Flood risk assessment is essential for flood control and disaster mitigation in arid and semi-arid river basins, where conventional univariate and bivariate frequency analyses struggle to capture nonlinear dependence among flood variables and often underestimate extreme synergistic risks. This study focuses on the Wulanmulun River Basin in Inner Mongolia and employs long-term observations from the Zuanlongwan and Wangdaohengta hydrological stations. A trivariate D-vine Copula model was constructed to jointly characterize peak discharge, total flood volume, and water level. Optimal vine structures differ between the stations (Qp-H-W and W-Qp-H) and outperform traditional Copula models in representing extreme joint risks. The ternary joint return periods reveal two distinct flood risk transmission modes, "jump" and "accumulation", and joint exceedance probabilities under low, medium, high, and ultra-high-risk scenarios are 6.4%, 31.95%, 37.64%, and 5.75% at Zuanlongwan, and 4.7%, 35.24%, 45.78%, and 0.53% at Wangdaohengta, indicating concentration in medium-to-high risk ranges. The validation at Longtouguai Station showed an error RSME of 0.0630 and an R-2 of 0.905, confirming the reliability of the model framework. These results indicate that the proposed framework can effectively capture multivariate flood dependencies and provide a scientific basis for flood control design, risk zoning, and emergency management of small and medium rivers in arid and semi-arid regions.

Arid and semi-arid regions cover more than 30

The relationship between lacustrine groundwater discharge (LGD) and dissolved organic matter (DOM) dynamics remains insufficiently understood, particularly during the freezing season. In this study, the spatial and seasonal variability and controlling factors of DOM in surface water and groundwater were investigated in a representative ice-covered lake basin (Ulansuhai Lake, China). Specifically, the effects of LGD on DOM sources and transport during non-freezing and freezing seasons were examined using fluorescence excitation-emission matrix spectroscopy coupled with parallel factor analysis (EEM-PARAFAC) and a 222Rn mass balance model (RMBM). During the non-freezing season, DOM was primarily composed of humic-like components (C1-C3), whereas during the freezing season, in addition to humic-like components (C1-C3), protein-like component (C4) was also identified. Quantitative source apportionment indicated that effluent (mean: 39%), soil (mean: 27%), fertilizers (mean: 20%), and plants (mean: 14%) were the dominant DOM sources. During the non-freezing season, LGD was relatively weak (mean LGD rate: 5.12 mm/d), and canal water served as the primary recharge source for both groundwater and lake water; consequently, allochthonous humic-like substances (components C1-C3) dominated the DOM pool. In contrast, during the freezing season, LGD intensified markedly (mean LGD rate: 17.49 mm/d), becoming the sole recharge source for the lake and amplifying the relative contribution of protein-like DOM (C4). Moreover, hydrological conditions and vegetation cover further regulated DOM dynamics by modulating LGD processes. Overall, this study highlights the crucial role of LGD in governing DOM dynamics and demonstrates the effectiveness of integrating optical techniques with isotopic mass balance modelling to elucidate biogeochemical processes under seasonally variable environmental conditions.

River runoff and suspended sediments exhibit high dynamic and multiscale variability, particularly during rainstorm-driven floods in arid and semiarid regions. Despite the widespread use of the hysteresis loop model, flood-event-scale hysteresis patterns and their controls remain poorly quantified in these environments. This study investigated the hysteresis patterns between runoff and sediment and identified the dominant factors affecting the hysteresis index (HI) in a typical semiarid catchment in northern China. Forty-eight flood events were classified by K-means clustering using the runoff depth (H), flood duration (T), and peak flood flow (Qp) into three types: A (low-intensity, medium-duration floods), B (medium-variability, long-duration floods), and C (high-intensity, short-duration floods). The monthly runoff and sediment load were decomposed via multivariate empirical mode decomposition (MEMD) to extract scale-specific dynamics, whereas event-scale hysteresis patterns were quantified separately to assess sediment transport mechanisms during floods. Figure-eight loops occurred most frequently (31.3%), followed by counterclockwise (29.2%), complex (22.9%), and clockwise (16.7%) loops. HI is governed by distinct hydrological factors for each loop type: flood peak timing skewness (FT) for clockwise loops; Qp, suspended sediment yield (SSY), mean suspended sediment concentration (Sm) and peak suspended sediment concentration (Sp) for counterclockwise loops; T and high traffic duration (HT) for figure-eight loops; and flood variability (FV) for complex loops. By combining MEMD-based scale decomposition with objective K-means event classification and HI quantification, we provide a transferable, scale-aware framework for diagnosing sediment transport regimes in semiarid basins.

Surface water-groundwater interactions are complex in the Great Bend region of the Yellow River. Clarifying water exchange and source differences among geomorphic units is important for understanding regional hydrological processes and improving water resources management. In this study, the Wulanmulun River Basin, a typical tributary of the Yellow River, was selected as the study area. A total of 90 water samples were collected, including 25 river water samples, 43 groundwater samples, and 22 precipitation samples. Gibbs diagrams, multivariate statistical analysis, PMF, and MixSIAR models were used to investigate the hydrochemical and isotopic characteristics, source contributions, and transformation relationships of surface water and groundwater. The results showed that, under the combined effects of rock weathering dissolution and evaporation concentration, the hydrochemical types of surface water and groundwater in the study area were dominated by Cl-Na and HCO3-Ca types, respectively. The source contributions of hydrochemical components differed between surface water and groundwater. Groundwater was mainly controlled by geological processes (41.1%), followed by agricultural input (31.3%); surface water was also dominated by geological processes (36.1%), but the influence of domestic input was enhanced (34.1%). In the hilly regions, groundwater discharge mainly recharged surface water, with an average contribution rate of 32.4%. In the plain and built-up areas, surface water infiltration mainly recharged groundwater, with average contribution rates of 34.2% and 25.0%, respectively. These results enhance understanding of surface water-groundwater transformation in the Yellow River's "Great Bend" and support scientific management and efficient utilization of basin water resources.

Environmental change can alter storm-flood response, but whether it modifies the baseline response magnitude or rainfall-response elasticity remains unclear in semi-arid urbanizing basins. This study developed a stability-validated power-law modelling framework to examine event-scale flood peaks in the Ulan Moron River Basin, northern China. Candidate influencing variables were screened using calibration-set tests, validation consistency, and bootstrap stability, and alternative model structures were compared by allowing environmental covariates to affect the scaling parameter (alpha), the rainfall-response exponents (beta and gamma), or both. Results showed that maximum 1-h rainfall intensity and total rainfall amount were the dominant storm controls, while environmental effects were more consistently represented through alpha than through beta and gamma. The dynamic scaling parameter alpha t decreased significantly over the study period, indicating a downward shift in baseline storm-flood response magnitude. From the early to late period, combined environmental change was associated with an approximately 69.8% reduction in alpha, mainly linked to integrated urban hydrological development, with additional contributions from temperature, surface energy partitioning, and vegetation regulation. Fixed-storm scenarios showed that environmental modulation produced the same relative effect across storm magnitudes but larger absolute flood-peak changes under larger storms. These findings suggest that environmental change primarily shifted the flood-response scale rather than producing robust storm-type-dependent changes in rainfall-response elasticity. The inferred effects should be interpreted as conditional, basin-specific associations rather than definitive causal estimates.

Studying the precipitation characteristics and extreme precipitation events in arid and semi-arid regions is of significant baseline value for optimizing water resource allocation and utilizing precipitation resources. Utilizing multi-scale ERA5 precipitation data from 1960 to 2023, this study focuses on the typical arid and semi-arid region of Ordos as the research area. Precipitation exceeding the 90th percentile was defined as extreme precipitation, and three indices—extreme precipitation amount (EPA), extreme precipitation frequency (EPF), and extreme precipitation proportion (EPP)—were used to investigate its characteristics in the study area. Additionally, three typical extreme precipitation events in recent years were analyzed to study the precipitation process of these typical events. The main results are as follows: The annual average precipitation in the study area ranges from 170.3 to 606.1 mm, with an average of 378.5 mm, which has been on a declining trend over the years, with an average annual decrease of 1.2 mm. Overall, 70% of the precipitation is concentrated in the months of June to September. The daily average of extreme precipitation in Ordos is 18.7 mm and the annual average number of extreme precipitation days ranges from 8 to 13 days, with an average annual number of extreme precipitation days being 11. Extreme precipitation accounts for more than 50% of the total precipitation. Among all areas analyzed, Jungar Banner demonstrates the greatest vulnerability to intense rainfall events. Typical extreme precipitation events in Ordos are characterized by short-duration heavy rainfall, with the rain peak ratio coefficients of the three events ranging from 0.62 to 0.72, exhibiting a distinct “post-peak” pattern. These findings provide scientific support for water resource management and disaster prevention strategies in arid and semi-arid regions.

To gain a deeper understanding of the evolving characteristics of floods across different ecosystems and their meteorological driving conditions, in this study, a transferable, all-season, multi-ecosystem methodology for setting meteorological threshold warnings for floods of various magnitudes is established. An in-depth analysis of the differences in evolving flood characteristics and driving conditions between ecosystems that have been intensely disturbed by human activities and natural/seminatural systems is conducted. A flood management measure is proposed, which draws on the ecological resilience of forest and grassland systems. We emphasize that the philosophy of flood control should shift from hard-defence to adaptive governance. In the Yellow River Basin, the universal driving factors of floods are precipitation, humidity, and sunshine, with atmospheric pressure requiring additional consideration in water bodies and wetlands. Winter floods are most sensitive to sunshine percentage and precipitation, whereas floods in spring, summer, and autumn are primarily driven by humidity and sunshine hours. Since 1985, flood risk in the Yellow River Basin has entered a phase of persistent intensification. This intensification has accelerated across the entire basin in the 21st century and has reached 4–57 times the average rate of increase. Flood severity will continually rise throughout the basin. In this study, the peak flood months for each ecosystem are identified, with the most significant increases in flood momentum observed in January and September. The traditional summer-and-autumn flood season boundary has extended into winter and spring, indicating that flood risk in the Yellow River Basin has broken through traditional seasonal patterns and is intensifying across all seasons and multiple ecosystems. Human activities could increase flood risk. However, agricultural and settlement systems, which are entirely human-dominated, have the potential to mitigate floods in the long term through sustained management. Deserts are the most prone to triggering floods, grasslands carry the hidden danger of long-term persistent increases in flood risk, and floods in water bodies and wetlands are prone to evolving into high-intensity events. The research findings provide decision-makers with actionable pathways for flood early warning and adaptive management.

Lakes in cold regions that experience seasonal ice cover are potential hotspots of greenhouse gas (GHG) emissions. However, the synergistic roles of groundwater, as a concealed external source, and dissolved organic matter (DOM) in regulating CH4 and CO2 dynamics and emissions remain poorly understood. Based on year–round observations at Ulansuhai Lake, this study revealed seasonal differences in the regulatory effects of lacustrine groundwater discharge (LGD) and microbially derived DOM. During the ice–free period, the LGD rate was low (5.53 mm/d). CO2 dynamics were primarily controlled by microbial decomposition of DOM, whereas CH4 dynamics were mainly influenced by direct LGD input and were positively correlated with DOM aromaticity and humification. During the ice–covered period, the LGD rate increased markedly, reaching a maximum of 18.09 mm/d, and became the sole external source of CO2. In contrast, CH4 was predominantly regulated by DOM, driven positively by humic component C2 and the humification index (HIX), a process facilitated by anaerobic microbial fermentation. Flux calculations indicated that, during the ice–free period, water–air emissions of both gases exceeded the corresponding LGD–derived inputs. However, when considering only the LGD pathway, its regulatory effect was stronger for CH4 than for CO2. During ice cover, water–air gas exchange ceased, while LGD–derived CO2 inputs (0.049–1.720 mmol/m2/d), together with CH4 produced via DOM–driven anaerobic fermentation, sustained greenhouse gas accumulation beneath the ice. These findings identify LGD as a concealed driver of CH4 and CO2 emissions with seasonal shifting roles and highlight its importance in carbon budget assessments of seasonally ice–covered lakes in cold regions. This study advances understanding of groundwater–derived carbon inputs in lake carbon cycling and provides both hydrological and biogeochemical insights for compiling greenhouse gas emission inventories from seasonally ice–covered lakes.

Under the combined effects of climate change and human activities, the water-sediment processes in the Yellow River Basin demonstrate significant evolutionary characteristics. Based on long-series hydrological and sediment data from the Inner Mongolia reach of the Yellow River collected between 1919 and 2024, this article investigates the water-sediment relationship using trend-sudden change-cycle identification and segmental diagnosis methods. Combined with data on reservoir construction and operation, land use, and vegetation index, it analyses the mechanisms driving changes at different phases and quantifies the contribution rate of each factor. The results obtained include: (1) Runoff and sediment transport on a centennial scale generally exhibited a significant decreasing trend (p < 0.01). Compared with the high flow and sediment period P2 (1933-1967), the average annual runoff at various stations during the multi-reservoir coordinated operation period P6 (2014-2024) decreased by 7.0-28.6%, while sediment transport decreased by 62.7-90.7%, with an even more significant reduction during the flood season. (2) The runoff volume changed abruptly in 1933, 1969, and 1986, and an additional abrupt change in sediment transport occurred in 2000, which coincided with the construction period of upstream hydraulic engineering. A cycle analysis showed that there were 38a, 14a, and 7a cycles in the runoff, while the sediment transport presented multi-scale fluctuation characteristics of 2-4a, 7-14a, 22-24a, and 46a. (3) The water-sediment relationship underwent a significant transition from a natural dominant phase (P1, P2) to a phase driven by both human activities and climate after 1968. Subsequently, human activities became the major factor influencing the modern water-sediment process, with contributions of more than 55% to runoff and more than 79% to sediment transport. (4) The construction of cascade reservoirs had a significantly stronger effect on sediment transport (mainly during the flood season) than on runoff. From 1980 to 2020, the expansion of forest land (+10.8%) and grassland (+1.3%) contributed to sediment reduction. On the other hand, the expansion of construction land (+37.1%) and the continuous increase in the Normalized Difference Vegetation Index (NDVI) (0.026/10a) led to an increase in evapotranspiration and ecological water use, which became significant factors in runoff reduction. This study reveals the driving mechanisms of water-sediment changes in the Inner Mongolia reach from both long-term and multi-phase perspectives. These results may serve as a quantitative reference for the water-sediment regulation and ecological protection of the Yellow River basin.

To promote the efficient utilization of water resources and coordinated socio-economic development in the Jizi Bend of the Yellow River， based on data from Ordos City from 2010 to 2023， a water resources carrying capacity （WRCC） evaluation system was constructed covering three dimensions： water resources， socio-economics， and ecological environment. A game-theoretic approach was applied to determine indicator weights by integrating the analytic hierarchy process （AHP） with subjective weighting and the entropy weight method with objective weighting. The Technique for Order Preference by Similarity to Ideal Solution （TOPSIS model） was used to comprehensively assess water resources carrying capacity levels and reveal their temporal evolution， while an obstacle-degree model was employed to quantitatively identify the main factors restricting WRCC improvement. The results indicate that： ① From 2010 to 2023， the overall water resources carrying capacity in the study area showed an upward trend； however， as of 2023， all evaluation units remained at Level III to IV， falling short of the acceptable carrying standard， indicating that the water resources carrying capacity still requires improvement. ② A core-periphery spatial differentiation pattern was observed within the region. The carrying capacity in the central urban area improved rapidly， reaching Level IV by 2023， whereas in the peripheral areas， due to factors such as water use structure and policy constraints， the improvement was slow， with the carrying capacity remaining at Level III over the long term. ③ Per capita water availability， annual precipitation， irrigation water use per unit of farmland， and COD emissions per 10 000 yuan of output value were the main obstacle factors limiting the improvement of regional water resources carrying capacity. The research findings can provide theoretical support and decision-making references for the optimal allocation of water resources， ecological environmental protection， and sustainable socio-economic development in resource-based cities of the Yellow River Basin.

Understanding how urbanization intensifies precipitation, especially in water-sensitive arid regions, is critical for climate adaptation. This study investigates the nonlinear and scale-dependent responses of precipitation to urbanization in the Ordos Plateau, China, an ecologically fragile arid region experiencing rapid urban expansion. We parameterized urbanization intensity and precipitation amplification to quantify their relationship across different quantiles (tau) and time scales using a mixed-effects nonlinear quantile regression model. This approach allowed us to control for site-specific variability and isolate the generalizable urbanization effect. Our results reveal a robust urbanization intensity-dependent precipitation enhancement. On average, across the time scales considered, a tenfold increase in urbanization intensity amplifies extreme precipitation (tau = 0.99) by approximately 0.590 units, which is about 4.6 times greater than its effect on light precipitation (tau = 0.01). Furthermore, the marginal contribution of urbanization to precipitation amplification diminishes with increasing urbanization levels but remains persistently higher for extreme events. Crucially, these nonlinear effects are more pronounced at monthly and seasonal scales than at the annual scale, underscoring a heightened vulnerability to short-term extremes. These findings reveal that urban expansion in arid and semi-arid regions disproportionately exacerbates extreme precipitation risks. It is therefore suggested that urban planning, water resource management, and disaster prevention policies integrate this nonlinear, urbanization intensity-dependent relationship to enhance climate resilience.

Instream ecological flow plays a crucial role in maintaining the fundamental functions of river ecosystems, but the uncertainty in its estimation is increasing under a changing environment. Considering the non-stationary behaviors of instream low flows is expected to enhance estimation precision. In this study, a non-stationary estimation scheme was proposed to quantify the dynamic processes of ecological flow at multiple time scales. Based on the deterministic components (trend, shift, and cycle) of the low flow time-series at a given time scale, time-dependent probabilistic fitting models were developed by considering different time-varying forms. The best model was then selected to improve the traditional 7Q10 method for estimating instream ecological flow. Focusing on a typical tributary of the Yellow River, the variations in its low flows at annual, seasonal, and monthly time scales were found to be dominated by periodic characteristics, with primary periods ranging from 22.3 to 49 years. The cause of this is possibly the water conservation capacity of the Kubuqi Desert. The results show that the cyclical time-varying model performed best in fitting temporal changes in low flows, contributing to the estimated ecological flows with clear cyclic variations. The model integrating periodic characteristics exhibits potential in enhancing the adaptability and interpretability to hydrological changes. In general, the ecological flow estimated for high-flow years was higher, while that for low-flow years was lower, aligning closely with hydrological periodic changes. These findings provide novel insights into ecological flow estimation and serve as a basis for decision-making in river ecological protection.

The asymmetric erosion and deposition patterns in river bends result in more pronounced erosion on the concave bank. Sediment transport, serving as the interaction medium between river flow and bank slope, plays a vital role in shaping the evolution of river bends. Thus, an in-depth investigation into the critical conditions for sediment initiation is essential for elucidating the stability mechanisms of river bank slopes. This study employs an innovative approach by integrating mechanical analysis with probabilistic methods to develop a mechanical equilibrium model for the critical sediment initiation on the concave bank. The model accounts comprehensively for factors such as the distribution of centrifugal forces, the pressure gradient arising from lateral pressure differences, and changes in flow momentum influencing sediment initiation within the bend. Furthermore, the model incorporates the stochastic nature of sediment particle distribution and exposure on the concave slope, calculating the expected relative exposure for uniform and non-uniform sediments. It subsequently determines the exposure coefficient for the additional mass force, ultimately deriving a formula to calculate the critical flow velocity required for sediment initiation on concave banks. Comparison with existing standard formulas for sediment initiation velocity in river bends reveals that the proposed formula demonstrates superior accuracy, with a root mean square error of only 0.0539. The findings indicate: 1) The lateral flow velocity in river bends is governed by mainstream velocity, curvature radius, and water depth, with the curvature radius influencing sediment initiation on the concave bank through lateral circulation forces. 2) Ignoring the effects of lateral circulation force, changes in flow momentum, and additional mass force results in calculated sediment initiation velocities being significantly lower than the measured values. 3) The sediment initiation velocity is primarily determined by effective gravity, lateral circulation forces, additional mass force, and the forces arising from changes in flow momentum, with the impact of each force growing as sediment particle size increases. 4) The concave bank slope indirectly affects the sediment initiation velocity by altering the components of each contributing force; a steeper slope facilitates easier mobilization of sediment.5) The effect of unsteady flow disturbances on sediment incipient velocity exhibits significant nonlinear characteristics with respect to particle size and disturbance intensity. 6) The momentum correction coefficient requires reasonable selection to avoid introducing non-physical errors through excessive correction. This study not only enhances the understanding of sediment initiation mechanisms on concave banks of river bends but also offers scientific evidence for river management and bank stability maintenance.

Characterizing hydrological dynamics and controlling mechanisms under extreme weather events is essential for ensuring water security and early-warning of disasters in semiarid regions. Isotope analyses for different waters were conducted in a pronounced drought-flood abrupt alternation period within the Wulanmulun River basin of the Inner Mongolia semiarid region. The evaporation characteristics of different waters in the basin were assessed based on isotopic records during this critical period. The results revealed that isotopes in lake water were the most enriched, followed by those in river water, and those in groundwater were the most depleted in both dry and wet periods. Conversely, the d-excess values exhibited the opposite pattern, indicating that the intensity of evaporation followed the order of lakes > rivers > groundwater. The spatial isotopic variation of river water was driven predominantly by evaporation, overshadowing the effects of elevation changes. The Craig-Gordon model revealed that evaporation losses for both rivers and lakes were significantly greater during the dry period than during the wet period. The evaporation loss in the mainstream Wulanmulun River water displayed marked spatial heterogeneity relative to the tributary, largely attributable to reservoirs that extended the water residence time, whereas impervious surfaces and urban drainage systems downstream accelerated the influx of "fresh" precipitation into the river. Lakes exhibited notable variations in their evaporation over inflow ratio (E/I) in both dry and wet periods, primarily due to their distinct functions as artificial urban lakes, which resulted in varying water renewal rates. Relatively weak evaporation observed in groundwater demonstrated its potential in regulating water resources seasonally. The findings of this study provide valuable insights for adaptive water management and reservoir operations in vulnerable semiarid basins.

Flash drought has attracted worldwide attention in recent years because of its destructive effects on ecosystems. However, the research on ecosystem loss caused by flash drought is insufficient. In this study, a basis for further understanding of the impact of flash drought on grassland was developed, and a method was established for quantitatively assessing productivity and economic losses in various types of grassland impacted by different levels of flash drought. The results showed that grassland productivity anomalies, productivity loss, and economic loss caused by flash drought vary with grassland type, flash drought type, and flash drought stage. During flash drought, the average NPP loss rate (NPPLR) is >60 %; and the minimum daily economic loss is >1200x106 yuan. In the early stage of mild drought, flash drought can cause the NPPLR to exceed 50 %, and the NPPLR increases significantly with increasing flash drought intensity. Flash drought-induced losses occur primarily in summer. The productivity anomalies, productivity loss and economic loss are greatest in meadow grassland, with the lowest values in desert grassland. Meadow grassland productivity suffers severe losses in almost all flash drought stages. The main stages causing flash drought-induced losses are mild drought in typical grassland and severe and extreme drought in desert grassland. The productivity and economic losses caused by strong evapotranspiration flash drought (SEFD) are greater than those caused by heat wave flash drought (HWFD), and SEFD is more likely to cause productivity loss. The flash drought-induced direct economic loss in grassland is higher than the indirect economic loss and the investment premium loss. In addition to agriculture and animal husbandry, flash drought has the greatest impact on the chemical products industry. The flash drought-induced losses of grassland ecosystems will continue to increase in the future.

Flash droughts have attracted worldwide attention because of their rapid outbreak and extensive influence. However, studies regarding the characteristics and effects of flash droughts in grassland ecosystems are insufficient. In this study, the frequency and intensity characteristics of flash droughts in the Xilinguole Grassland in China were studied. The response characteristics of the productivity of different types of grassland to flash droughts and the relationship between these characteristics and the drought tolerance of grassland were revealed. The results show that (1) flash droughts had the greatest impact on grassland net primary productivity (NPP) and rain use efficiency (RUE) in summer and spring, respectively, with a level of intensity above that of moderate drought. Strong evapotranspiration flash droughts (SEFD) require more attention from decision-makers than heat wave flash droughts (HWFD). A higher frequency and intensity of flash droughts had a greater impact on vegetation. (2) Flash droughts caused moderate negative anomalies in the NPP and RUE indices in more than 90 % of the grasslands. The longest lag time of the NPP response to flash droughts was 2 months, and NPP anomalies were affected by flash droughts for nearly 2 months. RUE was more sensitive to flash droughts than NPP. RUE responded to flash droughts within 10 days, with a decrease of more than 80 % in magnitude, which was 30 % higher than that of the NPP, and the duration of the anomaly was half that of the NPP. (3) Grasslands with a high sensitivity to flash droughts had shorter response durations, fewer abnormalities, better recovery abilities and better drought tolerance. The drought tolerance of grasslands did not increase in association with large NPP and RUE values. Desert grasslands were the most drought tolerant, while meadow grasslands were the least drought tolerant, with the highest risk of flash droughts. This study provides theoretical support for improving the ability of an ecosystem to cope with flash drought risk and scientific grassland management.

Amidst climate change, managing water and vegetation adaptability is vital in ecology and agriculture. Salix psammophila is key in deserts, maintaining ecological balance and combating desertification. Understanding its surface condensation and response to weather is critical for survival. This study aims to investigate the formation patterns of surface condensation water under S. psammophila at different irrigation levels and the influence of precipitation, temperature, relative humidity, and wind speed. Conducted in China’s Kubuqi Desert from June to September 2022, the study employed micro-lysimeters with four irrigation test sets and a control group, conducting detailed observations at various locations. The results indicate that S. psammophila significantly influences surface condensation water by blocking solar radiation and reducing wind speed. The drip irrigation system also regulates surface condensation water on S. psammophila. Moreover, meteorological factors such as 24 h maximum temperature, relative humidity difference, wind speed, and air vapor pressure deficit show significant correlations with surface condensation water formation. In conclusion, this study enhances the understanding of desert ecosystem water balance, vegetation adaptability, and efficient water resource utilization. It provides valuable scientific guidance for the conservation and restoration of desert ecosystems.

Grazing affects grasslands worldwide. However, the global patterns and general mechanisms of how grazing affects plant reproductive traits are poorly understood, especially in the context of different climates and grazing duration. We conducted a meta-analysis of 114 independent grazing studies worldwide that measured plant reproductive traits in grasslands. The results showed that the number of tillers of plant increased under grazing. Grazing did not affect the number of reproductive branches of forbs, but significantly reduced the number of reproductive branches of grasses. Grazing increased the number of vegetative branches of all plants and reduced the proportion of reproductive branches. Grazing significantly reduced the number of flowers in forbs. Seed yield in the two plant functional groups was reduced compared with no-grazing. Under grazing, the sexual reproduction of grasses decreased much more substantially than that of forbs. This may be due to biomass allocation pattern of grasses under grazing (i.e., belowground versus aboveground). Under grazing, plants tended to adopt rapid, low-input asexual reproduction rather than long-term, high-risk sexual reproduction. This study represents the first large-scale evaluation of plant reproductive trait responses under grazing and demonstrates that grazing inhibits sexual reproduction and promotes asexual reproduction. The effect of grazing on plant sexual reproduction was influenced by grazing intensity, mean annual precipitation, and grazing duration. These results will assist in the development of sustainable grazing management strategies to improve the balance between human welfare and grassland ecosystem health.

The goal of this study was to establish a comprehensive growth index (CGI) of grassland vegetation for monitor the overall condition of the grassland. Taking the desert grassland in Otuoke Banner, Ordos City, Inner Mongolia as the research object, this study integrates five indicators. First, the optimal band of the unmanned aerial vehicle hyperspectral data is optimized using the correlation analysis, successive projection algorithm (SPA), optimum index factor method, and band combination index method. A dual-band spectral index in good correlation with the CGI is then constructed in the optimal band. Afterwards, a CGI characterization model is established in accordance with the partial least squares regression (PLSR) algorithm and its accuracy is analyzed. Finally, the CGI of the study area is estimated. The experimental results are as follows. 1) The R 2 of models built using the training samples of the spectral indices corresponding to the optimal spectra screened by the SPA method was 0.7835, RMSE was 0.0712, and RE was 6.89%, less than 10%. The R 2 of the Validation samples was 0.7698, RMSE was 0.0471, and RE was 6.36%, less than 10%, highest precision. 2) Models were built using the spectral indices corresponding to the optimal spectra screened by the SPA method, and the CGI mean values were inverted. A comparison of the mean measured CGI values of the sample quadrat of the test area showed that the mean relative error was 3.82%. The results show that the vegetation growth of desert-steppe grasslands can be adequately monitored, providing technical support for the rapid and accurate diagnosis of grassland conditions. However, there are still shortcomings in this study. 1) The research area for this study was mainly in the desert steppe in Otuoke Banner, Ordos, hence the relevance and universality of the findings need to be verified, and subsequent experiments need to be carried out on desert steppes in other regions or even other types of grasslands to test the universality of the model. 2) In this study, the influence of soil background and litter on the spectral reflectance is not considered in depth. In addition, the influence of sensor observation angle and solar elevation angle on the inversion model demands further investigation efforts.