset over the Intergovernmental Panel on Climate Change (IPCC) Sixth Assessment Report (AR6) reference period of 1995–2014. We produce bias-adjusted and downscaled data over the historical period (1950–2014) and the future emissions pathways (2015–2100) for 25 GCMs in total. The output dataset is the Global Downscaled Projections for Climate Impacts Research (GDPCIR), a global, daily, 0.25∘ horizontal-resolution product which is publicly available and hosted on Microsoft AI for Earth's Planetary Computer (https://planetarycomputer.microsoft.com/dataset/group/cil-gdpcir/, last access: 23 October 2023).

Tipping points have gained substantial traction in climate change discourses. Here we critique the ‘tipping point’ framing for oversimplifying the diverse dynamics of complex natural and human systems and for conveying urgency without fostering a meaningful basis for climate action. Multiple social scientific frameworks suggest that the deep uncertainty and perceived abstractness of climate tipping points render them ineffective for triggering action and setting governance goals. The framing also promotes confusion between temperature-based policy benchmarks and properties of the climate system. In both natural and human systems, we advocate for clearer, more specific language to describe the phenomena labelled as tipping points and for critical evaluation of whether, how and why different framings can support scientific understanding and climate risk management. The tipping points framing is widely used in climate discussions but receives mixed feedback. This Perspective critiques it for oversimplifying the complexities of natural and social systems and failing to drive effective action, and offers recommendations for future improvements.