Peatlands are important global carbon sinks but face increasing fire vulnerability under projected warmer and CO2-rich climate conditions. Past high-CO2 periods in Earth’s history serve as an important prospect to improve our understanding of fire regime changes and linked ecosystem responses. Here, we report wildfires in Late Cretaceous temperate rainforests and wetlands that existed only ~ 900 km away from the South Pole. Multiple proxy evidence from macro- and microcharcoal, polycyclic aromatic hydrocarbons, maceral analysis, inertinite reflectance and charcoal morphology document frequent, low-temperature surface fires. The fire regime changes closely coincide with a hydroseral succession (i.e., terrestrialisation) from swamp forests to the earliest known coal-forming Sphagnum-peat bog from southern high latitudes. Despite fundamental differences in atmospheric and floristic composition, fire and vegetation in 90-million-year-old Cretaceous forests and peatlands appear to have interacted in ways remarkably similar to modern ecosystems, with recurrent burning maintaining open vegetation structure and promoting peatland development. Our deep-time study establishes fire as an intrinsic driver of high-latitude wetland ecosystems and provides critical constraints on fire regime responses and ecosystem resilience under warm, elevated-CO₂ greenhouse conditions. Pollen, charcoal, and biomarkers from the 90-million-year-old peatland sediments in Antarctica reveal that fire dynamics in the earliest peatlands of the warm, CO₂-rich Cretaceous world closely resembled those of modern ecosystems.
Oil refineries must transition to become climate-neutral, but their integrated production of fuels and chemicals makes this challenging. Carbon pricing systems can encourage cleaner production, but currently ignore carbon stored in plastic products. A promising approach is to credit this temporarily stored carbon, which would incentivize renewable materials. However, it remains unclear how such policies affect overall refinery operations. Here we show that crediting carbon stored in plastic products significantly reduces emissions, fossil fuel consumption, and production costs compared to current regulations. Using an optimisation model, minimising cost of integrated refinery production in Europe by 2050, we demonstrate that including this carbon credit leads to a more efficient system. Our findings provide guidance for policymakers designing carbon pricing mechanisms that can effectively accelerate the transition to sustainable refinery operations. Crediting carbon temporarily stored in plastics fundamentally changes refinery economics by 2050; the flow-based pricing approach achieves climate neutrality at lower cost and with less fossil oil use than current carbon pricing, according to an optimisation model.
Cascading slope failures are increasing as glaciers retreat and slopes adjust to a warming climate, yet reconstructing their full dynamics from precursor to deposition remains challenging. Here we show that the 28 May 2025 Blatten rock-ice avalanche, which mobilized about 9.3 million cubic metres, can be reconstructed from seismic, geomorphological, and geotechnical evidence and numerical modeling. The observations reveal the timing, source, volume, and friction of over two weeks of accelerating rockfalls and minor glacier collapses before the main failure, followed by four stages of avalanche motion. Only with drastically reduced friction can both depth-averaged and three-dimensional models reproduce the center-of-mass force history and the deposit distribution. These friction values reflect effective frictional weakening. Geotechnical observations suggest that transient excess pore pressures possibly contributed to this weakening, but are insufficient to rule out additional mechanisms such as ice frictional reduction, segregation, and fragmentation. This work provides a reproducible and robust framework for analyzing catastrophic avalanches under a changing climate. Seismic, geomorphological and geotechnical observations combined with numerical modelling suggest the failure happened in four kinematic stages preceded by over two weeks of escalating precursory activity and with high mobility driven by extreme frictional weakening.
The stability of vegetation regimes—the temporal persistence of plant assemblages—has fluctuated markedly since the last deglaciation. While climate is the primary driver of long-term vegetation dynamics, the effects of human land use change remain less clear. Here, we analyze global vegetation regime stability over the past 15,000 years using fossil pollen records and assess the impacts of land use type and intensification. Vegetation stability declined during the climatically variable deglaciation, increased in the early to middle Holocene when climate variability dropped, but declined sharply over the past 2000 years despite no increase in climate variability, with continent-specific differences in timing. Intensive agriculture and urbanism had the strongest negative effects on vegetation stability, followed by extensive agriculture and pastoralism, whereas foraging, hunting, gathering, and fishing had minimal impact. These findings suggest that recent declines in vegetation stability are linked to intensified human land use, highlighting the need to conserve natural land less intensively managed ecosystems. Intensive agriculture and urbanism destabilize vegetation far more strongly than pastoralism or foraging, revealing contrasting long-term impacts of human land use, according to an analysis that reconstructed global vegetation regime stability over the past 15,000 years using fossil pollen records.
Giant sequoias (Sequoiadendron giganteum) are adapted to a narrow ecological niche where they endure long, dry summers, during which their water use may exceed ~2 m3/day. To meet this demand, sequoias rely on subsurface water stores, the volume and distribution of which remains uncharacterized. We use elastic full-waveform inversion and geostatistical rock physics modeling to map and quantify subsurface water content in the Mariposa Grove of Giant Sequoias in Yosemite National Park. Our results show zones of high water content within the upper 2 m of the subsurface near drainages, whereas on hillslopes and ridges, areas of high water content tend to occur deeper or in more isolated zones. Additional analysis of the water content models shows that to avoid water stress throughout the summer, hillslope and ridgetop sequoias probably access substantial volumes of deep ( > 2 m) rock moisture from weathered bedrock, implying that the extensive rooting systems of giant sequoias probe deeper into the subsurface than previously thought. Giant sequoias on hillslopes and ridgetops can access rock moisture stored in weathered bedrock over 2 m deep, helping them avoid summer water stress, suggests a study using hydrogeophysics to quantify subsurface water content in the Yosemite National Park, USA.
While the societal implications of climate extremes have received considerable attention, the specific associations with crime dynamics remain unclear. Leveraging a unique nationwide crime panel dataset (over 8.6 million) parsed via large-scale textual analysis, we capture detailed offender- and case-level attributes and examine the correlation between climate extremes and crime dynamics across China’s prefecture-level cities. We reveal that temperature anomalies, droughts, and extreme heat events are consistently associated with increases in reported crimes. Property crimes showing stronger associations than violent crimes, suggesting these associations may reflect disruptions in economic and social resource structures. These associations are stronger among rural migrants, farmers, and young males in low-income areas, providing differentiated sensitivities across groups. Potential policy responses may include targeted measures that strengthen mental health support, economic inclusion, ecological sustainability, and local governance. Our findings highlight the complex interplay between climate extremes and crime dynamics, offering policy guidance across climate-vulnerable developing regions. Temperature anomalies, droughts, and extreme heat are linked to higher reported crime, particularly property crime, according to a study using panel regression analysis of more than 8.6 million court records from Chinese cities.
Abstract How El Niño–Southern Oscillation variability and diversity have changed over the Common Era, and how corals have responded to ENSO-induced temperature anomalies, remains unclear. Here, Porites microatolls from Arno Atoll, Marshall Islands, in the western Pacific were analyzed across three periods: 484–560 ± 6 CE (Late Antique Little Ice Age; LALIA), 1273–1332 ± 4 CE (Early Little Ice Age; early LIA), and 1954–2017 CE (modern). The early LIA record indicates significantly colder conditions and reduced Sr/Ca-derived temperature seasonality compared to today, consistent with contemporaneous Central Pacific δ¹⁸O records. Negative Sr/Ca-temperature anomalies in the LALIA record likely relate to the 536 and 540 CE volcanic eruptions. Comparison with Central Pacific records suggests that Central Pacific El Niño activity increased between the first and second millennium of the Common Era, while Eastern Pacific El Niño activity remained largely stable, or may even have been stronger than today. Thermal anomalies impacted only modern coral calcification, indicating reduced tolerance when close to the population’s thermal threshold, regardless of potential adaptation to local thermal regimes.
Fire danger can escalate rapidly when vegetation dries over days to weeks. Flash droughts, characterised by rapid root-zone soil moisture depletion under high evaporative demand, may accelerate wildfire spread. Here, we quantify relationships between flash droughts and wildfire spread dynamics across land-cover types in the conterminous United States, using a model-derived root-zone soil-moisture flash-drought metric and contrast these with conventional drought conditions defined by persistently low soil moisture without rapid onset. We find that fires following flash droughts exhibit faster early spread than fires not associated with flash droughts, with model estimates indicating about 1.3–1.8-fold faster early burning across major flammable land covers. Grasslands and croplands show the strongest acceleration, followed by open shrublands, woody savannas, savannas and deciduous broadleaf forests, consistent with faster drying and greater fuel continuity in fine-fuel ecosystems. In contrast, conventional drought conditions produce weaker and less consistent changes in fire spread. Our results demonstrate that rapid soil-moisture drawdown—rather than background dryness—appears to provide a more direct pathway linking drought to accelerated wildfire spread. We recommend that monitoring declines in root-zone soil moisture in the days to weeks preceding ignition could improve early warning of fire risk in ecosystems prone to rapid drying. Wildfires exhibit up to 1.8-fold faster early spread following flash droughts than under non-flash-drought conditions, suggesting a stronger role for rapid soil-moisture depletion than background dryness, according to a study using a root-zone soil moisture-based metric across US land-cover types.
Edaphic antibiotic resistance genes (ARGs) have garnered worldwide concern, yet mechanisms by which specific microbial taxa drive ARG variation under anthropogenic stress remain unclear. Here we show how opportunistic and sensitive taxa contribute to ARG propagation in soil microcosms exposed to conventional (polyethylene, PE) or biodegradable (polybutylene adipate terephthalate, PBAT) microplastics and four-generation tetracyclines. Compared with PBAT-only, the total abundance of ARGs increases progressively in soils co-exposed to PBAT and tetracyclines, ranging from 1.17-fold for the first-generation tetracycline to 2.87-fold for the fourth-generation tetracycline. Tetracycline and multidrug ARGs are markedly enriched under PBAT combined with high-generation tetracyclines, particularly fourth-generation omadacycline. This ARG proliferation coincides with generational enrichment of opportunistic taxa, whereas sensitive taxa exhibited weak or negative correlations. Molecular docking simulations reveal that enhanced resistance potential of opportunistic taxa stems from stronger binding affinities between resistance proteins and high-generation tetracyclines. These findings highlight the pivotal role of opportunistic microbes in ARG dissemination under biodegradable microplastics and newer antibiotic generations, advancing our mechanistic understanding of resistance proliferation in soil ecosystems. Strong binding of resistance proteins to high-generation tetracyclines exacerbates antibiotic resistance gene spread in soil exposed to biodegradable polybutylene adipate terephthalate and tetracyclines, based on soil microcosm experiments.
The conjugative transfer (CT) of antibiotic resistance genes (ARGs) during dynamic conversion of mackinawite (FeS) have received little attention. We disentangled the individual contribution of minerals-mediated and •OH-mediated processes towards CT. Low concentrations of FeS promoted CT frequency with a volcano-shaped change trend, which was jointly driven by minerals-mediated and •OH-mediated processes, with their relative contributions shifting from equalling to prevailing by the former as FeS concentration increased. However, high concentrations of FeS manifested significant inhibition to CT owing to growth coercion caused by FeS itself. The key of promoting CT lied in a cascade of up-regulated bacterial responses against oxidatation, involving membrane permeability, ATP synthesis, extracellular polymeric substances secretion, and quorum sensing. The stress triggered by excessive FeS impaired membrane integrity and caused overexpression of bacterial responses, thereby suppressing CT. These findings provide valuable insights into effects of active minerals on ARGs dissemination and developing pollution control technologies. Low concentrations of mackinawite (5-50 mg/L) promoted conjugative transfer through mineral- and hydroxyl radical-mediated processes, while high concentrations (75−100 mg/L) inhibited transfer due to growth coercion and refractory cell damage, based on model conjugation experiments.
Earthquakes disrupt terrestrial ecosystem carbon storage by triggering landslides and altering land cover, yet a global quantification of these carbon losses has been lacking. Here we quantify the global earthquake-induced carbon losses by combining vegetation models, land cover, carbon density, and earthquake records. Focusing on strong earthquakes worldwide between 2006 and 2018, we find that these events cause a cumulative loss of 38.86 ± 6.23 teragrams of carbon by 2022. Forest ecosystems require at least six years to begin recovery. These carbon losses are unevenly distributed across countries, with 84.6% of affected countries being emerging and developing economies. In Nepal and Ecuador, cumulative earthquake-induced carbon losses exceed half of their cumulative land-use change emissions during the post-earthquake period. Our study establishes a global benchmark for carbon losses from geophysical disturbances, offering a crucial reference for national carbon accounting and disaster resilience strategies. Strong earthquakes worldwide between 2006 and 2018 caused a cumulative loss of 38.86 ± 6.23 Tg of C by 2022, with emerging and developing economies being most affected, according to estimates derived from land cover, carbon density and global vegetation models
Framboidal pyrite serves as a critical archive of mineralization and paleoenvironmental history, yet its formation remains contested between biological templating and abiotic self-organization. Integrating geometric and crystallographic analyses of framboidal pyrite from the Gakkel Ridge, we reveal a hierarchical architecture defined by fractal geometry (fractal dimensions of 1.7–1.9) and systematic core-to-rim gradients in crystallinity and lattice strain. These findings underpin a unified three-stage growth model: initiation by synchronous nucleation, advancement via diffusion-limited aggregation into a fractal skeleton, and finalization through surface-energy minimization driving spheroidization. This mechanism provides a universal abiotic explanation for framboidal formation across diverse mineral systems. While biological activity may provide the initial supersaturation pulse, the characteristic texture remains a robust outcome of physical forces. The resulting architecture optimizes biosignature preservation, wherein defect-rich cores adsorb organic matter encapsulated by high-crystallinity rims. These findings resolve the long-standing biotic-abiotic dichotomy, establishing framboids as physically inevitable products and robust geochemical proxies for decoding mineralogical biosignatures in both terrestrial and extraterrestrial contexts. Microscopic and spectroscopic observations of framboidal pyrite from the Gakkel Ridge support a 3-stage growth model, which is compatible with, but does not require, biological activity and can explain the preservation of biosignatures in framboidal structures
Water vapor isotopes carry the integrated history of evaporation, condensation, mixing, and transport. Although previous studies have shown potential to improve forecasts under controlled conditions, real-world applications have been limited by systematic biases in both models and satellite retrievals arising from sparse measurements in the free troposphere. Here we assimilate mid‑tropospheric δD retrievals (peak sensitivity ~4.2 km) from the Infrared Atmospheric Sounding Interferometer into the Isotope‑incorporated Global Spectral Model and evaluate the added value beyond co‑assimilated temperature and specific humidity with identical spatial and temporal coverage. Assimilating δD improves 0–120 h forecasts of wind, temperature, specific humidity, and geopotential height, with the largest gains in the midlatitudes; heavy‑precipitation skill also increases for thresholds >3 mm per 6 h. Demonstrated in a coarse‑resolution configuration with limited observations, the results indicate that isotopic information strengthens transport tracking and hydrological constraints, motivating evaluation in operational high‑resolution forecasting systems. Assimilation of mid-tropospheric vapor isotopes significantly improves short-term weather forecasts and mean-to-extreme precipitation, especially at mid-latitudes, according to an analysis of 4.5 years of data from the Infrared Atmospheric Sounding Interferometer.
Improving the subseasonal-to-seasonal (S2S) prediction of heatwaves is a critical scientific challenge due to their severe societal impacts and the limited predictability at these lead times. Here we show two dominant and physically distinct sources of S2S predictability for the weekly hot days across North America. The leading predictable mode, with its strongest loading over the southern U.S., reflects an ocean-forced mode associated with a Pacific Decadal Oscillation-like sea surface temperature pattern. In contrast, the second most predictable mode, with centers of action over northwestern and southeastern North America, is dominated by atmospheric internal variability associated with the summer Arctic Oscillation and summer North Atlantic Oscillation. Skillful week-3 forecasts of hot days are further characterized by quasi-stationary wave patterns coupled with regional soil moisture deficit, revealing dynamical and land-surface ‘windows of opportunity’ for extended S2S predictability. These findings identify the physical mechanisms governing North American heat extreme predictability and provide a scientific basis for improving extended-range early warnings of heat extremes. Subseasonal-to-Seasonal Predictability of Summer Heat Extremes over North America arises from distinct oceanic and atmospheric modes coupled with soil moisture deficits, creating windows of opportunity for extended forecasting, as revealed by ensemble hindcast experiments.
Maintaining healthy thermal environments has long focused on people’s subjective satisfaction with temperature conditions – commonly termed thermal comfort. This approach, however, underrepresents how heat, thermal environments, adaptive response, and health outcomes interact. As climate change drives more frequent and intense heatwaves, there is an urgent need to move beyond thermal comfort and its close variations such as thermal stress, toward a broader, more integrated concept of thermal health. Here we show the Integrative Thermal Health (ITH) framework comprising eight key hallmarks across physiological and psychological health through scientific literature examination combining conventional bibliometric techniques with machine learning. These hallmarks sit within four key stages of the health continuum, from mortality to wellbeing and span the full 24-hour cycle of daily life. The framework provides a broader foundation for understanding how thermal environments—mediated by human activities—shape health, clarifies causal pathways, supports wider health indicators, and guides research and policy amid growing heat-related health risks. An Integrative Thermal Health framework with eight hallmarks across four health stages shows how thermal environments shape human health beyond thermal comfort, according to a bibliometric analysis and machine learning.
Abstract Elongated dykes in tectonic rifts feed extensive basaltic fissure eruptions, yet surface eruption sites are highly discontinuous and unpredictable. Here we show that along-strike topographic variations, and the associated stress gradients, correlate with this localisation. We investigate eruptive cones at the two most productive historical basaltic eruptions in Iceland, Laki (1783–1784) and Eldgjá (~939–940). Morphological analyses of > 300 cones reveal volume variations (10 0 –10 7 m 3 ) and segmentation along ~ 100 km of fissures. Comparing cone distribution and volumes along the rift zone with pre-eruptive stress models, we demonstrate that high topographic stress gradients inhibit cone growth. We identify a critical stress gradient threshold predicting where larger cones are inhibited, while horizontal stress gradient peaks predict the location of fissure gaps. Our findings establish an empirical and physical link between topographic stress and volcanic morphologies, complementing previous models with a quantitative framework that aims at improving the forecasting of along-rift vent locations and development.
Abstract Emperor penguins are threatened by climate change and their population is predicted to decline rapidly this century due to changes in Antarctic sea ice. Remote sensing is a key tool for monitoring emperor penguins but analysis has been restricted to spring/summer due to a lack of optical imagery during the polar night. Here we show Sentinel-1 synthetic aperture radar (SAR) can be used to observe the movement of penguin colonies throughout winter. We combine winter SAR imagery with medium-resolution summer optical imagery to track the year-round movement of colonies at Atka Bay, Coulman Island, and Cape Washington, 2017–2024. Although emperor penguins are less mobile than during summer, all three colonies exhibit movement in winter, but which varies between years and colonies. At times, different groups within the colonies synchronously move away from the fast ice edge despite its apparent stability. We demonstrate that SAR is an important tool for understanding emperor penguin behaviour throughout their breeding cycle, and particularly their response to environmental forcing and change during winter.
Large marine animals face severe threats from industrial fisheries globally, with bycatch recognized as a primary driver of population declines. This challenge is particularly pronounced in China’s seas and adjacent waters—a region of extraordinary biodiversity yet intense fishing pressure. Here we show a cumulative impact analysis framework, integrating information on species habitat suitability, and fishing effort stratified by gear type, to assess bycatch risks for threatened and protected species. Our results indicate that peak values of bycatch risk occur in the 23°N–25°N coastal waters of the Chinese mainland and off western Taiwan Island. On average, potential species bycatch occurred across 60% of their estimated ranges, and about 10% of the species may be bycaught in over 90% of their estimated ranges. Marine protected areas effectively mitigate risk, with internal bycatch risk one-third of that in surrounding areas. This study supports targeted conservation and fisheries management by prioritizing high-need species and sites. Bycatch risk in marine protected areas in the seas around China is one-third that of the surrounding area, showing they can effectively mitigate fishing pressures even with high bycatch risk along the continental shelf, according to a cumulative analysis framework integrating 10 years of data.
As global warming worsens urban heat, accurate urban heat modelling is key to cooling strategies, which traditionally overemphasises coarse land cover while overlooking detailed three-dimensional (3D) materials like building façades and roads. This study operationalised a high-throughput pipeline to quantify 3D material envelope information across five diverse United States cities: Boston, Chicago, Los Angeles, New York, and Seattle. We provided the first city-wide, street-level 3D material albedo dataset and revealed a consistent cross-city pattern: façade albedo shows a robust cooling association across diverse urban morphologies, supported by SHapley Additive exPlanations (SHAP) interpretation and counterfactual scenarios, whereas road albedo interacts nonlinearly with canyon geometry. Our findings suggest a previously unquantified material-geometry regime and indicate effective mitigation measures to prioritise the hottest low-density suburban areas where low-income households dominate; high albedo renewal can yield a cooling potential of up to 1.14 °C in Boston. This study highlights the critical role of 3D material albedo in urban heat, offering actionable insights for equitable climate adaptation. Façade albedo consistently reduces urban heat across five United States cities, with cooling of up to 1.14°C in Boston, while road albedo effects vary with urban geometry, according to city-wide street-level three-dimensional material albedo mapping and scenario analyses.
Global energy demand continues to rise, increasing the need for reliable renewable energy resources. Salinity gradient energy, generated by freshwater–seawater mixing, is an emerging renewable energy resource that remains insufficiently understood in estuarine and coastal regions. As this resource depends on climate-sensitive river discharge, water temperature, and salinity contrasts, its future evolution remains uncertain. Here, using state-of-the-art climate simulations, we quantify the projected changes in the global theoretical salinity gradient energy resource under global warming. The simulations consistently show an overall increase in this theoretical resource, with large regional heterogeneity. Enhanced river discharge associated with increased precipitation minus evapotranspiration is the dominant contributor, while higher mixed-water temperature further increases the resource and altered ocean–river salinity contrasts partly offset these effects. These results indicate that global warming may expand the global theoretical resource base of salinity gradient energy, providing a basis for assessing its potential contribution to future renewable energy portfolios. Global theoretical salinity gradient energy is projected to increase overall under greenhouse warming, with rising river discharge as the primary driver and warmer mixed-water temperatures providing additional enhancement, according to state-of-the-art climate simulations.
