Earths Future最新文献

Wildfires have increasingly affected human and natural systems across the western United States (WUS) in recent decades. Given that the majority of ignitions are human-caused and potentially preventable, improving the ability to predict fire occurrence is critical for effective wildfire prevention and risk mitigation. We used over 500,000 wildfire ignition records from 2000 to 2020 to develop machine learning models that predict daily ignition probability across the WUS and incorporate a wide range of physical, biological, social, and administrative variables. A key innovation of this work is development of novel sampling techniques for representing ignition absence. Unlike traditional purely random sampling or hyper-sampling, which does not account for temporally autocorrelated factors (such as droughts, insect outbreaks, and heatwaves) and spatially autocorrelated factors (such as proximity to human settlements, infrastructure presence, and fuel type), we introduce spatially and temporally stratified sampling of ignition absence. By drawing absence samples near the location and time of historical ignitions, we better captured the complex environmental and anthropogenic conditions associated with fire occurrence or lack thereof. Models trained without stratified sampling produced ignition probability maps that consistently overestimated fire risk during high fire danger periods, whereas models incorporating stratified fire absence samples more accurately captured the spatial and temporal variability of fire potential and achieved predictive accuracies exceeding 95%. In addition to operational utility for fire prevention and resource allocation, our approach offers insights into the drivers of wildfire ignitions and highlights the value of incorporating spatial and temporal structure in absence sampling for wildfire modeling.

Plants modify their functional traits in response to changing environmental conditions under climate change. However, it remains unclear whether tree planting alters patterns and acclimation of hydraulic traits across spatial scales. Here, we compiled a site-level data set of hydraulic traits in natural (NF) and planted forests (PF) to examine trait patterns and relationships, quantified environmental and ecological drivers on ecosystem-scale hydraulic traits of PF and NF across China, and computationally projected future trait acclimation using the space-for-time approach. We identified distinct differences in hydraulic traits between NF and PF, with PF exhibiting higher hydraulic safety but lower hydraulic efficiency than NF at the species level. NF demonstrated a trade-off between hydraulic efficiency and safety, whereas PF exhibited a contrasting positive correlation between these traits. We confirmed that both environmental and ecological factors influence ecosystem-scale hydraulic traits in NF and PF, although dominant drivers vary among specific traits. Projections under future climate scenarios suggest that, despite persistent differences in trait acclimation between NF and PF, both forest types tend to exhibit increased water-use efficiency and enhanced drought resistance in response to rising precipitation and air dryness. These findings provide a valuable benchmark for estimating potential changes in hydraulic traits under climate change, supporting improved simulations of carbon and water fluxes in response to climate and anthropogenic influences.

While the impacts of subsurface fluxes on mediating hydrologic response to droughts are often ignored, several studies indicate that baseflow can sustain rivers during droughts and decrease the vulnerability of water supplies. Therefore, given the increasing impacts of droughts on economic and environmental issues, understanding the baseflow drought (BFD) evolution and its drivers are critical. In this study, we quantify and analyze the long-term evolution of BFD characteristics across the Contiguous United States. We use long-term daily baseflow values of DeepBase data set and explainable machine learning models to identify and rank the climatic and physical drivers of BFD. Our analysis reveals notable regional disparities in BFDs, with western regions, particularly the Southwest, experiencing increased frequency and prolonged durations, while much of the eastern areas show declining trends. During the past decade, BFD frequency has been governed mainly by anomalies in the atmospheric water balance and by soil properties. Its duration has been primarily influenced by hydrogeologic attributes, and its intensity has been modulated most strongly by topographic setting. Highlighting the non-stationary and complex nature of BFD mechanisms, our results have practical implications for water resource management and drought adaptation strategies.

This study employs a novel application of Impact Chains combined with tailored metrics to conduct a cross-country comparative analysis of the same hazard event. This is a research approach currently absent from the literature, yet essential for developing transferable Disaster Risk Management (DRM) lessons. Given the increased frequency of hydrometeorological hazards under climate change, the importance of such lessons cannot be overstated. Focusing on the devastating 2021 floods in Europe, this study investigates (1) how did the interplay of impacts, vulnerabilities, and adaptation options produce divergent disaster outcomes in two contrasting European contexts, and (2) what transferable lessons for DRM can be elicited. The selected case studies focus on the areas hardest impacted by the 2021 flood events in Romania (Alba County) and the Netherlands (Limburg Province). The three-tiered analysis performed at the level of impacts, vulnerabilities, and adaptation options shows that response in the two countries was well-directed, while vulnerability mitigation remains a critical problem, differing between the two case studies. Another highlight is that development in flood-prone areas and flood protection standards below (flood) hazard return periods ranked as the most influential vulnerabilities in both case studies. This research provides a new methodology for comparative disaster analysis, offering critical insights for scientists and practitioners in the face of increasingly frequent and impactful hydrometeorological hazards.

Less than 6% of US farmlands are cover cropped, an on-farm management practice with potential to sequester carbon and provide environmental co-benefits (e.g., improved soil health and water quality). Despite their promise as a nature-based climate solution that can enhance soil carbon storage, cover crops remain underutilized, in part due to farmer perceptions that benefits are not assured and management risk is high. This scoping review synthesizes research from multiple disciplines to identify persistent knowledge gaps that limit both the effectiveness and sustained adoption of cover cropping. We focus on Midwestern US agroecosystems, where scientists generally agree that cover cropping has the potential to increase soil organic carbon and contribute to long-term carbon sequestration. Our synthesis reveals critical challenges across domains: carbon outcomes are highly variable across space and time; water and nutrient dynamics exhibit tradeoffs and context dependence; economic returns remain difficult to quantify; and adoption patterns are shaped by feedbacks between perceived risk, observed outcomes, system constraints, and social factors such as norms and identity. We use Ostrom's social-ecological systems framework to structure our analysis across biophysical, economic, and social domains, linking scientific uncertainty to real-world implementation barriers. The review culminates in a set of research priorities designed to advance transdisciplinary work on cover cropping, clarify its climate mitigation potential, improve the design of private and public interventions, and support adaptive management.

Groundwater irrigation plays a crucial role in sustaining India's food security and rural livelihood. However, the carbon footprint of its energy (72% coal-based) requirement is also equally alarming. Under such scenarios, the Farm Power Subsidy (FPS) policy for groundwater irrigation, one of India's costliest public support systems, needs a relook, keeping in view the Sustainable Development Goals (SDGs). Although aimed at bridging social disparities, it is observed that FPS encourages simplification and intensification of the food production system, leading to inter-state and district scale inequality in income. States implementing FPS, even if experiencing physical water scarcity, are growing water guzzling crops for economic benefits. Conversely, without adequate access to energy, states reeling from economic water scarcity are witnessing low farm incomes. India consumes about a quarter of the world's groundwater for irrigation, and also leads the energy use. Notably, since the mid-2000s large-scale implementation of the FPS policy, the CO₂ emission (equivalent) has been rising by 5.77 million tons annually, now surpassing 100 million tons. These outcomes bear testimony to the globally accepted view that those least responsible for causing greenhouse gas emissions are the hardest hit. We suggest repurposing of agricultural policies, without compromising farmers' income, for efficient management of energy and water.

Carbon emissions from land use change (LUC) play a critical role in the global carbon budget and serve as a fundamental basis for national land-use planning. However, existing estimates are typically derived from direct human activities, while the environmental indirect effects driven by both natural and anthropogenic factors are often overlooked. To date, no studies have explicitly differentiated between the direct and environmental indirect effects of interannual LUC. This omission may bias estimates of LUC-induced carbon emissions and undermine the effectiveness of carbon reduction policies. Here, we proposed a novel accounting method that employed a process-based dynamic vegetation model to quantify interannual LUC-induced carbon fluxes (LUCF) in China from 1990 to 2020, further distinguishing them into direct (LUCF-d) and environmental indirect fluxes (LUCF-ind). The results indicate that China's overall LUCF from 1990 to 2020 functioned as a carbon source, with a trend shifting from a source in the 1990s to a sink in the past decade. Over the past 30 years, LUCF-ind sequestered 1.39 ± 1.05 TgC yr⁻¹, offsetting approximately 25% of the emissions from LUCF-d. The South and Central regions are the primary areas for the trade-off between LUCF-d and LUCF-ind, with indirect environmental effects reducing direct emissions by 13.3% and 13.7%, respectively. This study aims to enhance the effectiveness of national-level land use carbon reduction policies by integrating environmental indirect effects and providing methodological references for other countries.

Atmospheric rivers (ARs) are key drivers of hydrological change, but their role in sustaining and reshaping hydrological responses in Chinese river basins remains insufficiently quantified. We detect ARs for 1950–2023 using integrated water vapor transport (IVT) dual percentile thresholds with morphological filtering for elongated, persistent features; attribute precipitation, extreme precipitation, runoff, and soil moisture via a 1.5° axis buffer; model contributions with zero-inflated beta (ZIB) regression; and trace moisture sources during AR-driven floods with TROVA. Results reveal a south-to-north contrast, with southern basins exhibiting high summer IVT (∼600–1,100 kg m⁻¹ s⁻¹) from long ARs (∼13,000 km) that sustain antecedent wetness and amplify floods, while since the 1980s ARs have shortened and weakened, and hydrologic responses have declined in central and northern basins. Basin mean ARs contributions peak in the Huai River Basin (16.5%, 70.0%, 17.9%, 3.2%) and are lowest in the Southwest Basin (0.1%, 0.8%, 0.1%, 0.04%) for precipitation, extreme precipitation, runoff, and soil moisture, respectively. ZIB indicates a declining mean AR influence on soil moisture in the east and increasing shares of extremes in the southeast; antecedent soil moisture is the strongest covariate of AR-induced precipitation. During floods, dominant sources can reach 74.7% from the East Asia Monsoon (Yangtze River Basin, 1998), 55.6% from the West Pacific Tropics (Pearl River Basin, 2009), and 32.5% from the South Asia Monsoon (Pearl River Basin, 1985). These findings show that ARs sustain water availability and reshape hydrological responses, informing flood risk management and water security planning in a warming climate.

There is an increasing interest amongst policymakers in understanding the implications of high-impact low-likelihood (HILL) risks for climate mitigation, adaptation and resilience. Whilst extreme sea level rise scenarios have been used and there is awareness of some HILL risks, in practice there are currently few scenarios which can be applied in risk assessments. Here we present two sets of HILL climate scenarios for the UK, complementing existing UK climate projections. Both are based around storylines describing physically-plausible changes, were developed using observations, models and theory, and describe HILL drivers of change as inputs to impact models or stress tests. The storylines provide a narrative framework for understanding risk, and indicative quantifications provide the basis for quantitative risk assessments. One set describes six storylines for transient climate change to 2100 and beyond, reflecting plausible forcings and system responses outside the range conventionally assumed. These describe enhanced global warming, rapid reductions in aerosol emissions, volcanic eruptions, enhanced Arctic Amplification, changes to ocean circulation, and accelerated sea level rise. The other set describes extreme monthly and seasonal anomalies, representing hot, cold, wet, dry and windy extreme years. This set includes storylines describing persistently anomalous weather.

In a global-scale nuclear war, massive explosions, intense heat, and radioactive fallout would cause extensive harm to humanity and ecosystems. Further, previous studies of even regional-scale nuclear conflicts show that the smoke from large-scale fires caused by such weapons could lead to global-scale ozone loss. However, combustion studies show that urban fires release key ozone-depleting substances that were not previously considered in nuclear war studies, particularly chlorine and bromine compounds. Recent wildfire studies have also shown that high solubility of hydrochloric acid in oxidized organic smoke particles can greatly enhance chlorine-driven ozone loss. For the first time, here we simulate the impacts of a nuclear war on the ozone layer using a chemistry-climate model that accounts for fire-related halogen emissions as well as HCl solubility and heterogeneous chemistry in smoke particles. Our results show that a regional war scenario with 5 Tg of soot could result in a ∼40% reduction in the global ozone burden, nearly twice as much as previous studies. The calculated ozone losses exceed ∼80% over the Arctic, comparable to those observed when the Antarctic ozone hole was discovered and hence represent an Arctic ozone hole. Ozone losses also reach ∼50% over mid-latitudes in the Northern Hemisphere, including highly populated areas. The enhanced ozone loss compared to previous studies is attributable to combined non-linear effects from the incorporation of halogen emissions and updated heterogeneous chemistry in smoke particles. Such ozone losses lead to a large increase in surface ultraviolet exposure, posing grave risks to humanity and ecosystems.