Urban vegetation is vital for urban environment quality and human health. However, human activities contribute to atmospheric factors like Carbon dioxide (CO2) and particulate matter (PM2.5 and PM10) concentration, impacting urban vegetation growth. Vegetation productivity indicates vegetation growth status, with most current estimates in urban areas based on terrestrial carbon cycle models. This study addresses the significant spatial variation of urban and explores data fusion techniques using Landsat 8 and MODIS to acquire high-resolution remote sensing data. Using the evergreen broadleaf forest in Shenzhen as a case study, we modeled the CO2 and Aerosol Optical Depth (AOD) atmospheric stress factors of Vegetation Photosynthesis Model (VPM) for Tianxinshan (TXS) and Yangmeikeng (YMK) research forests using regression algorithms. The selected simulation accuracy indices were R2=0.51, RMSE=0.11, MAE=0.85 for VPM_urban in urban areas, and R2=0.65, RMSE=1.65, MAE=1.30 for rural areas. After validation, the Person r of the GPP simulated by VPM_urban in the urban area represented by TXS research forest increased from the original 0.62 to 0.71. This demonstrates that using VPM_urban can provide a more accurate assessment of the growth status of evergreen broadleaf forests in urban areas, offering valuable insights for urban planning and management.

Accurate evapotranspiration (ET) estimates are vital in the water-stressed Haihe River Basin (HRB); however, the suitability and error modes of existing ET products remain unclear. This study develops an end-to-end diagnostic framework to reveal error mechanisms, define applicability limits, and link site-level error processes to basin-scale performance patterns. The framework's novelty lies in its multi-pronged attribution strategy designed to disentangle complex error sources. It leverages high-precision lysimeters and an eddy covariance (EC) network to separate and validate evaporation versus transpiration errors. It then integrates a physics-based scenario matrix to diagnose error patterns under distinct hydrothermal stresses, and employs a generalized additive model (GAM) to quantify nonlinear relationships between these errors and environmental drivers. The framework reveals systematic, condition-dependent error patterns: at site scale, products overestimate bare-soil evaporation but underestimate cropland transpiration; at basin scale, land-surface models generally show higher skill in near-natural mountain regions, whereas their skill is reduced in intensively managed, irrigation-dominated plains. Failure to capture the spring irrigation-driven ET peak is one factor contributing to this contrast. These irrigation-induced biases directly affect irrigation scheduling, agricultural water budget, and crop water requirement estimates. Product choice should be context-specific: GLEAM demonstrates a superior ability to capture irrigation signals in managed plains, whereas reanalysis products are more reliable in near-natural mountains. These investigations can help optimize water resource utilization and allocation in agricultural fields.

This study collected air monitoring and flux data from the Tianxinshan (TXS) Urban Research Forest and Yangmeikeng (YMK) Rural Research Forest in Shenzhen in 2020, and used structural equation model (SEM) to explore the effects of carbon dioxide (CO2), particulate matter (PM2.5 and PM10), ozone (O3) and other key environmental factors including light, temperature, and water on Gross Primary Productivity (GPP) and ecosystem respiration (Reco) by standardized path coefficients. Variations in the atmospheric concentrations of CO2, PM2.5, PM10 and O3 between urban and rural vegetation were found to influence vegetation growth to a degree comparable to the impact of critical environmental factors such as temperature and water. Notably, the CO2 fertilization did not stimulate vegetation growth in rural Shenzhen, and high concentrations of CO2 in urban hindered photosynthesis. In contrast, increased PM2.5 and PM10 concentrations had a positive impact on promoting photosynthesis particularly in polluted urban areas. Moderate concentrations of O3 in the urban environment can also enhance vegetation growth. Our findings clearly demonstrate that the polluted atmosphere driven by human activities notably affect the urban vegetation growth. These highlight the crucial need to integrate these factors into the ecosystem carbon cycle modeling and the environmental management.

Groundwater level (GWL) variations in the arid regions of Northwest China are driven by both natural processes and human activities. Identifying causal links between hydrological variables is fundamental to understanding groundwater evolution and conducting dynamic simulations. This study integrates the Mann–Kendall test, Seasonal-Trend decomposition using Loess, and the Peter and Clark Momentum-threshold and Momentary Conditional Independence (PCMCI) causal inference to analyze GWL variation characteristics and causal response processes across seven sub-basins in the Tarim Basin using multi-source remote sensing data. Results show an overall decline in GWL, primarily in the north-central part of the basin, with the Kaidu–Konqi River Basin reaching a maximum rate of 0.51 m/year. The trend components reveal localized depletion alongside broad stability, while seasonal components exhibit three types of temporal shifts in fluctuations. A mismatch exists between the prevalence of environmental influences and their causal strength. Daytime land surface temperature (LSTD), surface runoff (RO), and evapotranspiration (ET) show the highest detection frequencies, yet volumetric soil water in layers 2 (SWVL2) and RO exhibit the largest ranges in strength and drive variations at specific sites. Response times are asymmetric. Negative effects from ET on GWL transmit quickly, while positive recovery is slow. Conversely, positive recharge from volumetric soil water in layer 1 (SWVL1) is faster than its negative lag. At the basin scale, surface processes recharge GWL while mediating indirect influences from other variables. Climate and agricultural irrigation act as direct sinks. Depending on local conditions, three regional patterns emerge: direct climate-driven depletion, obstructed shallow water retention, and indirect compensation from agricultural water use. Causal networks indicate that RO and SWVL1 have the highest centrality and dominate water output, whereas SWVL2 acts as a passive receiver. Pathways from the surface to GWL are also asymmetric. The most frequent path involves step-by-step infiltration along RO → ET → SWVL1 → SWVL2 → GWL. In contrast, the paths with the highest cumulative strength are shorter and faster, specifically RO → ET → GWL and RO → SWVL1 → GWL. The identified pathways and lag parameters provide a direct basis for groundwater dynamic modeling and water resource management in the basin.

In recent years, large-scale macroalgae blooms with two dominant species of Ulva prolifera (U.prolifera) and Sargassum horneri (S.horneri) have occurred frequently and concurrently in the Yellow Sea and East China Sea. Existing remote sensing models for distinguishing between U.prolifera and S.horneri have achieved high accuracy in specific scenarios but exhibited weak generalizability, which stemmed primarily from training data inadequately representing diverse scenarios. We first established a more representative sample set from 2015-2023 GaoFen-1, Wide Field of View imagery. This sample set encompassed diverse scenarios, including clear water, turbid water, and areas affected by thin clouds and sun glint. Through spectral analysis, we found that: 1) influenced by multiple factors, algal spectra exhibited significant variability and approximately 30% of algal pixels even had similar spectral shapes; 2) combining difference spectra (spectral difference between algae and surrounding water) and water turbidity could effectively distinguish these fuzzy pixels. Subsequently, we developed a random forest classification model, with input features encompassing three dimensions: algal spectra, difference spectra and turbidity. The overall accuracies of our model on both validation set and test set were higher than 90%. Compared to existing methods, our model could effectively distinguish spectrally similar algal pixels and improved classification accuracy by over 10%.

Investigating dryland phreatic water assets requires an in-deep understanding of depth to water table (DWT). However, current DWT methods suffer from limited accuracy and demand refinement. Therefor, this study suggests one novel integrated model cascaded by twin, assimilation, and DWT submodels. The twin submodel clones land surface model (LSM) with machine learning method (MLM) to capture LSM uncertainties, then the assimilation submodel develops an eigen-uncertainty-weighted four-dimensional variational assimilation framework to optimize LSM outputs using multiple remote sensing (RS) actual evapotranspirations (AETs), thereby, the DWT submodel proposes one physical mechanism based equation with dynamical parameters constrained by optimized LSM outputs and in situ observations. Its effectiveness is evaluated through five pairs of experiments conducted in the Tarim river basin (TRB), China. Results corroborate that the RMSE, MAPE, MAE, R 2 of its estimated DWTs are improved by 21.9%-36.1%, 52.9%-58.3%, 49.5%-59.7%, 2.6%-9.3%, respectively, compared to those from pure-MLMs using original LSM outputs against 84 validation wells across 15 basins. Additionally, the DWT trend obtained well reflects the temporal variability and spatial heterogeneity of the TRB from 2000 to 2020. Owe to its LSM independence and solid physical mechanism, the integrated model ameliorates DWT estimate through a novel insight into the dynamics between phreatic water and heat by maximizing the advantages of multi-source quality RS AETs and LSM outputs without the adjoint models and running of LSMs.

Accurately assessing the Water Conservation Capacity (WCC) of the Water Conservation Area (WCA) in the Yellow River Basin (YRB) is imperative for regional ecological security, yet it remains challenging because of the intricate interplay among climatic and anthropogenic drivers. This study proposes a novel integrated framework that couples the process-based InVEST model with a data-driven Random Forest (RF) algorithm to evaluate the spatiotemporal dynamics of WCC from 2000 to 2022 using annual water yield and Water Conservation Quantity (WCQ). Results reveal that annual water yield and WCQ range from 80.17 to 218.68 mm and 3.56 to 10.99 mm, and optimal Grade I WCC is predominantly concentrated in the central-southern Yellow River Source Area (YRSA), the Southern Mountain Tributary Area of the Wei River (SMTAWR), and the western Yiluo River Basin (YLRB). RF-based factor importance analysis indicates that climatic factors (precipitation, potential evapotranspiration) and anthropogenic factors (NDVI, population, GDP, flow velocity coefficient) are the primary drivers of WCC, while natural structural factors (soil depth, slope, saturated hydraulic conductivity, plant available water content) exert relatively minor effects. By quantitatively disentangling the relative contributions of climatic, natural structural, and anthropogenic factors to WCC, the proposed InVEST-RF framework advances watershed WCC assessment. Moreover, it provides a transferable methodological tool for ecohydrological evaluations in global watersheds, particularly under the context of changing climate and evolving land use trajectories.

Based on five MOD13Q1 data in 2000,2005,2010,2015,and 2020,combined with digital elevation model（DEM）,this study analyzed the distribution and variation characteristics of vegetation coverage in Qilian Mountains Nature Reserve under different periods and terrain conditions.The results show that:（1）The fractional vegetation coverage（FVC）was calculated by the pixel binary model, and it was found that the average annual FVC of Qilian Mountain Nature Reserve from 2000 to 2020 increased from 58.75%（2000）to the recent 63.25%（2020）,the vegetation recovery condition is good.（2）By calculating the transition matrix and state index, it is found that the composition and structure of FVC in Qilian Mountains Nature Reserve has changed from 2000 to 2020,and the overall situation tends to be stable and improved.The improved area of low grade FVC（FVC＜20%）is 1960.439 km~2;the improved area of lower grade FVC（20%＜FVC＜40%）is 1 476.438 km~2;the improved area of higher grade FVC（40%＜FVC＜65%）is 1 816.625 km~2;the area of high grade FVC（65%＜FVC＜100%）is degraded 378.5 km~2.（3）The vegetation restoration status of Qilian Mountain Nature Reserve is different under the conditions of different altitudes, slopes and slope aspects.The vegetation coverage area increases significantly in areas with altitudes of 2500-4000 m, slopes of 5°-25° and semi-shady slopes.The research results can provide ideas and basis for vegetation restoration under different terrain conditions in Qilian Mountains Nature Reserve.

Root zone soil moisture (RZSM) has a direct impact on ecosystem function, vegetation growth and food security, and plays a vital role in global climate system, water and carbon cycles. However, large variations and uncertainties still exist in RZSM across the globe under the warming climate. In this study, we applied comparison map profile (CMP), Theil-Sen regression and partial correlation analysis to investigate the spatial and temporal changes of RZMS and its driving factors from 1981 to 2017 by using three soil moisture products-ERA5, GLDAS and MERRA-2. Results showed that RZSM derived from three products presented a similar spatial pattern that the highest RZSM values occurred in tropical forest and cold areas, followed by subtropical, while the relatively low RZSM values were observed in arid and semiarid regions. Globally, RZSM decreased in all of three datasets with a rate of -0.14 x 10(-3) m(3) m(-3) yr(-1) on average (p < 0.001), which was largely correlated with temperature anomalies. Spatially, the RZSM trends greatly varied, with 21-31% of global land areas experiencing a significant decreasing trend and 7-24% for an increasing trend, respectively, confirming their different sensitivities to climate change. Temperature-driven RZSM dominated 19-29% of global land areas and was primarily distributed in northern high-latitude areas. The areas dominated by evapotranspiration were mainly in arid and semiarid areas, accounting for 29-44% of global land areas. Precipitation dominates the remaining 36-45% of global land areas mainly in eastern America and Europe, suggesting variations in the dominance of environmental factors on the spatial patterns of RZSM trend. Our findings will deepen our understanding of the impacts of climate change on the long-term trend of global soil moisture, and will be greatly critical to global soil water resource protection and management under the warming climate.

Understanding the sensitivity of vegetation growth and greenness to vegetation water content change is crucial for elucidating the mechanism of terrestrial ecosystems response to water availability change caused by climate change. Nevertheless, we still have limited knowledge of such aspects in urban in different climatic contexts under the influence of human activities. In this study, we employed Google Earth Engine (GEE), remote sensing satellite imagery, meteorological data, and Vegetation Photosynthesis Model (VPM) to explore the spatiotemporal pattern of vegetation growth and greenness sensitivity to vegetation water content in three megacities (Beijing, Shanghai, and Guangzhou) located in eastern China from 2001 to 2020. We found a significant increase (slope > 0, p < 0.05) in the sensitivity of urban vegetation growth and greenness to vegetation water content (S-LSWI). This indicates the increasing dependence of urban vegetation ecosystems on vegetation water resources. Moreover, evident spatial heterogeneity was observed in both S-LSWI and the trends of S-LSWI, and spatial heterogeneity in S-LSWI and the trends of S-LSWI was also present among identical vegetation types within the same city. Additionally, both SLSWI of vegetation growth and greenness and the trend of S-LSWI showed obvious spatial distribution differences (e.g., standard deviations of trends in S-LSWI of open evergreen needle-leaved forest of GPP is 14.36 x 10(-2) and standard deviations of trends in S-LSWI of open evergreen needle-leaved forest of EVI is 10.16 x 10(-2)), closely associated with factors such as vegetation type, climatic conditions, and anthropogenic influences.

As remarkable human-induced temperature anomalies on the land surface, variations of urban heat island (UHI) and its driving factors have been investigated in numerous studies. However, few studies discussed the spatiotemporal heterogeneity of the driving forces exerted by land surface energy fluxes, i.e., net radiation, sensible heat, latent heat and heat storage, on UHI behaviors at large scale and long term. In this study, a comprehensive application of multisource datasets and statistical methods have been implemented based on land surface energy balance theory, the spatiotemporal variations of surface UHI intensity (urban-rural temperature difference) and changes of their driving forces have been quantified. The results demonstrate the dynamics of UHI intensity in 32 major cities of China from 2003 to 2017 are generally coherent with the common perception, the overall surface UHI intensity is 4.57 K higher in summer than in winter. The spatial variations of the fluxes that alter UHI intensity can be largely attributed to the varied energy interactions between vegetated/paved surface and atmosphere and the differences of background temperature and precipitation, the contribution of latent heat to UHI changes declines nearly 40% from semiarid/arid climate at the north to subtropical humid climate at the south, while the contributions of other fluxes are stable. The temporal changes of the effect of these fluxes, however, imply more complex mechanisms. The contributions of sensible heat and latent heat to UHI intensity variations are three times and eight times larger in the warm season than in the cold season respectively, indicating the influence of seasonality of background temperature, precipitation and vegetation. The low contributions of these fluxes in the cold season also suggest the significant effect of other driving forces such as anthropogenic heat, especially in semiarid/semihumid climate zones. This study highlights the temporal shifts of major driving forces of UHI intensity, the mitigation tactics for UHI in different cities and seasons should be customized for better validity.

Satellite-derived land surface temperature (LST) is critical for retrieving terrestrial evapotranspiration (ET); however, its availability is limited by low spatial resolution and inclement weather conditions. This study develops a spatiotemporal regression strategy that can downscale 1-km Moderate Resolution Imaging Spectroradiometer (MODIS) LST product to 250-m resolution and simultaneously gap-fill the missing values. The proposed methodology synergistically uses random forest (RF) model and geographically weighted regression, which are, respectively, available for demonstrating the nonlinear correlation between LST and explanatory variables and for calibrating the RF-derived residuals. The study is conducted across a region of ~1.49 million square kilometers in northern China. The coupled model creates a 250-m spatial resolution LST product with the root-mean-square error (RMSE) of 2.32 and 1.87 K when compared with field observations and reference Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) LST, respectively. Meanwhile, it minimizes the constraint of LST availability due to inclement weather conditions with RMSE of 2.69 and 2.31 K relative to field observations and reference images, respectively. The results further reveal that remote-sensing-derived ET using the 250-m downscaled LST data is fairly accurate with the relative errors of 6%–9% as evaluated with flux measurements. The 250-m modeled ET retrievals exhibit a more intense hydrological response to the water use conditions compared with the 1-km remotely sensed ETs and Noah land surface model ETs. This study may benefit land surface hydrology research and water resource management.

Marine floating raft aquaculture (FRA) monitoring is significant for marine ecological environment and food security assessment. Synthetic aperture radar (SAR) based monitoring is considered to be the effective way for FRA identification because of its capability for the all-weather application. Considering the poor generalization ability and extraction accuracy of traditional monitoring methods, we proposed a semantic segmentation model called D-ResUnet to extract FRA areas from Sentinel-1 images. It has the U-Net-like structure but combines the pre-trained ResNet34 as the encoder and dense residual units in the decoder. The experiments showed the effectiveness and superiority of FRA extraction based on the proposed D-ResUnet compared with the other three state-of-the-art semantic segmentation models.

As ???the third pole of the world???, the land surface temperature (LST) of the Qinghai-Tibet Plateau (QTP) has a profound impact on the climate of central Asia and even the whole earth. Studying the impact of the LST over QTP depends on long time and high spatiotemporal resolution LST dataset. However, the unavailability of such dataset has hindered LST-related researches: one of the most important reasons is that traditional spatiotemporal fusion methods such as Spatial and Temporal Adaptive Reflectance Fusion Model (STARFM) have heavy computation demands to process big data. To fill this gap, this paper outlines one new cloud spatiotemporal fusion method by combining Google Earth Engine and STARFM to develop for the first time a 3-hourly 30 m LST dataset over the QTP from 2000 to 2020 through fusing Landsat and GLDAS-2.1 derived LST data. The outlined method first fuses the LSTs obtained from Landsat and GLDAS-2.1 data within one year to synthesize the LST of the entire QTP on one base time, and then the base time LST is spatiotemporally fused with GLDAS-2.1 LSTs on the base and prediction times to derive the 3-hourly 30 m QTP???s LSTs on prediction times. The outlined method provides a promising technical scheme for batch processing big data in combining traditional spatiotemporal fusion methods with cloud computing platforms. Derived LST dataset is validated by station observations at multiple time and spatial scales to have high accuracy, which provides a guarantee for analyzing water and heat exchange and climate change over the QTP.

Soil moisture (SM) is a crucial component for understanding, modeling, and forecasting terrestrial water cycles and energy budgets. However, estimating field-scale SM based on thermal infrared remote-sensing data is still a challenging task. In this study, an improved Flexible Spatiotemporal DAta Fusion (FSDAF) method based on land-surface Diurnal Temperature Cycle (DTC) model (DFSDAF) was proposed to fuse Moderate Resolution Imaging Spectroradiometer (MODIS) and Advance Spaceborne Thermal Emission and Reflection Radiometer (ASTER) land-surface temperature (LST) data to generate ASTER-like LST during the night. The reconstructed diurnal LST data at a high spatial resolution (90 m) was then utilized to drive a two-source normalized soil thermal inertia model (TNSTI) for the vegetated surfaces to estimate field-scale SM. The results of the proposed methods were validated at different observation depths (2, 4, 10, 20, 40, 60, and 100 cm) over the Zhangye oasis in the middle region of the Heihe River basin in the northwest of China and were compared with the SM estimates from the TNSTI model and other SM products, including AMSR2/AMSR-E, GLDAS-Noah, and ERA5-land. The results showed the following: (1) The DFSDAF method increased the accuracy of LST prediction, with the determination coefficient (R2) increasing from 0.71 to 0.77, and root mean square error (RMSE) decreasing from 2.17 to 1.89 K. (2) the estimated SMs had the best correlation with the observations at the 10 cm depth (with R2 of 0.657; RMSE of 0.069 m3/m3), but the worst correlation with observations at the 40 cm depth (with R2 of 0.262; RMSE of 0.092 m3/m3); meanwhile, the modeled SMs were significantly underestimated above 40 cm (2, 4, 10, and 20 cm) and slightly overestimated below 40 cm (60 and 100 cm); in addition, the field-scale SM series at high spatial resolution (90 m) showed significant spatiotemporal variation. (3) The SM estimates based on the TNSTI for the vegetated surfaces are more capable of characterizing the SM status in the root zone (~80 cm) or even deeper, while the SMs from AMSR2/AMSR-E, GLDAS-Noah, or ERA5-land products are closer to the SM in the surface layer (the depth is less than 5 cm). The TNSTI provided favorable data supports for hydrological model simulations and showed potential advantages for agricultural refinement managements and smart agriculture.

Under the background of global warming, understanding the dynamic of vegetation plays a key role in revealing the structure and function of an ecosystem. Assessing the impact of climate change and human activities on vegetation dynamics is crucial for policy formulation and ecological protection. Based on the Global Inventory Monitoring and Modeling System (GIMMS) third generation of Normalized Difference Vegetation Index (NDVI3g), meteorological data and land cover data, this study analyzed the linear and nonlinear trends of vegetation in northern China from 1982 to 2015, and quantified the relative impact of climate change and human activities on vegetation change. The results showed that more than 53% of the vegetation had changed significantly, and 36.64% of the vegetation had a reverse trend. There were potential risks of vegetation degradation in the southwestern, northwestern and northeastern parts of the study’s area. The linear analysis method cannot disclose the reversal of the vegetation growth trend, which will underestimate or overestimate the risk of vegetation degradation or restoration. Climate change and human activities promoted 76.54% of the vegetation growth in the study area, with an average contribution rate of 51.22% and 48.78%, respectively, while the average contribution rate to the vegetation degradation area was 47.43% and 52.57%, respectively. Vegetation restoration of grassland and woodland was mainly affected by climate change, and human activities dominated their degradation, while cropland vegetation was opposite. The contribution rate of human activities to vegetation change in the southeastern and eastern parts of the study area was generally higher than that of climate change, but it was the opposite in the high altitude area, with obvious spatial heterogeneity. These results are helpful to understand the dynamic mechanism of vegetation in northern China, and provide a scientific basis for vegetation restoration and protection of regional ecosystems.

Abstract. Soil carbon isotopes (δ13C) provide reliable insights at a long-term scale for studying soil carbon turnover. The Tibetan Plateau (TP), called “the third pole of the earth” is one of the most sensitive areas to global climate change and exhibits an early warning signal of global warming. Although many studies detected the variability of soil δ13C at site scales, a knowledge gap still exists in the spatial pattern of topsoil δ13C across the TP. To fill the substantial knowledge gap, we first compiled a database of topsoil δ13C with 396 observations from published literatures. Then we applied a Random Forest (RF) algorithm – a machine learning approach, to predict the spatial pattern of topsoil δ13C and β (indicating the decomposition rate of soil organic carbon (SOC), calculated by δ13C divided by logarithmically converted SOC). Finally, two datasets – topsoil δ13C and β with a fine spatial resolution of 1 km across the TP were developed. Results showed that topsoil δ13C varied significantly among different ecosystem types (p < 0.001). Topsoil δ13C was −26.3 ± 1.60 ‰ (mean ± standard deviation) for forests, 24.3 ± 2.00 ‰ for shrublands, −23.9 ± 1.84 ‰ for grasslands, −18.9 ± 2.37 ‰ for deserts, respectively. RF could well predict the spatial variability of topsoil δ13C with a model efficiency of 0.62 and root mean square error of 1.12 ‰, enabling to derive data-driven δ13C and β products. Data-driven topsoil δ13C varied from −28.26 ‰ to −16.95 ‰, with the highest topsoil δ13C in the north and northwest TP and the lowest δ13C in Southeast or South TP, indicating strong spatial variabilities in topsoil δ13C. Similarly, there were strong spatial variabilities in data-driven β, with the lowest β values at the east and middle TP, indicating a higher SOC turnover in the east and middle TP compared that of other regions in the TP. This study was the first attempt to develop a fine resolution product of topsoil δ13C and β across the TP, which could provide an independent data-driven benchmark for biogeochemical cycling models to study SOC turnover and terrestrial carbon-climate feedbacks over the TP under climate change. The data-driven δ13C and β datasets are public available at https://doi.org/10.6084/m9.figshare.16641292.v2 (Tang, 2021).

"一带一路"沿线地区水资源短缺且空间分布不均衡,虚拟水贸易实现了对水资源的远距离空间调配.以2010-2018年"一带一路"沿线59个国家和37种农作物为研究对象核算各国农作物虚拟水贸易,利用标准差椭圆、Moran's I指数、LISA指数刻画农作物虚拟水贸易的时空格局特征,通过地理探测器和地理加权回归模型分析农作物虚拟水贸易的驱动因素及其空间异质性.研究发现:①沿线各国"低耗水-高出口型"作物占比4.013％,"高耗水-高出口型"作物占比1.926％.②沿线各国农作物虚拟水贸易的进口格局呈收缩趋势,出口格局呈扩张趋势,且其局部存在一定的集聚特征,进口的"高-高集聚"区域主要处于南亚地区,出口的"高-高集聚"区域主要分布在中东欧地区.③各显著驱动因素能较好地解释沿线各国农作物虚拟水净出口量,各因素中国内生产总值呈负相关驱动特征,耕地面积呈正相关驱动特征,而人口规模、森林面积和邻国接边数在各单元间呈正负两极的差异驱动特征.