Soil organic carbon stocks (SOCS) play a critical role in mitigating atmospheric CO2 and enhancing ecosystem resilience, yet their controls in high-mountain environments remain insufficiently constrained. Using an integrated framework combining field sampling, remote sensing, machine learning, and multi-source environmental datasets, this study quantifies SOCS variability along elevation gradients across the climatically diverse region of Gilgit-Baltistan, northern Pakistan. Results reveal three dominant patterns. First, SOCS distribution is jointly regulated by elevation and climate, with humid mid-to-high altitudes functioning as net carbon sinks, whereas lower-elevation semi-humid and semi-arid regions exhibit persistent SOCS declines. This spatial-altitudinal coupling highlights the sensitivity of mountain SOCS to aridity across both vertical and horizontal landscapes. Second, climatic drivers exert contrasting effects: precipitation enhances SOCS accumulation in humid zones by promoting vegetation productivity, while rising temperatures in semi-humid and semi-arid belts accelerate carbon losses through enhanced decomposition. Third, vegetation-soil interactions amplify sequestration potential along elevation gradients. Grasslands developed on loamy soils at mid-to-high elevations store significantly higher SOCS than other land use-soil combinations, while carbon stocks decline sharply with soil depth, reflecting the vertical constraints imposed by shallow mountain soils. Together, these findings demonstrate that SOCS dynamics in fragile high-mountain ecosystems emerge from the interaction of elevation, climate variability, vegetation cover, and soil properties. Recognizing this multidimensional control is essential for designing zone-specific management strategies that enhance carbon sequestration, reduce vulnerability to climate change, and support regional contributions to global climate mitigation and sustainable development goals.

This study combined continuous field monitoring of soil moisture dynamics in maize fields with stable isotope techniques and numerical modeling to clarify field water balance, evapotranspiration processes, and support efficient water use and stable high yields in arid regions. A full-process quality-control system covering sampling, extraction, and modeling was established, with dynamic monitoring across seedling, jointing, and filling stages. Stratified soil sampling (0–200 cm at 10 cm intervals) and multi-organ plant sampling with liquid nitrogen flash-freezing were performed, achieving analytical precision of δ¹⁸O ≤ 0.1‰ and δ²H ≤ 1‰. We developed and validated revised Green-Ampt, Bayesian water source partitioning, and isotope fractionation models. Maize organ fractionation coefficients ranked as leaf (6.5‰) > bract (4.8‰) > tassel (4.5‰) > stem (2.9‰). Isotope tracing showed infiltration depths of 20–40 cm, 60–80 cm, and 100–120 cm for light, moderate, and heavy rainfall, respectively. The Bayesian model revealed stage-dependent water use patterns, and a leaf δ¹⁸O irrigation threshold of −5‰ (Y=−32.5δ¹⁸O−112.5, R²=0.82, p<0.001) was identified, above which yield loss exceeded 20%. The revised infiltration model (NSE=0.85) reduced prediction error to ±15%, and a non-destructive root depth inversion model (R²=0.87) was constructed. Bract and tassel isotopes were verified as novel water-use efficiency indicators. These models and thresholds provide practical tools for precision water management in arid maize-production systems.

Soil erosion and its coupled interaction with water environment have become critical threats to drinking water security in the hilly red soil regions of southern China. This study targeted the Luofu River Watershed, a typical agricultural headwater catchment in the upper reaches of Dongjiang River Basin. We integrated the Revised Universal Soil Loss Equation (RUSLE), field monitoring, geospatial analysis, and principal component analysis (PCA), coupling coordination degree model and Pearson correlation analysis to systematically quantify the spatiotemporal patterns, driving mechanisms of soil erosion, its coupling relationship with water environment, and the comprehensive efficiency of soil and water conservation (SWC) measures. The results revealed that hydraulic erosion is the dominant erosion type across the watershed. The average soil erosion modulus was 892 t·km⁻²·a⁻¹, which falls into the mild erosion grade at the watershed scale, while local high-intensity erosion hotspots were widely distributed. Moderate and intense erosion areas only accounted for 8.38% of the total watershed area, and these regions were mainly concentrated in densely populated agricultural villages. Sloping farmland acted as the primary sediment source, contributing 67.2% of the total soil loss. Significant positive coupling relationships were detected between soil erosion and water quality degradation: the erosion modulus showed extremely significant positive correlations with sediment concentration and total phosphorus (TP) concentration, with correlation coefficients of 0.96 and 0.93 respectively (P < 0.001). TP derived from soil erosion contributed 78.5% to the total agricultural non-point source TP loading, which pushed the water environment carrying capacity (WECC) close to the early warning threshold. Natural factors explained 58.2% of the total variability in soil erosion, among which topographic (LS) factor and rainfall erosivity (R) were the dominant natural controls. Anthropogenic activities accounted for 41.8% of erosion variation and served as key amplifiers for local high-intensity erosion. The combined application of engineering, ecological and agronomic SWC measures reduced total soil loss by 60.1% and TP loading by 45.8%. The zoned classified management framework proposed in this study can provide scientific support for soil erosion control and drinking‑water‑source protection within the Luofu River watershed in the upper Dongjiang River Basin, and may serve as a potential practical reference for comparable small agricultural headwater catchments in southern China’s red‑soil hilly regions. Further local calibration would be required before direct application to other watersheds.

Vegetation restoration is reshaping dryland river basins, yet its hydrological consequences remain uncertain because greening can both enhance land-surface water retention and increase biological water consumption. This uncertainty is particularly relevant for the Yellow River Basin, where large-scale ecological restoration has coincided with marked changes in climate and runoff. Here we examine naturalized runoff, climate variability and vegetation greenness across the basin from 1982 to 2018 to determine whether post-restoration greening was associated with altered climate–runoff relationships. Natural runoff changed from a broad decline during 1982–2000 to a post-2000 recovery, although the recovery was not spatially uniform. Precipitation remained the primary positive predictor of runoff, while NDVI emerged as the most important secondary predictor. Temperature and solar radiation showed weaker and generally negative associations. Statistical pathway decomposition further indicated that vegetation greenness modified how climatic signals were associated with runoff. The area where NDVI-mediated pathways reinforced precipitation effects increased from 54.19% to 61.47%, and the area where they partly offset negative temperature associations increased from 33.41% to 47.83%. These results suggest that vegetation recovery did not produce a uniform hydrological response, but was linked to a reorganization of climate–vegetation–runoff associations. Restoration assessments in dryland basins should therefore consider not only vegetation gains, but also how greening changes runoff sensitivity to climate.

Ecosystems along the eastern margin of the Qinghai-Tibet Plateau (EQTP) are highly fragile and extremely sensitive to climate change and human disturbances. To quantitatively assess climate-induced ecosystem responses, this study proposes a Climate-Induced Productivity Index (CIPI) based on the Super Slack-Based Measure (Super-SBM) model using remote sensing data from 2001 to 2020. The results reveal persistently low CIPI values (0.47-0.53) across major ecosystem types, indicating widespread vulnerability to climatic variability. Among these ecosystems, forests exhibit the highest CIPI (0.55), followed by shrublands (0.54), croplands (0.53), grasslands (0.51), and barelands (0.43). The Theil index analysis further demonstrates significant intra-group disparities, suggesting that extreme climatic events amplify CIPI heterogeneity. Moreover, the dominant environmental drivers differ among ecosystem types: the Palmer Drought Severity Index (PDSI) primarily constrains grassland productivity, solar radiation (SRAD) strongly influences shrub and cropland systems, whereas subsurface factors exert greater control in forested regions. This study provides a quantitative framework for evaluating climate-ecosystem interactions and offers a scientific basis for long-term ecological monitoring and security planning across the EQTP.

With the implementation of various ecological restoration and regional development policies, both urbanization and soil conservation in the Loess Plateau (LP) have advanced substantially, but conflicts between development and the ecological environment are becoming increasingly prominent. Soil conservation is a key regulating ecosystem service, and understanding its interaction and coupling mechanism with urbanization is crucial for achieving sustainable urban development. This study used multi-source remote sensing data and evaluated soil erosion and conservation status over the past 30 years based on the revised universal soil loss equation (RUSLE) model. In addition, the coupling coordination degree model (CCDM) was employed to investigate the spatiotemporal characteristics and reveal the mechanism of the coupling coordinate degree (CCD) between soil conservation rate (SCR) and urbanization index (UBZ). Results indicate that: (1) soil erosion intensity (SE) has been well controlled in the Loess Plateau in recent decades, and the SCR has an obvious upward trend, but there is a risk of soil erosion worsening in the urban agglomeration area; (2) the coupling coordination degree between UBZ and SCR has improved significantly in the research period, but it is still concentrated in the southeast of the Loess Plateau and around the provincial capital cities; (3) there is a very significant logarithmic function fitting relationship between UBZ and CCD. At present, UBZ is the main factor leading CCD, and cities need to pay attention to controlling SE to maintain the stable growth rate of CCD when they develop to a certain period. This study pioneers the integration of multi-source remote sensing data with the RUSLE-CCDM framework to quantify a logarithmic relationship between urbanization and soil conservation coupling in the LP, proposing a phased constraint-incentive-restoration governance strategy aligned with environmental thresholds.

Mass coral bleaching has become a hallmark ecological signature of anthropogenic climate change, yet the structural mechanisms governing its spatial and temporal evolution remain poorly resolved. Traditional assessments emphasize surface thermal anomalies and strong El Niño phases but increasingly fail to explain the persistence of bleaching, its regional asymmetry, and the instability of ecological thresholds. Here, using long-term datasets (1993-2020) of ocean temperature, bleaching alerts, El Niño-Southern Oscillation (ENSO) indices, coral-algal cover, and subsurface thermal profiles, we identify a regime shift from episodic bleaching toward chronic thermal exposure. This shift reflects a reduced dependence of bleaching on ENSO intensity, as even moderate ENSO phases, including both El Niño and La Niña events, are now associated with widespread bleaching, suggesting progressive erosion of thermal thresholds under sustained ocean warming. We further propose a "Structure-Pathway-Response" framework to capture the structural reconfiguration of bleaching risk: (i) heat convergence along eastern continental margins (e.g., East Asian Seas, Caribbean); (ii) poleward transport and topographic retention of subsurface heat along continental slopes; and (iii) vertical accumulation through isotherm deepening that elevates bleaching risk. We identify two dominant heat-retention regimes: a current-deflection mode in the Northern Hemisphere and a thermal-stacking mode in the Southern Hemisphere. These patterns increase regional vulnerability by broadening the spatial extent and persistence of thermal anomalies, reflecting processes not fully captured by surface Sea Surface Temperature (SST) variability. Our findings highlight the limitations of surface-only monitoring systems and underscore the need for thermodynamically informed, region-specific early-warning frameworks. Protecting structural thermal refugia and managing heat pathways will be critical for sustaining coral reef resilience in a rapidly warming ocean.

This study evaluates grain production eco-efficiency (GPEE) and constructs a Rural Land Management Institution (LMI) index to quantify rural land policies in China and examine their impacts on GPEE in the Yellow River Basin (YRB). A hybrid analytical framework integrating an environmental preference-based Super-SBM model, Tobit regression, and mediation analysis is applied to socio-economic and ecological data from 53 cities in the YRB during 2000-2021. The main findings are as follows: (1) GPEE in the YRB was generally inefficient and remained below the national average throughout the study period. (2) Significant spatial and temporal heterogeneity is observed in GPEE, pure technical efficiency, and scale efficiency, with positive spatial autocorrelation characterized by high-high (H-H) and low-low (L-L) clustering patterns. (3) Pure technical efficiency contributes more to overall GPEE than scale efficiency, which remains relatively weak. (4) Changes in arable land are strongly influenced by rural land policies, with pronounced regional heterogeneity in LMI across different areas. (5) Rural land policies affect GPEE through two main channels: grain production technology utilization capability and land scale management. (6) In balanced grain production and consumption areas (BGPCA), LMI significantly improves both pure technical efficiency and scale efficiency, thereby enhancing GPEE, whereas in major grain-producing areas (MGPA), the effect of LMI on GPEE is limited, indicating environmentally unsustainable production practices. These findings highlight the need for more targeted and region-specific policy interventions by the central government to promote sustainable improvements in grain production eco-efficiency.

Elevated CO2 (eCO2) significantly affect the carbon-water cycle in terrestrial ecosystems especially for gas exchange and water use efficiency (WUE). Therefore, in this study, we have conducted a meta-analysis to quantitative statistical means among studies and discuss how WUE responds to eCO2 under pathway (C3 and C4), four enrichment methods and soil types based on 124 peer-reviewed studies and 1474 observations to provide an in-depth overview of how these factors interact under future CO₂ scenarios. Key findings reveal that: (1) C₃ crops, such as potato and tomato, show significantly greater increases in WUE compared to C₄ crops like maize, with effect sizes of 13.96 and 7.02 for plant-level WUE (WUEₚ), suggesting that C₃ crops may be more advantageous in water-limited environments due to reduced photorespiration under eCO₂; (2) soil type substantially modulates WUE responses, with clay soils, due to their high water-holding capacity, demonstrating the highest WUE enhancements (effect sizes of 7.87 for WUEₚ and 12.54 for yield WUE, WUEᵧ), while sandy soils, characterized by rapid drainage, showed limited improvements; and (3) greenhouse and growth chamber studies displayed the highest WUE improvements, while FACE experiments, which better replicate real-world conditions, indicated smaller WUE increases due to environmental variability, underscoring the need for a hybrid approach that merges controlled data with field insights to develop practical, water-efficient agricultural strategies. Collectively, these findings highlight the potential for crop- and soil-specific strategies to optimize WUE under elevated CO₂, offering valuable insights for sustainable agriculture and climate adaptation.

Sustainable coordination of the Water-Energy-Food (WEF) nexus in the Yellow River Basin (YRB) faces challenges due to complex terrain-climate-human interactions, with their coupling mechanisms underexplored. To address this issue, we developed an innovative hybrid framework integrating the Super-SBM model, game theory combination weighting based Coupling Coordination Degree (CCD) model, Random Forest, and PLS-SEM to analyze the spatiotemporal patterns and driving mechanisms of WEF nexus across 53 YRB cities from 2003 to 2023. Our key findings reveal: (1) WEF efficiencies in the YRB remain low (WRUE = 0.57, EUE = 0.54, GPE = 0.79) with divergent trends-WRUE and EUE fluctuate, while GPE steadily improves. High-efficiency cities (e.g., Ordos, Yulin) contrast with low-efficiency ones (e.g., Xi'an, Zhengzhou), revealing spatial mismatches between resource endowment and functional roles. (2) The basin-wide WEF Nexus Coupling Coordination Degree (WEFCCD) exhibits an inverted U-shaped trajectory (0.57 -* 0.62-*0.58) and, overall, reflects a marginal coordination level. While 51 % of cities show improvement, 49 % experience a decline, highlighting development pathways that are largely influenced by resource endowments. (3) PLS-SEM results, interpreted via the PSR framework, reveal spatially differentiated drivers: in the upstream, terrain-induced pressures dominate and constrain coordination (path coefficient = -1.24); in the midstream, terrain and erosion pose ongoing pressures, but warming temperatures improve state conditions and crop yields (+0.36), with terrain-climate interactions shaping human impacts; in the downstream, strong human pressures and effective responses-such as innovation-enhance WEF coordination (+0.48). Based on these insights, we propose a zoning-based governance strategy to align regional resources with targeted WEF management.

This study aims to elucidate the adaptation mechanisms of prehistoric settlements in the Yellow River Basin to extreme climate events spanning from the pre-Yangshao period (9.0-7.0 cal. ka BP) to the Bronze Age (4.0-2.0 cal. ka BP). By integrating archaeological site data with ancient environmental records and employing spatial statistical methods such as kernel density estimation and nearest neighbor analysis, this research reconstructs the spatial and temporal evolution of settlement patterns under climate stress. Furthermore, it discusses the relationship between these patterns and natural disasters, uncovering the mechanisms of human-land interaction and the adaptive strategies that shaped regional resilience during periods of environmental stress. The results reveal three key findings. Firstly, the distribution of settlements exhibits significant spatial differentiation; while the middle reaches consistently maintained high density, a secondary core emerged in the upper Hehuang Valley during the Yangshao period. Secondly, settlement patterns demonstrate simultaneous trends of expansion in scale and reduction in number, indicating an increase in social complexity and stratification, which laid the groundwork for early state formation. Thirdly, the selection of settlement sites reflects clear climate adaptation: the average elevation rose by 56.98 m from the Yangshao to Longshan periods, and the proportion of settlements located farther from water sources increased by 12.3 % during the late Longshan period, strongly correlating with flood sediment records. Additionally, erosion and sedimentation in downstream areas provided fertile land, playing a crucial role in sustaining social development. This study illustrates that mid-Holocene climate variability, particularly in the form of intensified flooding and environmental pressures, drove adaptive strategies such as elevation migration and spatial differentiation. These dynamics, coupled with emerging sociopolitical structures, facilitated the rise of early complex societies in the middle and lower Yellow River Basin.

Quantifying soil denudation, sediment transport, and deposition processes in watersheds over millennial timescales is crucial for a comprehensive understanding of soil erosion dynamics. However, most current studies on soil erosion over such timescales primarily rely on the qualitative interpretation of field-based sedimentary records. Southwest China underwent substantial environmental changes during the Holocene, with soil erosion being notably influenced by long-term human activities. In this study, we selected a crater lake catchment with a straightforward sediment source as the case study area. The Landlab landscape evolution model was employed to simulate and reconstruct temporal and spatial soil erosion processes. Higher erosion rates were observed during 10.9-10.3 ka, 8.8-7.9 ka, 3.9-3.2 ka, and after 1.0 ka, with the lowest level recorded between 7.7 and 4.2 ka. During the Holocene, approximately 64 % of the catchment area underwent denudation in gullies and mountains, while 36 % of the area, including flatlands and lakes, was deposited. Our findings indicate that vegetation exerted a significantly greater impact on soil erosion compared than precipitation. The simulated erosion rate increased rapidly after similar to 1.0 ka with an average increase of 1.2 times, driven by human activities. This study highlights the need for additional field geological evidence to validate the model.

Investigating the history of wind-dust deposition and soil erosion is crucial for understanding the relationship between geomorphologic formation, monsoon evolution and the current state of soil erosion on the Loess Plateau. This study utilizes a 50-m sedimentary record from Lake Luyanghu (LYH) in the southern Loess Plateau. We developed a chronological framework using optically stimulated luminescence (OSL) dating and stratigraphic comparisons. By applying end-member modeling of grain size, we identified various sediment sources and quantitatively reconstructed the dust deposition fluxes and soil erosion modulus during the late Quaternary. Additionally, we examined the patterns of sediment source evolution with environmental changes at LYH. Our findings reveal that dust constitutes an average of 32.3% of the lake sediments, with an average deposition rate of 40.2 g/cm2/ka. Notably, after the Last Glacial period, dust deposition became the dominant component of the lake sediment. During the Marine Isotope Stages (MIS) 5b, 5d, 4 and 2, enhanced dry climatic conditions, decreased vegetation cover and composition, leading to significant soil erosion. Based on the results of this research, future studies should aim to reconstruct the watershed's erosion and deposition processes throughout geological periods by integrating aspects of lake evolution, such as water levels and developmental stages.

Study Region: Egypt is a country located in northeastern Africa. Study Focus: The research evaluated the random forest (RF) and extreme gradient boosting (XGB) as single models and the models' hybrid to predict the ETo for the baseline and future (2015-2099) period from Shared Socioeconomic Pathways (SSP1-26, SSP2-45 and SSP5-85) based on 18 GCMs models. New Hydrological Insights for the Region: The hybrid model has performed better than single models; compared RF and XGB to RF-XGB, the RMSE values were decreased in all zones esepically in zone 3 by 16.2 %, these results indicate that the highest performances of all models are observed in the middle and south Egypt, which exhibit the strongest correlation between temperature and ETo. For the SSP5-8.5 scenario, the ETo increased over the years for all zones; the ETo will increase by 4.38 %,3.71 %, 4.27 %, 2.16 %, 3.26 %, 1.35 %, 5.22 % at the year 2099 compared to the year 2015 for zone 1, 2, 3, 4, 5, 6 and 7 respectively. The T-min and T-max are the most critical factors that affect the ETo in all zones in the baseline and future scenarios. This study provides important insights into applying machine learning models to estimate ETo and its implications for future water management strategies. Such models hold promise for significantly enhancing regional agricultural water-resource planning and management.

Ensuring the security of China’s rice harvest is imperative for sustainable food production. The existing study addresses a critical need by employing a comprehensive approach that integrates multi-source data, including climate, remote sensing, soil properties and agricultural statistics from 2000 to 2017. The research evaluates six artificial intelligence (AI) models including machine learning (ML), deep learning (DL) models and their hybridization to predict rice production across China, particularly focusing on the main rice cultivation areas. These models were random forest (RF), extreme gradient boosting (XGB), conventional neural network (CNN) and long short-term memory (LSTM), and the hybridization of RF with XGB and CNN with LSTM based on eleven combinations (scenarios) of input variables. The main results identify that hybrid models have performed better than single models. As well, the best scenario was recorded in scenarios 8 (soil variables and sown area) and 11 (all variables) based on the RF-XGB by decreasing the root mean square error (RMSE) by 38% and 31% respectively. Further, in both scenarios, RF-XGB generated a high correlation coefficient (R2) of 0.97 in comparison with other developed models. Moreover, the soil properties contribute as the predominant factors influencing rice production, exerting an 87% and 53% impact in east and southeast China, respectively. Additionally, it observes a yearly increase of 0.16 °C and 0.19 °C in maximum and minimum temperatures (Tmax and Tmin), coupled with a 20 mm/year decrease in precipitation decline a 2.23% reduction in rice production as average during the study period in southeast China region. This research provides valuable insights into the dynamic interplay of environmental factors affecting China’s rice production, informing strategic measures to enhance food security in the face of evolving climatic conditions.

This paper presents a comprehensive evaluation of urban water resource utilization efficiency (UWRUE) in relation to comprehensive water resource utilization capabilities (CWRUC) across 53 cities in the Yellow River Basin (YRB) of China from 2003 to 2020. This study uses an innovative hybrid approach that integrates the Meta-frontier Super-SBM model and the Tobit model to address both efficiency and the factors that contribute to regional disparities in UWRUE across different technological frontiers. Our research not only uncovers significant inefficiencies in UWRUE but also identifies critical regional differences rooted in CWRUC, and reveals the resource curse effect in the YRB. Additionally, the analysis highlights that scale efficiency plays a more pivotal role for improving UWRUE than pure technical efficiency. One of the key contributions of this paper is the identification of the U-shaped relationship between disposable income and UWRUE, alongside the negative marginal impact of population density. Importantly, this study differentiates between the U-shaped relationship between industrial development and UWRUE in the midstream and downstream regions and the inverted U-shaped relationship observed in the upstream region. This paper offers nuanced insights into the dynamics of water resource management in the YRB and contributes to the broader discourse on sustainable urban water governance. The findings offer practical policy recommendations and particularly emphasize the need to promote water conservation awareness, advance water-saving technologies, and enhance governance frameworks to address regional disparities.

This study quantifies urban energy eco-efficiency (UEEE), examines the technological gap ratio (TGR) in regional energy use, and analyzes the impact of China's urbanization on UEEE. We propose a hybrid model that combines the Meta-frontier Super Epsilon-based Measure (Super-EBM) model and a Tobit model, using data from 271 Chinese cities between 2003 and 2020. The results indicate that the overall UEEE is ecologically inefficient throughout the study period. TGR has a positive impact on UEEE, with notable regional differences. However, the marginal effect of TGR on UEEE in the east region is lower than that in the other three regions. In addition, the urban population growth and UEEE exhibit a U-shaped relationship in the east and central regions, while an inverted U-shaped relationship is observed in the west region. The relationship between urban economic growth and UEEE is positive, accompanied with diminishing marginal effects in the central and west regions, whereas fixed asset investment negatively affects UEEE in the four regions, showing different marginal effects. These findings highlight the importance of considering technology and local urbanization in policy making.

The distribution and flow of geothermal energy in the North China region are influenced by geological structures and rock thermal properties. However, the interactions between geothermal energy, groundwater movement, and hydrogeochemical processes are not well understood. In this study, a numerical model was developed to simulate these interactions under different geological conditions. The results showed that the stress field of geological structures affects the flow path and velocity distribution of geothermal fluid, leading to distortion and uneven temperature distribution. The distribution of high-mineralized zones of geothermal water indicated the main recharge area in the northern part of the North China Basin (Henan). Hydrochemistry analysis revealed a dynamic balance between centrifugal and centripetal flows in sandstone geothermal water. The study provides important insights for the exploration and development of underground geothermal water, as well as the assessment and optimization of geothermal resources in large basins.