Search

• We provide a fine-scale daily quantification of the time-accumulation and -lag effects of precipitation on carbon fluxes. • Time-accumulation and -lag effects significantly amplify the combined effects of precipitation on carbon fluxes. • The synthesis of decadal site observations with broad regional scale analyses reveals a time-delayed response of carbon fluxes to precipitation. Precipitation (PRE) plays a vital role in hydrological processes, ecological vegetation, and land-atmosphere interactions in semi-arid regions. Previous research has mainly focused on the impact of PRE on large-scale regional climate change and ecological evolution. However, there have been few studies on the long-term effects of PRE on carbon fluxes in these regions, especially the time-accumulation and -lag effects. Here, we employed observational data from the Semi-Arid Climate and Environment Observatory of Lanzhou University (SACOL) and integrated multiple data sources, including remote sensing and carbon flux simulation data, to quantitatively assess the lagged response of carbon fluxes to PRE and elucidate the underlying mechanisms from multiple perspectives. Characterization of PRE, soil water content (SWC) and carbon fluxes at SACOL qualitatively reveals the existence of a time-delayed response of carbon fluxes to PRE, both on monthly and finer daily temporal scales. The average lagged response of net ecosystem exchange (NEE) and gross primary productivity (GPP) to accumulated PRE (APRE) is approximately 42 days. When considering time-accumulation and -lag effects, the combined direct and indirect effects of APRE on NEE and GPP increase by 0.37 and 0.58, respectively. Notably, preceding APRE primarily exerts a direct effect on current carbon fluxes, whereas the impact of SWC at a depth of 0.1 m is primarily mediated through the memory effect of preceding APRE, resulting in an indirect effect on carbon fluxes. These findings emphasize the importance of preceding APRE. Significantly, our subsequent study indicates that the delay in NEE and GPP responses to APRE also extends to approximately 40 to 50 days at the regional scale. Our findings emphasize the significant time effects of APRE on carbon fluxes, and considering these effects will contribute to a better understanding of the interplay between PRE and vegetation over semi-arid regions in China. [ABSTRACT FROM AUTHOR]

• Low SM and high VPD on GPP showed a continuous increase in drought stress. • Compared to the effect of VPD on GPP, SM is the main drought driver in 60 % of areas, affecting 10 % more of GPP. • The intensity of the impact of SM on GPP is rapidly increasing and the dominant region is expanding. • The effects of SM and VPD on GPP exhibited significant variations across different vegetation and climate gradients. Decreasing soil moisture (SM) and increasing vapor pressure deficit (VPD) are the main drought affecting factors of terrestrial vegetation productivity. Nevertheless, the impact of continued warming on the changing trend of SM and VPD constraints affecting vegetation productivity remains uncertain. Understanding the complex interactive effects of SM and VPD on vegetation is crucial for assessing drought risk and its ecological implications. This study aims to comprehensively quantify the trends in the respective contributions of SM and VPD to vegetation ecosystem productivity in China from 2000 to 2022. The results showed that both low SM and high VPD had significant inhibitory effects on gross primary productivity (GPP) anomaly. Furthermore, the magnitude of water stress on vegetation productivity increases, and is accompanied by a gradual expansion of areas experiencing significant water deficit during the study period. By decoupling the interactions between SM and VPD, we found that SM and VPD were the primary influencing factor of water stress on vegetation productivity, accounting for 60 % and 40 % of the total vegetated area, respectively. Notably, the effects of SM and VPD on GPP exhibited significant variations across different vegetation and climate gradients. The amount of land dominated by the SM constraint expanded, and SM gradually played an increasingly important role in the water stress of GPP over the past two decades. The results of the study are important to accurately assess the interrelationship between vegetation and climate in the context of climate change. [Display omitted] [ABSTRACT FROM AUTHOR]

• Structural traits showed hump (or reverse) patterns along the elevational gradient. • Macroclimate and elevation contributed to 90 % of the microclimate variability. • Vegetation displayed a stronger buffering effect on microclimate at day than night. • Soil moisture presented opposite effects on microclimate between day and night. • Lapse rates showed a seasonal fluctuation at daytime but were stable at nighttime. Tropical mountain forests are important biodiversity hotspots, which host disproportionally high number of endemic species. However, the potential impacts of climate change in these areas are uncertain. A key factor contributing to this knowledge gap is that climatic conditions experienced by organisms inside tropical forests (i.e., microclimate) remain largely understudied. Due to the effects of topography and vegetation, the understory microclimate can differ substantially from free-air conditions (i.e., macroclimate). This study aimed at unveiling vegetation structural characteristics and microclimatic patterns along an elevational gradient in a highly diverse tropical mountain ecosystem (Mount Kenya), by combining hundreds of terrestrial laser scanning measurements with a two-year time-series of microclimate observations. Our results showed that macroclimate temperature and elevation contributed >90 % to the microclimate variability in our study area. The influence of vegetation and soil moisture in regulating temperature differed substantially between day and night, as well as in different periods of the year. The contribution of vegetation to microclimate variation during the day was two times higher than that during the night. Soil moisture had a cooling effect on microclimate temperature during daytime, while the opposite pattern was observed at night. These differences affected lapse rates, which showed a clearly seasonal fluctuation during diurnal periods but a relatively stable pattern in night periods. The diurnal temperature range was regulated by combined effects of vegetation structure, elevation, and soil moisture. Finally, we were able to detect subtle changes in forest structure caused by historical selective logging. These changes resulted in a legacy effect on microclimate, thus demonstrating that human-induced disturbances have long-lasting impacts on tropical mountain ecosystems. These results improve our understanding of the climatic conditions experienced by organisms inside highly diverse African mountain forests, as well as how these conditions are regulated by vegetation structure and environmental factors. [ABSTRACT FROM AUTHOR]

• Seasonal and inter-annual variations of NEE were examined. • The daily NEE shifts between PAR-dependent and NDVI-dependent with changes in SWC during the growing seasons. • SWC mainly controlled the inter-annual variability in annual NEE. Alpine ecosystems may be contribute to increasing atmospheric carbon dioxide (CO 2) concentrations under intensified climate change on the Tibetan Plateau (TP). The eddy covariance technique was used in this research to examine how net ecosystem CO 2 exchange (NEE) and its component fluxes, gross primary production (GPP), and ecosystem respiration (ER) varied over 10 years (from 2010 to 2019) in an alpine meadow on the eastern TP. This alpine meadow acted as a sink of CO 2 , with a mean of –92.5 ± 54.9 g C m–2 year–1 during the measurement period. The exponential ER response model is unsuitable for water stress conditions, potentially resulting in an overestimation of ER and GPP. The intra-annual variations of daily GPP and ER were more related to photosynthetic active radiation (PAR), soil temperature (Ts) and normalized difference vegetation index (NDVI), but the daily NEE during growing seasons shifts between PAR-dependent and NDVI-dependent with changes in soil water content (SWC). At monthly and seasonal scales, the sensitivity of GPP to NDVI and water conditions was higher than that of ER. The inter-annual variability in NEE was significantly related to the SWC. These findings are critical for understanding the response mechanism of carbon dynamics to the changing biotic and abiotic conditions in the TP. [ABSTRACT FROM AUTHOR]

• In-situ measurements of eddy covariance are used to examine the Bowen ratio over the Tibetan Plateau. • The Bowen ratio is found to be sensitive to soil moisture in arid and semiarid regions. • Vapor pressure deficit is the dominant factor influencing the Bowen ratio in arid and subhumid regions. Turbulent heat fluxes across the surface are an important mechanism of land-atmosphere coupling. But there is still a lack of sufficient observational measurements, particularly over the climate sensitive Tibetan Plateau (TP). This paper examines the partitioning between sensible and latent heat fluxes during growing season using the Bowen ratio as a diagnostic based on eddy covariance measurements from 12 observational sites located on the TP with the altitudes ranging between 3327 m and 4509 m above sea level. The results show an average Bowen ratio of 1.13, indicating that sensible heat flux is only slightly dominant in surface energy balance. For different climate zones, Bowen ratio varies from 0.56 to 1.05 in the semi-arid zone, 0.53 and 0.57 in the subhumid region, and 2.73 to 3.11 in the arid zone. The Bowen ratio shows sensitivity to soil dry and wet condition, with higher values during drier soil conditions. In the arid to semi-arid regions, the Bowen ratio shows a clear positive correlation with the Drought Soil Index (DSI) and a negative correlation with the Normalized Difference Vegetation Index (NDVI), suggesting sensitivity to soil moisture conditions. In the sub-humid climate zone, vapor pressure deficit (VPD) is the dominant factor influencing the Bowen ratio. In the wet and dry transition zone, soil moisture, VPD, NDVI and land-air temperature difference all have a role to play. [ABSTRACT FROM AUTHOR]

• SIF performs well in tracking GPP during dry conditions in a semi-arid grassland. • SIF and GPP exhibit parallel responses to environmental variables under dry conditions. • Compared to wet conditions, the role of physiological information regulation in SIF and GPP increases under dry conditions. Solar-induced chlorophyll fluorescence (SIF) has recently emerged as a promising tool for estimating gross primary production (GPP). To date, there is ongoing debate regarding whether the strong correlations between SIF and GPP persist under dry conditions. Here, we conducted continuous far-red SIF measurements in a semi-arid grassland from 2017 to 2019 to investigate its association with GPP. Our findings revealed strong correlations in the seasonal patterns of SIF and GPP during dry conditions (R2=0.79). After disentangling the effects of photosynthetically active radiation and soil water content, we observed consistent responses to environmental variables in both SIF and GPP under dry conditions. Furthermore, we conducted a dominance analysis to assess the contributions of physiological and non-physiological components to the variations in SIF and GPP. Our results demonstrated a substantial increase in the contributions of physiological components to both SIF (wet: 19.29%; dry: 60.23%) and GPP (wet: 20.89%; dry: 28.38%) during dry conditions, highlighting a shift towards enhanced physiological regulation of SIF and GPP in response to dry conditions. In conclusion, our findings provide valuable insights into the GPP-SIF relationships in semi-arid grasslands under dry conditions. These insights hold the potential to refine and constrain model predictions under climate change. [ABSTRACT FROM AUTHOR]

• This study examined characteristics of NEP and CUE at different ecosystems in the alpine region. • Qilian Mountains, as a climate-sensitive region, acted as an effective carbon sink. • Wetland/swamp had the highest NEP, followed by forest, while shrub had the lowest NEP in the Qilian Mountains. Warming leads to significant loss of CO 2 in high-altitude regions (HAR), posing threat to the carbon sink of terrestrial ecosystem. Additionally, the spatial distribution of environmental factors and underlying surfaces also determine the carbon sink pattern. Therefore, it is necessary to systematically explore the carbon sink of HAR. Based on it, choosing the Qilian Mountains (QLM) as the study area, the continuous observation data of 14 eddy covariance in different ecosystems was used to analyze the variation characteristics of carbon use efficiency (CUE) and net ecosystem primary productivity (NEP), which is helpful to systematically understand the response of carbon cycle to climate change in alpine ecosystem. The research results indicated that the QLM serves as an effective carbon sink (13 of the sites yielded a net carbon sink), owing to the combined influences of environmental factors and vegetation characteristics. Annual NEP varied across the 14 sites, ranging from -192.6 to 524.5 g C/m2/yr. Limited observation indicated that wetland/swamp had the highest carbon sink, followed by forest, and shrub have the lowest carbon sink in this study. Along the altitudinal gradient, both gross primary productivity (GPP) and ecosystem respiration (Re) demonstrated a declining trend (P < 0.05), while, CUE displayed an increasing trend. Soil temperature and photosynthetically active radiation dominated the variation in carbon exchange and CUE along the altitudinal gradient. However, soil moisture was the dominant factor in drought ecosystem. This study provides basis for the assessment of carbon sink of the HAR. [ABSTRACT FROM AUTHOR]

• A single-point sensor cannot evaluate soil moisture in a heterogeneous environment. • TSA provides a better understanding of sensor network distribution. • The CRNP offers a reliable solution to capture SSM in challenging environments. • CRNP provides a reference for SM product evaluation in forested mountain terrain. Soil moisture is critical to the Earth's water cycle as it influences water exchange between the land, atmosphere, and oceans. This study assessed the potential of a cosmic-ray neutron probe (CRNP) to monitor soil moisture dynamics in a forested mountain region, which is critical for water resource management and ecosystem health. Accurate soil moisture monitoring is essential for effectively managing forested mountainous areas, but various limitations exist in such environments. To overcome the complexities of soil moisture monitoring, an observatory consisting of a CRNP, automatic weather sensors, and dataloggers was established. Fifteen capacitance sensor (CS) stations were placed strategically within the CRNP's footprint to ensure strict evaluation. Although each CS measurement revealed the heterogeneity of the experimental site, temporal stability analysis determined the representative soil moisture. The validation of the CRNP and the average results of the 15 surface CS sensors revealed that they were strongly correlated (R = 0.94). It means that the CRNP can describe the dynamics of soil moisture that fluctuate according to physical properties, underscoring its reliability in estimating intermediate scale soil moisture. This correlation remained robust despite differences in rainfall patterns over three years. These findings indicate that the CRNP may be a valuable tool for monitoring soil moisture in a challenging environment. Additionally, the evaluation of the Advanced Microwave Scanning Radiometer-2 (AMSR2), Advanced Scatterometer (ASCAT), and Global Land Data Assimilation System (GLDAS) demonstrated the usefulness of CRNP in supplying reference data. The AMSR2 showed an inadequate temporal pattern (R = 0.35), and the ASCAT achieved the highest correlation agreement (R = 0.67) with the CRNP. The GLDAS also showed close agreement, although with some discrepancies (R = 0.52). The results of this study support the utilization of CRNP and provide insight into establishing and expanding a soil moisture network. [ABSTRACT FROM AUTHOR]

• SM and VPD were coupled due to land-atmospheric interactions. • VPD exert more extensive and intense impacts on vegetation production than SM. • In China, high VPD dominated over high SM in promoting vegetation productivity. Vegetation production is critical for carbon sequestration in terrestrial ecosystems. However, the relative contribution of soil moisture (SM) and vapor pressure deficit (VPD) to vegetation production in ecosystems remains debated. This study combined SM and VPD data with satellite observations of solar-induced fluorescence (SIF) from 2003 to 2018 to examine the contribution of SM and VPD to long-term variations in vegetation production using binning approach. Constraint lines based on restricted cubic splines were established to investigate the threshold for the responses of vegetation production potential to VPD and SM. The results revealed that VPD had greater impacts on vegetation production in Chinese terrestrial ecosystems than SM, with VPD-dominated and SM-dominated areas accounting for approximately 84 % and 16 % of the total national area, respectively. The mean values of SIF changes under the independent influence of SM and VPD were 0.034 mWm−2nm−1sr−1 and 0.130 mWm−2nm−1sr−1, respectively, indicating higher sensitivity of SIF to VPD. The effects of SM and VPD on ecosystem production varied across aridity gradients, with VPD making a significant contribution to forest ecosystem production. However, determining the relative effects of SM and VPD in ecosystems with low vegetation cover was difficult. Further exploration revealed that the maximum potential for terrestrial ecosystem production increased with VPD, peaking at a VPD of about 1.275 ± 0.068 kPa. Beyond this range, the maximum production potential declined. Our study confirms that VPD has a stronger and more extensive influence on the trend and variability of ecosystem production in China than SM. [Display omitted] [ABSTRACT FROM AUTHOR]

• Tree water uptake does not necessarily correspond to soil water content. • Deferred wet season onset by two months increases tree's groundwater uptake. • 25% increase in wet season precipitation reduces tree's shallow soil water uptake. • Trees adapt to precipitation changes through water uptake and stomatal regulation. Gaining insights into the impacts of altered precipitation patterns on tree species' water uptake sources and water usage strategies is crucial for water resource management, but a comprehensive understanding is lacking. We explored water uptake patterns and water-related traits of four dominant tree species (Evergreen: Carallia brachiate, Symplocos poilanei, Schefflera heptaphylla ; Deciduous: Tetradium glabrifolium) in two altered seasonal precipitation patterns (deferred wet season [DW] and wetter wet season [WW]) and a control (CK) in a secondary tropical forest. Using stable isotopes (δ2H and δ18O) and MixSIAR models, we determined water uptake sources of the target species from a 60 cm soil profile and groundwater. Our results showed that the DW treatment increased (2.3% to 11.5%) groundwater utilization by trees in June and November compared to the CK, and promoted greater water uptake by trees from 0 to 40 cm soil layer in March. However, the WW treatment had a mild impact on water utilization in plants, leading to a reduction in water uptake by 1% - 3.8% from the 0–40 cm soil layer for the four tree species in September. The four studied tree species primarily rely on soil water, with the deciduous T. glabrifolium exhibited more flexible water uptake strategies compared to the other three evergreen species. For plant traits, S. heptaphylla exhibited larger stomatal size under DW treatment, paralleling the elongated stomatal in C. brachiate under the DW treatment. Conversely, S. poilanei and T. glabrifolium exhibit greater flexibility in balancing midday water potential. Under altered precipitation patterns, these plants adopt different morphological and/or physiological strategies, along with shift their water uptake sources to adapt to fluctuating water conditions. Overall, this study improves our understanding of plant water use in secondary tropical forests and emphasizes the importance of considering species-specific responses in forest management under potential precipitation pattern changes. [ABSTRACT FROM AUTHOR]

• The Dajiuhu peatland contributed to mitigating global warming in both the short and long term. • Different regulatory pathways of environmental factors on carbon fluxes in the growing and non-growing season. • Vascular plants could play a crucial role in the regulation of carbon fluxes. Peatlands present substantial terrestrial carbon reservoirs actively engaged in land-atmosphere carbon exchanges. Ongoing variations in carbon dynamics, attributed to global climate change, underscore the necessity of elucidating the connections between carbon fluxes and meteorological parameters. Limited studies have delved into disparities in carbon fluxes and the mechanisms mediating them during growing and non-growing seasons on a long-term scale, particularly in subtropical peatlands. Consequently, this study utilized the eddy covariance technique in the Dajiuhu peatland to investigate five-year CH 4 and CO 2 fluxes, longside meteorological characteristics. The Dajiuhu peatland functioned as a CO 2 sink and CH 4 source, ultimately serving as a carbon sink. A substantial portion of the annual fluxes was measured during the non-growing season, with elevated CO 2 uptake and CH 4 emission during the growing season. Further analyses revealed that photosynthetically active radiation (PAR), soil temperature (Ts), air temperature (Ta), and soil water content (SWC) were critical environmental parameters influencing fluctuations in CO 2 and CH 4 fluxes. Two distinct controlling pathways for CO 2 fluxes were identified. In the growing season, PAR emerged as the most influential factor in CO 2 flux variations, and increasing Ts weakened CO 2 absorption by decreasing SWC. During the non-growing season, Ts emerged as the most influential factor in CO 2 exchanges, with lower Ts leading to decreased CO 2 uptake. Additionally, CH 4 flux variations were significantly regulated by PAR and Ts in both periods. The results illustrated that vascular plants played a pivotal role in the dynamics of carbon between peatlands and the atmosphere. Furthermore, it was discovered that the Dajiuhu peatland contributed to mitigating global warming in both the short and long term, with SGWP identified as a more reliable evaluation index. Our findings underscore the essential ecological function of the Dajiuhu subtropical peatland and provide a theoretical foundation for scientifically managing carbon in natural subtropical peatlands. [ABSTRACT FROM AUTHOR]

• Spring phenology (SOS) advanced in China's forest with north-south disparity. • Soil temperature is the main driver of SOS variability for China's forest. • Coupling temperature and soil moisture effects on SOS were simulated by SEM. Changes in forest phenology regulate ecosystem structure and function and have a feedback effect on climate, however, the primary factors controlling forest spring phenology remain unclear. Here, we used long-term satellite records to monitor forest spring phenology (start of the growing season, SOS) trends in China from 2000 to 2018 and quantified the coupling effect of preseason hydrothermal factors on SOS by using a structural equation model (SEM). Our results showed that solar-induced chlorophyll fluorescence (SIF) retrieved SOS (SOS SIF) and enhanced vegetation index (EVI) retrieved SOS (SOS EVI) significantly (p < 0.05) advanced by 3.7 ± 0.1 d·decade–1 and 3.2 ± 0.1 d·decade–1, respectively, from 2000 to 2018. Regions with high advance trends were primarily distributed in low-latitude regions, whereas low advance trends were observed in the high-latitude regions. Partial correlation results showed that soil temperature (ST) dominates SOS variation in approximately half of the forest areas, and sensitivity analysis indicated regions with higher mean ST and maximum temperature (T max) tended to have earlier SOS but the effects of minimum temperature (T min) were reversed. The SEM further demonstrated that ST is the dominant factor in SOS variability, moreover, ST not only directly affects SOS, but also indirectly affects the SOS by influencing the soil moisture. These findings highlight the key role of ST in regulating changes in forest spring phenology in China, which is absent in most phenology models but should be considered when predicting forest phenology and associated carbon budgets. [ABSTRACT FROM AUTHOR]

• The temporal scale of the original Biome-BGC model was refined from a daily scale to an hourly scale. • A theoretical interception module was incorporated into the hourly-scale Biome-BGC model for correcting canopy interception, to further enhance the accuracy of carbon and water fluxes simulations in subtropical coniferous forest. • In comparison with the original daily-scale model, simulations from the hourly-scale model showed better agreement with measurements at the Qianyanzhou forest site, with lower annual average GPP, Re , ET, and canopy interception. • Although the accuracy of GPP and ET simulations did not show a significant improvement after canopy interception correction, it notably enhanced the accuracy of canopy interception, and soil moisture simulations. The process-based Biome-BGC (Biome Biogeochemical Cycles) model is widely used to simulate the fluxes of carbon and water of terrestrial ecosystems. While exhibiting excellent performance in simulating carbon cycle processes, this model provides a relatively simple simulation of the water cycle processes, particularly in canopy interception, which may result in significant errors in evapotranspiration (ET) and soil moisture estimations. This study initially refined the temporal scale of the original Biome-BGC model from a daily scale to an hourly scale, and then incorporated a theoretical interception module into the hourly-scale model for correcting canopy interception algorithm, to enhance the accuracy of hydrological processes simulations. The modified model was tested using the half-hourly observed meteorological and eddy covariance flux data at the Qianyanzhou subtropical coniferous forest site. In comparison with the original daily-scale model, simulations from the hourly-scale model showed better agreement with measurements at Qianyanzhou forest site, with 30.61 % and 40.92 % lower annual average gross primary productivity (GPP) and ET simulations. Although the accuracy of GPP and ET simulations did not show a significant improvement after canopy interception correction, it notably enhanced the accuracy of canopy interception and soil moisture simulations. The canopy interception rates were 55.8 % and 44.1 % for the original daily-scale and hourly-scale model, obviously overstimated compared with canopy correction simulation (31.1 %), leading to a significant underestimation of runoff. The canopy interception and runoff simulations after interception correction were proven relatively more reasonable. Additionally, the coefficient of determination (R2) for measured vs simulated soil water content (SWC) were 0.38, 0.51 and 0.60 for the original daily-scale, hourly-scale, and canopy interception correction simulations respectively, with a notable improvement in accuracy. This study suggested that improvements in temporal scale and canopy interception for process-based biogeochemical model can provide a better understanding of the carbon and water exchanges in response to instantaneous meteorological conditions and support improved water resource management. [ABSTRACT FROM AUTHOR]

• Measurement errors of ε were determined for eight EM sensors in high EC solutions. • Visualized measurement errors for each sensor under varying EC conditions. • Verified feasibility of insulation-coated probes for measuring ε in high EC conditions. Soil water content stands as a pivotal parameter across numerous research disciplines and engineering applications. Electromagnetic (EM) sensors have gained widespread usage in monitoring soil water content. While commonly employed sensors are suitable for typical agronomic- or landscape-soil environments, they may exhibit significant deviations under conditions of high-salinity or -electrical conductivity (EC). We investigated sensing limits under the influence of increasing EC and the resulting accuracy of dielectric permittivity (ε) determinations using commercial soil water content sensors. Tests were conducted using eight commercially available EM sensors (CS655, HydraProbe II, TDR-305N, TDR-310N, TDR-315N, TEROS 12, TEROS ONE, and WET150). For these uncoated rod sensors, most were accurate within ±3 % relative permittivity errors in saline solutions within the range of 0‒4 dS m−1 while TDR-305N, TEROS ONE and WET150 maintained an accuracy of 5 % for EC values up to 9.24 dS m−1. We explored extending the working range of some sensors to high-saline environments by adding an insulating coating on sensor rods. Since the coated probe not only perceives the characteristics of the immersed medium but also interacts with the coating material itself, a calibration model was evaluated for determining the actual ε of the evaluated medium. The strong correlation between a four-point calibration and the model demonstrates the effectiveness of calibrating coated EM sensors for ε determination using only a limited number of calibration materials. The findings presented here could serve as a valuable resource when selecting EM sensors for use in high-saline environments. Results also shed light on the potential to extend the sensing range through the integration of insulating coatings. [ABSTRACT FROM AUTHOR]

• Identified the importance of deep soil water (DSW) in the ecohydrological cycle. • Extreme dry years had 19–26% more DSW uptake than extreme wet years across varying deep soil desiccation (DSD). • DSD inhibited cumulative actual transpiration (T) and evapotranspiration (ET), with the adverse effect exacerbated by an increase in DSD and a decrease in precipitation. • T/ET ratio is little constrained by precipitation due to deep root water uptake. Ecohydrological processes in water-limited ecosystems are sensitive to deep soil water (DSW), but it remains unclear on how and to what extent deep soil desiccation (DSD) affects these ecohydrological processes, especially during meteorological drought. To address this gap in understanding, we used the Hydrus-2D model — calibrated by continuously observed soil water contents from 0–450 cm soil depth, transpiration (T), and evaporation during the 2018–2021 growing seasons in a dryland apple orchard to determine the ecohydrological response under varying degrees of DSD (> 200 cm) across a range of precipitation levels. The results indicated that deep-rooted apple trees used soil water to a depth of up to 1000 cm, with the proportion of DSW increasing in extremely dry years, causing DSD that inhibited cumulative T and evapotranspiration (ET), which occurred to a greater extent with higher DSD and lower precipitation. Severe deep desiccation and the lack of DSW resources decreased cumulative T by 19.3–23.8% and 37.5–42.5% and ET by 13.0–14.4% and 11.2–23.5%, respectively, across all precipitation amounts in comparison to the cases of no deep desiccation, mild deep desiccation, and moderate deep desiccation. Furthermore, extremely dry years had higher transpiration to evapotranspiration (T/ET) ratios than extremely wet years, independent of the DSD scenario, and precipitation dominated the T/ET change once DSW was unavailable. These findings enhance our understanding of the impacts of the combination of deep soil drought and meteorological drought on ecohydrological processes in water-limited ecosystems and have significant implications for future vegetation dynamics and water resource management in a changing climate. [ABSTRACT FROM AUTHOR]

• Species-specific SWP and SMI functions are developed to improve O 3 risk assessment. • SMI can represent soil moisture limitations in flux-based O 3 risk assessment. • Modelled SMI tends to overestimate soil moisture, especially in dry seasons/years. • SWP is more capable than SMI in showing the different plant water saving strategies. • The depth at which the soil moisture is measured has an effect on POD y estimation. The Mediterranean region chronically experiences high levels of tropospheric ozone (O 3) that can affect the health of vegetation. However, limiting plant growing conditions, such as low soil moisture, may restrict the stomatal phytotoxic ozone dose (POD) absorbed by vegetation, modulating O 3 detrimental effects. Atmospheric chemistry transport models that estimate POD for O 3 risk assessment of effects on vegetation species, such as the European Monitoring and Evaluation Programme (EMEP), have adopted the soil moisture index (SMI) to consider the influence of soil moisture on POD. The objectives of this study were the parametrization and validation of the SMI effect on stomatal conductance (g s) for improving the POD estimation and O 3 risk assessment for different vegetation species under water-limiting growing conditions, using field data collected in Italy and Spain and a literature review. The modelled SMI from EMEP proved to be a good indicator of soil moisture dynamics across sites and years, although it showed a general tendency to overestimate soil moisture availability for plants, particularly in the driest seasons. New parametrizations derived for modelling SMI effects on g s under Mediterranean conditions proposed in this study stress the importance of using species-specific parameters for species showing contrasting water-saving strategies in contrast of the current approach of using a simple relation between SMI and g s for all the species. Furthermore, g s modelling parametrizations based on soil water potential (SWP) were found to be more suitable than SMI for local scale estimation of POD under water-limiting conditions. Further consideration of rooting depth and distribution will be required in the future to determine the soil depth at which the soil moisture should be measured in POD modelling, since these features represent one of the most important uncertainties affecting the estimation of POD that could not be addressed with the present database. [ABSTRACT FROM AUTHOR]

• CropSMPAC model performed well on heat transfer under various mulching factors. • Transparent film (TM) warmed up all day, while black film (BM) only at night. • Closer soil-mulch distance increased soil temperature for BM, while reduced for TM. • CropSMPAC model captured weakening effect of crop canopy on film mulching warming. The impact of diverse mulching factors on crop growth depends on their influences on heat transfer, while the precise effects of these factors on heat transfer remain unclear. To address this knowledge gap, we employed the CropSMPAC model to simulate energy fluxes and soil temperature under varying mulching conditions. Our study integrated a soil column experiment and a three–year field experiment. The soil column experiment encompassed 13 treatments, incorporating 3 plastic film colors, i.e., transparent film (TM), black film (BM), and silver–grey film (GM), and 2 mulching ratios (f m), i.e., 100 % and 75 %, as well as 2 distances between soil and film (Z sm), i.e., 0 and 5 mm, along with a control treatment (no mulching). The filed experiment comprised 2 treatments, i.e., film mulching (FM) and no mulching (NM), utilizing TM with a f m of 97.98 % and Z sm of 5 mm under FM condition. Results demonstrated the robust performance of the CropSMPAC model in predicting hourly soil surface temperature, hourly soil temperature in the night at 10 cm depth, daily soil water content at 10 cm depth across varying mulching scenarios. Furthermore, the model aptly captured soil temperature, net radiation flux (R n) and soil heat flux (G) during the maize growth stages under both FM and NM conditions. For daily soil temperature at 10 cm depth, the root mean square error (RMSE) were 1.71 and 2.71 °C, Nash–Sutcliffe efficiency coefficient (NSE) were 0.79 and 0.55, and determination coefficient (R2) were 0.76 and 0.75 for FM and NM, respectively. Corresponding values for daily R n were 37.3 and 42.7 W m–2 (RMSE), 0.56 and 0.47 (NSE), and 0.72 and 0.66 (R2), they were 8.5 and 6.9 W m–2 (RMSE), 0.44 and 0.56 (NSE), and 0.62 and 0.72 (R2) for daily G. Both measurements and simulations revealed that TM increased soil temperature in the daytime and night. In contrast, BM and GM raised soil temperature only in the night. The soil temperature under f m of 100 % was higher than under f m of 75 % for both TM and BM. Film mulching with Z sm of 5 mm contributed to an increase in soil temperature compared with Z sm of 0 mm for TM, while led to a reduction for BM. Additionally, a dense crop canopy helped mitigate the fluctuations in G and soil temperature, and the warming effect of plastic film mulching also weakened with the increase of canopy coverage. [ABSTRACT FROM AUTHOR]

• Canopy temperature was warmer than air temperature during most of the growing season. • T c :T air slopes increased following a water- to energy-limited gradient. • We introduced a framework linking canopy temperature, radiation, and soil moisture. Canopy temperature (T c) plays an important role in regulating the rates of mass and energy fluxes at the leaf surface. Better understanding of the relationship between T c and water availability may enable more accurate monitoring of ecosystem functioning in a changing climate. Here, we used high spatiotemporal resolution thermal infrared cameras deployed at three eddy covariance flux tower sites along a water- to energy-limited gradient – including a predominately water-limited grassland/shrubland site, a seasonally water-limited evergreen needleleaf forest, and a predominantly energy-limited deciduous broadleaf forest – to determine T c seasonality and its relationship with gross primary productivity (GPP) and environmental drivers. We found midday T c was generally warmer than air temperature (T air) during the growing season (T c :T air slope: 1.14–1.27) for all sites. Water-limited sites exhibited higher positive T c deviations from T air (2.30 ± 0.06 °C) compared to the energy-limited site (1.29 ± 0.09 °C) partly due to their reduced latent heat fluxes during water-limited periods. We further found that the T c :T air slope increased with site aridity, namely for 1.14 slope for the grassland, 1.15 for the evergreen forest, and 1.27 for the broadleaf forest. Peak GPP occurred when T c was higher than T air across all sites, with peak GPP at the grassland site occurring at +1.1 °C (T c -T air) and peak GPP at the broadleaf evergreen site occurring at +2.2 °C (T c -T air). T c -T air dynamics were mostly associated with soil water content at water-limited sites where canopies undergo a substantial cooling during the transition from dormancy to the peak GPP, while net radiation played a crucial role at the energy-limited site where the canopy heats up compared to T air over the same phenological transition. Our findings provide novel insights into T c -ecosystem water availability links, highlighting the drivers of T c -T air across diverse ecosystems in various phenological stages, which has implications for ecosystem management in a changing climate. [ABSTRACT FROM AUTHOR]

• Study on forest heterogeneity's impact on water/energy partitioning in boreal forest. • A process-based model and local forcing were used to quantify land-surface fluxes. • Soil water content dynamic is shaped by physiographic factors and less by climate. • An insightful case study on CLASS's local-scale physical processes reliability. • Total evapotranspiration stays uniform but shows great variation in its partitioning. High latitude regions, including the circumpolar boreal biome, are experiencing important changes in the availability of usable surface water because of climate change. In this context, an adequate representation of the land-atmosphere interaction is critical to ensure optimal management of current and future water resources, forest management, and climate prediction. However, the task is particularly intricate in high-latitude boreal forest, as land surface model faces several challenges due to the unique environmental conditions and ecological characteristics. The objective of this study is to quantify the impact of forest landscape heterogeneity, specifically stand leaf-area index (LAI), soil texture, and drainage regime, on surface water and energy balance in a small boreal high-latitude sub-catchment. To this end, hydrometeorological conditions at seventeen 20×20 m plots in a 1-km2 boreal forest sub-basin are simulated using the Canadian Land Surface Scheme (CLASS), a land surface model, at the point scale. The subplot-scale soil texture, drainage regime, and vegetation characteristics and type are based as closely as possible on field measurements and observations for the 17 plots. The model-driven experiment comprises two sets of simulations using CLASS, each employing the same model setup and run for the 17 experimental plots. The main set employs meteorological forcing from a local micrometeorological tower within the sub-basin to investigate the plot-to-plot variability of albedo, energy fluxes, and soil state variables. A second set of simulations is conducted using meteorological forcing from the ERA5-Land reanalysis, which spans from 1986 to 2022. This data provides a longer time series, enabling a more accurate representation of the interannual climatic variability in the sub-basin. The results of the main and secondary sets of CLASS simulations are used to assess the plot-to-plot and temporal variability of several key hydrometeorological variables by calculating a monthly spread. In brief, the following conclusions and broader implications can be drawn from the findings: i) The simulated total annual evapotranspiration remains relatively uniform between plots despite notable variation in its partitioning from plot to plot. ii) In the presence of a full snowpack, the albedo exhibits substantial heterogeneity at the subplot scale, linked to the canopy's LAI. iii) Local soil properties, drainage regime, and vegetation structure and type exhibit substantial influence on the plot-to-plot variability in soil water content. iv) When parameterized with localized observations and measurements, CLASS can represent and be responsive to the complex dynamics of energy and water fluxes at the plot scale within the heterogeneous surface of boreal forests. [ABSTRACT FROM AUTHOR]

• We report eddy-covariance measured carbon fluxes and the water use efficiency of an intensively managed forage field. • More carbon was removed as harvest than returned as manure, resulting in a net carbon import/export deficit. • Changes in precipitation and soil temperature impacted annual carbon uptake and the effect was modified by harvesting frequency. • Separating ecosystem respiration into heterotrophic and autotrophic components suggested that the individual components of ecosystem respiration may have been affected by harvest. • Crop harvest suppressed gross primary productivity more than evapotranspiration which decreased water use efficiency. Intensively managed grasslands have been found to be either carbon (C) sources or sinks depending on management and climate. This study reports the net ecosystem production (NEP) and latent heat fluxes (λE) from a managed forage field at a dairy farm in Agassiz, British Columbia, Canada. The forage crop (ryegrass and tall fescue) was harvested up to 6 times a year. The field received multiple applications of dairy manure slurry and was also fertilized with inorganic nitrogen. Eddy-covariance measurements of NEP were combined with C imports (manure additions) and exports (harvested biomass) to determine the net ecosystem C balance (NECB), and values of gross primary production (GPP) and λE were used to determine water use efficiency (WUE). In terms of environmental controls on NEP, variability of daytime NEP was well described by fitting measured incoming photosynthetically active radiation with a rectangular hyperbolic light-response curve, but variability in nighttime NEP was less effectively described by soil temperature and soil moisture. After accounting for C imports and exports, the NECB of the field was -315 ± 141 and -51 ± 148 g C m-2 y-1 (± indicates the uncertainty range) during the 2020 and 2021 study years, respectively, indicating C was lost from the field and was strongly influenced by C imports and exports relative to NEP. Higher than normal soil moisture and precipitation as well as higher than normal air temperature were both found to suppress GPP and ecosystem respiration (R e), but annual NEP was more impacted by soil moisture in the first year (2020) due to its effect of lowering GPP compared to high air temperature (including the 2021 Pacific Northwest heat dome) and low soil moisture in the second year due to their greater impact on R e relative to GPP. Crop harvests were found to substantially reduce both GPP and WUE which suggests that the intensity of management in terms of harvest frequency could be modified to improve long-term C sequestration. [ABSTRACT FROM AUTHOR]

• We addressed the single impact of soil warming on tree water use and canopy stomatal conductance of two different tree species. • Tree species with anisohydric stomatal strategy could enhance water use under soil warming, while isohydric species like S. superba decrease tree transpiration with warming in subtropical regions. • Whole-tree hydraulic traits are hardly affected by climate warming. • Soil water content is less influenced under warming in subtropical regions. As air temperature and vapor pressure deficit (VPD) increase continuously, forests are losing more water through evapotranspiration, with large consequences for local and global hydrological cycles. In regions with high vegetation cover, soil warming can be more pronounced than warming of the air because of reduced air movement in the forest understory and high heat conduction in the soil. Hence, soil warming could further amplify climate change impacts on forest water uptake and use. Yet, the independent effects of persistent soil warming on tree transpiration remain highly uncertain, especially in sub-tropical forests where soil temperature rise has been considerable in recent years. We carried out a long-term soil warming experiment (+2 °C) in a mixed subtropical forest dominated by the relatively anisohydric Acacia auriculiformis and isohydric Schima superba. Throughout the dry and wet seasons, we tracked whole-tree transpiration (E L), night-time water recharge (E L_night), hydraulic conductivity (K), and canopy stomatal conductance (G s) on mature trees of both species. We found that E L and G s of A. auriculiformis increased (+ 41 % and + 5 %, respectively) while it decreased in S. superba (- 5 % and - 21 %, respectively) with soil warming. However, no changes in K were found for either species. Moreover, soil warming did not change the sensitivity of E L and G s to VPD in both species, but the reference canopy stomatal conductance (G sref) was influenced. Our results suggest that soil warming significantly alters tree canopy transpiration, independently of air temperature. In sub-tropical climates, species with anisohydric stomatal strategy, such as A. auriculiformis , could take up more water and continue to transpire under both high and low VPD, potentially leading to increased nutrient and carbon uptake, and growth. In contrast, isohydric species like S. superba seem adversely affected by soil warming, which may compromise its dominance in subtropical forests in the future. [ABSTRACT FROM AUTHOR]

• SWA dominated the changes in water cost of poplar plantation under dry-heat events. • Dry-heat events caused GPP-ET decoupling in high SWA conditions. • Water use strategy was sensitive to the interaction between SWA and VPD. • Land-atmosphere feedback increases the potential risk of poplar plantation. Terrestrial ecosystems are increasingly threatened by more frequent and intensified compound dry-heat conditions, exacerbated by land-atmosphere feedback under changing climate. The water-use strategy (WUS), which determines the trade-off between carbon assimilation and water consumption, is a key metric for exploring how terrestrial ecosystems respond to environmental stresses. However, how water availability influences their water-use strategies in response to dry-heat conditions is still not well understood. Here, we compared the response of WUS to dry-heat conditions in two poplar plantations with different soil water availability (SWA) in North China, represented by the Priestley-Taylor coefficient (α), and applied an interpretable machine learning algorithm to disentangle the independent and interactive effects of the main biophysical factors on WUS under dry, heat, and their compound conditions. We found that the plantation with higher SWA (high-SWA plantation) showed a more permissive WUS (i.e., increased marginal water cost of carbon gain) under dry-heat conditions than the low-SWA one due to the decoupling of evapotranspiration and gross primary productivity, and less sensitive to α and vapor pressure deficit (VPD) variations. During dry conditions, heat made the WUS more permissive in the high-SWA plantation and more conservative in the low-SWA one. Moreover, the α-VPD interaction was progressively strengthened as intensified dry-heat stresses until reaching a threshold (α = 0.7, VPD = 2.2 kPa in the low-SWA plantation and α = 0.85, VPD = 2.6 kPa in the high-SWA plantation). Thus, our study emphasizes that land-atmosphere feedback exacerbates the vulnerability of poplar plantations to dry-heat conditions. [ABSTRACT FROM AUTHOR]

• The three-dimensional identification method is helpful for the objective description on the dynamic process of drought event. • Drought event pairs matched using a novel spatiotemporal connection procedure can profoundly reveal the complex linkage between two drought types. • Drought propagation features under various response categories are comprehensively depicted from a higher dimension. Studies show that meteorological and agricultural droughts are inherently correlated, and understanding the spatiotemporal propagation mechanism between these two drought types is crucial for drought early warning and mitigation. In this study, the standardized precipitation evapotranspiration index (SPEI) and standardized soil moisture index (SSMI) are used to evaluate meteorological drought and agricultural drought in northwestern China from 1960 to 2018. A three-dimensional clustering method is then adopted to capture drought events, and analyze drought dynamics. Furthermore, meteorological-agricultural drought event pairs are matched from a higher dimension (i.e., space and time), offering profound insights into drought propagation mechanisms on both temporal and spatial scales. The results indicate that agricultural drought exhibits longer duration but smaller magnitude compared to meteorological drought based on identification results, and the dynamic migration process of individual drought event is visually and explicitly described from a three-dimensional perspective. Using a novel spatiotemporal matching procedure, 53 drought event pairs are successfully matched. Following matching, the magnitude of drought events escalates, characterized by prolonged duration, expanded area, increased severity, and lengthened migration distance. Minor meteorological drought events are less likely to cause agricultural drought events, while drought events with larger magnitudes are prone to induce agricultural drought events in various ways. The propagation of most paired drought events involved in the C1, C2 and C4 categories moves in the lateral direction with the relatively long average propagation distance, while moving in the vertical direction for the C3 category. Additionally, the season of paired drought events with the highest occurrence is spring, spring, summer, and autumn for the C1, C2, C3 and C4 categories, respectively. From a higher perspective, the propagation process from meteorological to agricultural drought is effectively and visually depicted, characterized by pooling effect, lengthening effect, lagging effect, and attenuation effect. The findings of this study contribute to a better understanding of drought development and propagation mechanisms, providing solutions to the application restrictions of traditional methods in regions with broad spatial coverage. [Display omitted] [ABSTRACT FROM AUTHOR]

• A downscaling-and-fusion framework was proposed for generating complete 1 km SSM. • BME method considering multi-spatial-scale uncertainties was presented for fusion. • The completeness of fine-scale SSM is directly and significantly improved. • The accuracy and variations of downscaled SSM are well preserved into fused SSM. Surface soil moisture (SSM) plays a crucial role in global energy exchanges and serves as a key requirement for drought monitoring and agriculture production. Microwave remote sensing (RS) SSM products at coarse scale (e.g., 0.25°) can hardly meet the requirements for agricultural and meteorological applications, which necessitate seamless fine-scale SSM data. Current methods for generating complete fine-scale SSM often involve initially filling spatial gaps in SSM ancillary variables (e.g., land surface temperature). However, the uncertainties introduced during the filling process are not adequately and quantitatively considered and utilized in subsequent downscaling process. In this study, we developed a downscaling-and-fusion framework to generate spatiotemporally complete fine-scale SSM (1 km) in the central Tibetan Plateau (TP) of the southwestern China. Firstly, the incomplete downscaled SSM was retrieved by disaggregating the bias-corrected European Space Agency's Climate Change Initiative (ESA CCI, 0.25°) SSM based on apparent thermal inertia (ATI), and then the Bayesian maximum entropy (BME) method considering multi-spatial-scale uncertainties was employed to spatiotemporally merge downscaled and coarse-scale CCI SSM for seamless fine-scale SSM. The completeness of valid merged SSM pixels and pixels unable to retrieved SSM sums up to over 99.9 %. Comparisons with in situ SSM measurements show that merged SSM with more samples have comparable overall performance and slightly better station-level performance than the downscaled SSM with less samples. Station-level statistical metrics averages are superior to those of the overall evaluation for the merged SSM, especially in terms of R (0.592 versus 0.441) and ubRMSE (0.083 versus 0.115 m3/m3). The merged SSM can effectively reflect temporal variability of the in situ SSM measurements, accurately capture the changes of SSM caused by the precipitation events, and preserve the fine-spatial-scale variations of the downscaled SSM. These results demonstrate the success of the proposed downscaling-and-fusion framework in generating spatiotemporally complete fine-scale SSM. [ABSTRACT FROM AUTHOR]

• The summer SMC showed an interdecadal fluctuations, but shifted from drying to wetting at a cut-off point in 2009 over the SAVR of the Loess Plateau. • Spatially, SMC in summer tended to increase in high-value areas (≥90 m3/m3) and decrease in low-value areas (<70 m3/m3). • More summer SMC over the Loess Plateau was associated with larger water vapor budget and favorable circulation pattern. • Enhanced moisture accumulation on the Loess Plateau was attributed to westward expansion of WPSH and anomalous low-pressure system over the Lake Baikal. Soil moisture is a key regulator of hydrological cycle and is closely related to vegetation growth and climate changes, which is mainly replenished by summer precipitation on the Loess Plateau. However, the response of soil moisture content (SMC) anomalies during summer on the Loess Plateau to large-scale teleconnection remains unclear. Here, using the empirical orthogonal function (EOF) method, this study extracted a homogenized area with significant variability of normalized difference vegetation index (NDVI) (p <0.01), which was called sensitive areas for vegetation restoration (SAVR) of the Loess Plateau. Based on 1 km daily soil moisture dataset named SMCI1.0 at top 1 m, we demonstrated that there was an interannual fluctuation of summer SMC from 2000 to 2020 over the SAVR of the Loess Plateau. Spatially, SMC during summer in the SAVR of the Loess Plateau exhibited an enhanced regional difference characterized by 'dry to drier and wet to wetter'. Importantly, abnormal SMC during summer in the SAVR of the Loess Plateau can been attributed to distinct circulation pattern and anomalous water vapor transport, which showed asymmetrical distributions between positive and negative anomalous years of summer SMC. Due to the activities of western Pacific subtropical high (WPSH) and low-pressure system around the Lake Baikal (an anomalous anticyclone prevailed over the western north Pacific), the positive (negative) anomalous years of summer SMC witnessed net water vapor influx (outflux) anomalies to the Loess Plateau compared to the climatology, corresponding to anomalous upward (sinking) motions, further leading to more (less) SMC in summer over the SAVR of the Loess Plateau, respectively. The conclusions of this study have implications for excavating early warning information of SMC anomalies and natural restoration sustainable management on the Loess Plateau. [Display omitted] [ABSTRACT FROM AUTHOR]

• Precipitation seasonality determines the main types of evapotranspiration coupling. • Hydrological memory influences evapotranspiration in some regions of the basin. • Water feedback increases with distance from the Atlantic Ocean. The hydrological cycle of the Amazon Basin is characterized by a multitude of interactions between the atmosphere and surface processes, involving several temporal and spatial scales. Evapotranspiration is a key variable in the water and energy balance, particularly in this watershed that is home to the world largest rainforest. This study analyzed evapotranspiration in the Amazon Basin using the concepts of Coupling (one way relationship between two variables), Memory (temporal permanence of a state or anomaly in the behavior of one variable) and Feedback (two way relationships between two variables), during a 62-year period (12/1959–11/2021). This study found that the seasonality of precipitation, mainly due to the South American Monsoon, determines when evapotranspiration is mainly driven by (or coupled with) net radiation at the surface or soil moisture. Moreover, in the regions where seasonality of precipitation is strong, the previous hydroclimatic conditions due to hydrological memory (reaching just over 90 days in shallow soils) influences the evapotranspiration during the relatively dry season. Regarding water feedback, this study found that the dependence of seasonal precipitation of the Amazon Basin from their hydrobiological processes (fraction of precipitation recycled by Amazonian evapotranspiration) increases with the distance from the Atlantic Ocean, mainly during the non-monsoonal circulation. This process is particularly remarkable in the Amazon-Andes transition zone. However, the spatial pattern and height of water recycled in this watershed is mainly determined by the dominant circulation pattern of atmospheric fluxes in each season. [Display omitted] [ABSTRACT FROM AUTHOR]

• Deep rooted vegetation significantly leads to soil water deficit with the depth and precipitation. • Relative deep soil water deficit in artificial forest decreases with precipitation increase. • Precipitation self-regulation responses to deep soil water deficit in space. • Soil water deficit is more sensitive in the sub-shallow soil layers at a depth of 1800 cm. In semi-arid and semi-humid regions, deep soil water resources are crucial for vegetation restoration to mediate the ecosystem sustainability. Yet it is still limited on understanding soil water deficit (SWD) caused by vegetation restoration and its response to soil depth and precipitation. To obtain a more accurate measurement of regional SWD, we conducted a two-year soil water monitoring to a depth of 1800 cm under typical vegetation restoration with a precipitation gradient ranging from 300 mm/y to more than 600 mm/y across the Loess Plateau of China. With increasing precipitation gradient, soil water storages in 0–1800 cm depth under both arable cropland and artificial vegetation significantly increased (p < 0.05). Compared to arable cropland, artificial vegetations (forest and shrub) had a significantly greater SWD regardless of either precipitation gradient or soil depth. In particular, the deep-rooted vegetations (R. pseudoacacia and C. korshinskii) on the Loess Plateau demonstrated the greatest SWD in the 0–1800 cm layer due to their great water consumption. Indeed, a positive relationship between the accumulated SWD and the accumulated root biomass was detected, suggesting that the increased root biomass in soil depth may promote soil water consumption. In addition, the relative SWD in artificial vegetation at a depth of 1800 cm decreased significantly (p < 0.05). These findings reveal that there is a mechanism of precipitation self-regulation on the deep SWD, of which is more sensitive in the sub-shallow soil layers. We conclude that the SWD depends on vegetation restoration, soil depth, and precipitation self-regulation, offering a reference for how afforestation affects deep soil water in the Loess Plateau region. This finding also benefits to mediate water balance processes and develop sustainable land management in water-limited regions around the world. [ABSTRACT FROM AUTHOR]

• GRU performed well in simulating NEE with R2 of 0.61–0.93 at single site. • Radiation parameters contributed most to the variability of modeled NEE. • Contributions of individual variables appear complex during the non-growing season. • Simulation accuracy reached to 0.7 and 0.8 across all sites and unique land-cover types. The complex forcing mechanisms of biophysical drivers on ecosystem productivity, as well as some unknown relationships, stipulates the application of artificial intelligence techniques in modeling the ecosystem carbon cycle. In particular, various deep learning algorithms have been successfully applied to comprehend temporal relationships, though some issues regarding capturing temporal features for modeling CO 2 fluxes across time scales remain unsolved. In this study, a gated recurrent unit (GRU) model was trained for simulating half-hourly net ecosystem exchange of CO 2 (NEE) using 12-year continuous flux data (2009–2020) from seven experimental bioenergy crops in southwest Michigan, USA. The fields were historically managed either as corn-soybean rotation agricultural (AGR) or Conservation Reserve Program (CRP) lands and planted to no-till corn (AGR-C and CRP-C), restored-prairie (AGR-Pr and CPR-Pr), switchgrass (AGR-Sw and CRP-Sw) and smooth brome grass (CRP-Ref). We compared NEE simulations for the entire year, growing season, and non-growing season datasets, and analyzed relative importance of biophysical variables on variations of NEE. GRU performed well in simulating NEE at single site, with R2 of 0.89–0.93, 0.85–0.90, and 0.61–0.85 for the annual, growing season, and non-growing season datasets, respectively, which provided comparable accuracy and 6 % of run-time shorter than those of long short-term memory. R2 reached 0.70 and 0.80 across all sites and unique land-cover types. Radiation parameters contributed most to the variability of NEE. Contributions of individual variables appear more complex during the non-growing season and include incoming shortwave radiation, day of year, temperature, Monin-Obukhov stability, wind direction, and soil water content. The six most important forcing variables are consistent with our current understanding, although their ranked importance varied by ecosystem type and modeling scale. These results bring renewed insights in modeling observations at seasonal scales and provide guidance for variable selection to scale up measurements from one site to multiple land-cover types across the landscape. [ABSTRACT FROM AUTHOR]

• The influence path and degree of environmental factors on energy partitioning at different time scales were studied by SEM. • LE was the main consumption item of daytime energy during the growing seasons, even during the shooting and leaf-fall stages. • The influence of arid advection on energy partitioning is obvious during growing seasons, especially at the fruiting stages. • Advection induced by irrigation can be weakened properly by choosing a reasonable irrigation time. In this study, a two-year experiment in an irrigated vineyard was conducted to investigate the variations in energy fluxes (net radiation, R n ; latent heat flux, LE; sensible heat flux, H ; and soil heat flux, G) and quantify the influence of arid advection and environmental factors (vapor pressure deficit (VPD), R n , air temperature, wind speed, precipitation, and volumetric soil water content) on energy partitioning in arid Northwest China. For the diurnal variation, the peak values of the energy fluxes appeared at approximately 14:00, except for G (which appeared at approximately 16:00). LE was the main consumer of daytime (R n >0) energy during the growing season (average LE/(R n – G) was 87% and 89% in the two years), even during the shooting and leaf-fall stages. Arid advection, which affects energy partitioning mainly by increasing atmospheric evaporative demand and providing heat energy, contributed 6–60% and 1–56% of the average daytime LE during the two growing seasons, especially at the fruiting stage. Moreover, advection accounted for more than half of average daytime energy imbalances. We found that the arid advection induced by irrigation could be attenuated by selecting a reasonable irrigation time. According to the structural equation modeling (SEM) analysis, R n had the strongest direct and positive regulation effect on LE at the half-hourly and daily scales, indicating that LE was limited by energy rather than water in the study area. The influence of R n on LE gradually weakened with an increase in the time scale, whereas the effect of VPD increased, which may have been due to the smaller time-lag effect between VPD and LE. The diurnal variations in LE at approximately 08:00–19:30 and 19:30–08:00 were mainly controlled by the direct positive effects of R n and VPD, respectively. The results obtained in this study will provide a better understanding of surface processes and help improve water resource management in arid agricultural areas. [ABSTRACT FROM AUTHOR]

• The changes of land use/cover and climate in TRSR after 2000 were evident. • After 2000, the water conservation in TRSR increased significantly. • Climate change dominates the increase in water conservation. • The water conservation is highest for YL, followed by YTZ, and lowest for LC. • The effect of ecological measures is limited, but it is obvious in Yellow River. The water conservation function is one of the important ecosystem services. Quantifying water conservation of the Three-River Source Region (TRSR) of the Qinghai–Tibet Plateau in China and its attribution are essential for maintaining the water needs of the local ecosystem and providing sufficient water resources to downstream countries and regions. However, there is a lack of comprehensive knowledge about the variations of water conservation on muti-temporal (annual, monthly) and spatial scales, and its attributions in land use/cover change (LUCC) and climate change (CC). In this study, the Soil and Water Assessment Tool (SWAT), a distributed hydrological model with flexibility, high adaptability, and explicit physical mechanism, was used to accurately estimate the water conservation of TRSR from 1981 to 2020, and then the attribution analyses were conducted on the TRSR and its three subregions of the Yellow River Source (YL), the Yangtze River Source (YTZ), and the Lancang River Source (LC). The results show that (1) The environmental changes including LUCC and CC in TRSR after 2000 were significant and dominated the overall trend. (2) The multi-year average water conservation in the TRSR was 65 mm and highest for the YL, followed by the YTZ, and lowest for the LC. Water conservation decreased initially and then increased, with an overall increase (0.94 mm/a). Water conservation increased by 45.3% in the later period, mainly occurred in the summer. The most significant increasing trend occurred in the YTZ, while LC was a key water conservation reserve. (3) CC dominated the increase in water conservation of TRSR with contributions of 83.5% where increased precipitation was the main reason. Increased evapotranspiration was responsible for water conservation reduction in most areas of the LC. (4) The effects of ecological measures were limited (accounted for 12.72%), but obvious in the YL (51.31%). [ABSTRACT FROM AUTHOR]

• Almost all of deep soil available water was consumed as apple trees aged in the subhumid area of Loess Plateau. • A distinct adaptive response of apple trees to permanently reduced deep soil water seems to occur primarily by increasing their hydraulic safety rather than reducing hydraulic efficiency. • The plasticity of the embolism resistance of trees should be considered while using hydraulic models to predict trees hydraulic failure during long-term droughts. Deep soil water is important for trees to survive droughts in arid and semiarid regions, but often irreversibly decrease as tree age due to limitations in soil infiltration rates and precipitation. How adaptable is the hydraulic system of trees undergoing the depletion of deep soil moisture reserves? To answer this question, we examined 12 traits characterizing the hydraulic structure of apple tree branches along a plantation age gradient (7 - 26 years) under uniform soil and climatic conditions. We found that almost all deep soil available water was consumed (reduced by up to 98%) as apple trees aged, as a result, the fraction of roots located in water deficit layers increased by up to 70%. Reduced water availability in deep soil significantly increased xylem embolism resistance (P 50) by 8%, without altering xylem-specific hydraulic conductivity, vessel diameter, or vessel density. A distinct adaptive response to permanently reduced deep soil water availability seems to occur primarily by increasing tree hydraulic safety with no reductions in hydraulic efficiency, through which mechanism trees may acclimatize to deep soil water reduction. [ABSTRACT FROM AUTHOR]

• Ecohydrology model simulates water partitioning in tropical coffee plantation. • Water balance reapportionment was sensitive to calibration targets. • Transpiration data needed in calibration to capture green water contribution to total AET. • Including all calibration targets increase model performance and uncertainty. • Shaded coffee transpires more and recharges less water compared to sun plot. Evaporation (E) and transpiration (Tr) are the key terrestrial water fluxes to the atmosphere and are highly sensitive to land cover change. These ecohydrological fluxes can be measured directly only at small scales, such as individual plants or under laboratory experiments. Modelling is needed to upscale E and Tr estimates to plot, hillslope and catchment scales. However, model-derived ecohydrological water partitioning of E and Tr can be ambiguous, particularly when models are trained using hydrometric data and soil moisture. To test the influence of different types of data (i.e., sap flux-derived Tr, Eddy Covariance-derived actual evapotranspiration (AET) and measured soil water (SW)) on model calibration and subsequent water partitioning, we developed the low-parameter plot scale ecohydrology model EcoHydroPlot applied to a data-rich experimental agroforestry plot in humid tropical Costa Rica. The model was able to simulate SW well when calibrated with any data type, but large differences emerged in the E and Tr flux partitioning. Using only hydrometric data for calibration resulted in parameter configurations that produced greater E over Tr fluxes (Tr/AET < 0.5). The opposite was seen for model calibration using Tr data (median Tr simulations with KGE > 0.6), resulting in Tr/AET ratios close to the observed ∼0.9. Further, using all measurements simultaneously (including AET, SW and Tr) did not improve simulated water partitioning. We only found small differences between sun and shade locations with slightly greater average shaded coffee transpiration at the expense of lower upper SW, higher deeper SW and less groundwater recharge compared to sun exposed coffee. This work can inform measurement priorities for applications with relatively simple conceptual ecohydrology models and emphasizes the importance of transpiration estimates for model calibration beyond tropical environments. [ABSTRACT FROM AUTHOR]

• Atmospheric and soil drought stimulate the nighttime water-use behaviors of trees. • The RE makes up a major part of WU n in semi-arid areas with a mean RE % of 65.7–68.0%. • The Ѱ pd and VPD n affect RE % in the soil dry and wet stages, respectively. Recent studies have revealed that the active nighttime water-use behaviors by trees have critical physio-ecological significance in survival and growth in adversity. However, the characteristics and biophysical control of nighttime water-use of silvicultural trees are not clear. We continuously monitored sap flow and leaf water potential (predawn and midday) of Populus simonii of different tree sizes as well as environmental factors during the 2021 and 2022 growing seasons (May-September) in a semi-arid area of Loess Plateau. Our results showed that both nighttime water-use (WU n) and the ratio of nighttime to daily water-use (WU n %) increased significantly with tree size, and had larger values in the drought year than in the normal year. The WU n % of different tree sizes ranged from 10.24 to 13.01% and can reach up to 36.49% during prolonged soil drought. Besides, the WU n % of different tree sizes were significantly positively and negatively correlated with nighttime vapor pressure deficit (VPD n) and soil water content, respectively. By partitioning the components of WU n , we found that as tree size increased, the stem water refilling (RE) and nighttime transpiration increased obviously, but the ratio of RE to WU n (RE %) did not change. The RE of different tree sizes constituted the major part of WU n with an average RE % of 65.7–68.0%, and the RE % was larger in the drought year than in the normal year. Additionally, the RE % of different tree sizes increased with decreasing predawn leaf water potential (Ѱ pd) during the soil dry stage and with increasing VPD n during the soil wet stage. Given the increased risk of atmospheric drought and soil drought, it can be expected from our findings that the nighttime water-use behaviors of trees become more active in the future, which represents a significant performance of drought adaptation strategies of silvicultural species. [ABSTRACT FROM AUTHOR]

• Site-specific soil water supply drives canopy conductance of pine, oak and beech. • Minimum tree water deficit is a valuable proxy for tree water supply. • The operational space of canopy conductance determines tree-water use strategies. • Beech and oak adjust their operational space at a drier site more than pine. Tree-specific canopy conductance (G c) and its adjustment play a critical role in mitigating excess water loss in changing environmental conditions. However, the change of G c sensitivity to environmental conditions due to drought remains unclear for European tree species. Here we quantified the environmental operational space of G c , i.e., the water supply (soil moisture, tree water deficit) and demand conditions (vapor pressure deficit) under which G c ≥ 50% is possible (G c50 OS), at two sites with different soil water availability for three common European tree species. We collected sap flow and dendrometer measurements for co-occurring Pinus sylvestris, Fagus sylvatica and Quercus petraea growing under different soil hydrological conditions (drier/wetter). These measurements were combined with meteorological variables and soil moisture conditions in five depths. Dendrometer measurements were used to confirm soil water availability patterns. For all analyses, the contrasting soil hydrology between sites was the main driver of G c response. At the drier sites, F. sylvatica and P. sylvestris reduced their water consumption in response to decreasing soil water supply earlier in the growing season than Q. petraea. However, our analysis on the G c50 OS revealed that at the drier sites, F. sylvatica and Q. petraea reduced the extent of their G c50 OS to a higher degree than P. sylvestris. This indicates a higher level of G c50 OS adjustment to the drier site conditions for the two broadleaved species. These differences were more pronounced when using the dendrometer-derived tree internal water status as proxy for tree water supply. Our results provide preliminary evidence for diverging short-term G c responses when temperate trees are exposed to prolonged reduction in water availability. These findings suggest that G c50 OS can help to constrain species-specific predictions of water use by mature trees, especially when combined with high-resolution water potential measurements. [ABSTRACT FROM AUTHOR]

• Estimation of actual evapotranspiration (ET) in forest vegetation using inverse simulation of soil parameters. • Eddy covariance measurements over grass were well-correlated to inverse ET values. • Four-year mean cumulative growing season ET estimates are reported for four vegetation species. • Annual cumulative ET in descending order: aspen > deep rooted conifer > shallow rooted conifer > sage > grass. Evapotranspiration (ET) is a major component of water balance and is crucial for understanding the hydrological processes, however due to highly heterogeneous distributions of soils and plants, accurate quantification of ET remains a major challenge in montane ecosystems. In this study, water use for four common vegetation communities was numerically modeled using Hydrus-1D over four continuous growing seasons (2009 to 2012), supported by a network of soil water content sensor measurements. Plant species included i) aspen (Populus tremuloides) with understory dominated by grass/forb (rudbeckia occidentalis, Bromus carinatus and Elymus trachycaulu), ii) conifer (Picea engelmannii and Abies lasiocarpa), iii) grass (dominated by B. carinatus and E. trachycaulu, et al.), and iv) sage (Artemisia tridentata). The numerical model simulated water transport within the soil and water loss from soil evaporation and plant transpiration processes. Simulations were compared with temporal and spatial dynamics of soil water content, measured at 0.1 m, 0.25 m and 0.5 m within each of 36 subplots at 30-minute intervals. Numerically simulated ET in the sage/grass meadow was compared with ET determined from a centrally-located eddy covariance tower. The simulations effectively predicted soil moisture and ET of the montane plant communities during summer dry down. Results indicate vegetation type showed no significant difference (<5 %) in ET comparing four growing seasons, except where abnormally wet conditions occurred. Four-year mean cumulative growing season ET estimates for aspen, deep-rooted conifer, shallow rooted conifer, sage, and grass were 43.0, 40.2, 28.7, 28.8 and 26.1 cm, respectively. [ABSTRACT FROM AUTHOR]

• Irrigation changes precipitation. • Increased irrigation produces shallower planetary boundary layer. • Increased irrigation increases latent heat flux and reduces sensible heat flux. • Increased irrigation increases equivalent temperature and reduces air temperature. Land use land cover change, including irrigation, impacts weather and climate. In this paper a precipitation event that occurred on the morning of 23 July 2018 during the Great Plains Irrigation Experiment (GRAINEX) is investigated. Six Weather and Research Forecasting (WRF) model-based experiments were conducted, which involved the increase or decrease of soil moisture by 5 % and up to 15 % over the irrigated croplands. These changes approximated soil moisture content in response to different levels of irrigation applications. An additional experiment, where irrigated land use was changed to grassland, was conducted to capture pre-irrigation land use and its impacts. It was found that regardless of strength of irrigation, precipitation decreased. In addition, the model did not produce precipitation over non-irrigated land use. When grassland replaced irrigated agriculture, increases in precipitation were estimated. With increased irrigation, latent heat flux increased compared to the control simulation and decreased when irrigation decreased. On the other hand, sensible heat flux was decreased compared to control when irrigation increased. The planetary boundary layer over irrigated land use was shallower than over non-irrigated land use while over grassland it was higher than irrigated but lower than non-irrigated land use. The changes in precipitation, the surface energy balance, and the planetary boundary layer served as a reminder of irrigation's complex effects on the atmosphere. Additional analysis of other precipitation events during GRAINEX would be helpful to better understand the effects of irrigation. [ABSTRACT FROM AUTHOR]

• A modified potential source water line was proposed to identify total δ2H offsets. • Cryogenic vacuum extraction (CVE) accounted for 75.1 % of total xylem δ2H offsets. • Radial water transport (RWT) explained 24.9 % of total xylem δ2H offsets. • Water availability and plant physiology dominantly affected RWT-induced δ2H offsets. The positions and magnitudes of deuterium offsets between bulk xylem and corresponding source waters are under debate and quantifying them is essential for isotope-based ecohydrological investigations. In this study, stable isotopes (δ2H, δ18O, and δ13C), iteration method, and rehydration experiments were combined to quantitatively determine the magnitude of cryogenic vacuum extraction (CVE)- and radial water transport (RWT)-induced deuterium offsets using one riparian tree species Salix babylonica L. A modified potential water source line (MPWL) was proposed to identify the total δ2H offsets between bulk xylem and source waters. The relationships between δ2H offsets induced by CVE or RWT and plant water content, leaf δ13C values, soil water content (SWC), and the depth to the water table (WTD) were investigated. Results showed that the bulk xylem waters in different tissue positions of S. babylonica showed −7.0 ‰ to −4.0 ‰ deuterium depletion relative to MPWL at four different sites (p < 0.01). The isotopic compositions of sap water coincided well with MPWL on the dual-isotope plot at the four sites. The CVE- and RWT-induced δ2H offsets accounted for 75.1 % and 24.9 % of the total δ2H offsets, respectively. The CVE-induced δ2H offsets were significantly negatively correlated with plant water content. In comparison, the RWT-induced δ2H offsets were negatively related to plant leaf δ13C values, trunk water content, and SWC, but positively correlated with WTD. This study provides a quantitative contribution of two major sources of deuterium offsets. The results provide critical insights into isotope-based plant water source identification and evapotranspiration partitioning. [ABSTRACT FROM AUTHOR]

• Coexisting species exhibited large overlaps of water sources. • Plants shifted their water uptake to deeper soil as soil moisture increased. • Proportions of plant-available water and gravity water adjusted plant water uptake. • Plants at the lowest altitude obtained more shallow soil water. Soil water is crucial for maintaining forest health by providing plants with a dependable water source. However, the understanding of the impact of soil moisture on plant water uptake, particularly in conditions of excess soil water, is limited. In this study, we investigated root water uptake strategies of dominant species along an elevational gradient in the subalpine region of the eastern Qinghai–Tibet Plateau. Our results demonstrated divergent relationships between soil moisture and water uptake fractions in the shallow soil layer, where moisture generally exceeds field capacity (θ f c) during the growing season, and in the deep soil layer, showing negative and positive associations, respectively. This disparity is linked to soil water components: when plants mainly used shallow soil water, the relative proportion of plant–available water (the part of water between wilting coefficient and θ f c) in the shallow soil was smaller compared to instances where plants mainly used deep soil water. Conversely, the pattern for gravity water was the opposite. This phenomenon is attributed to precipitation replenishing shallow soil with gravity water and deep soil with plant–available water, leading to poor aeration in the shallow soil and an abundance of plant–available water in the deep soil, and plants tended to shift to deep soil water as a result. Furthermore, the decrease in the relative proportion of plant–available water in the shallow soil along the elevational gradient results in plants in the lowest-altitude broadleaf forest obtaining more shallow soil water and less deep soil water, which can also enhance our understanding of how soil moisture components influence plant water–use strategies. Additionally, no hydrological niche segregation was observed among species in this humid environment. These results can improve our comprehension of soil–plant interactions in subalpine areas. [ABSTRACT FROM AUTHOR]

• Relative impact of SM and VPD on vegetation during droughts conditions. • How drought events alter the dominant water stress of the ecosystems. • High VPD has a stronger control effect on vegetation than low SM. • Droughts could alter the dominant water stress factors in forests and grasslands. Low soil moisture (SM) and high vapor pressure deficit (VPD) have a significant impact on the terrestrial ecosystem productivity through different mechanisms. The sources of relative impact of these two water stress factors are difficult to separate due to the strong interaction between the land atmosphere, especially during drought events with an increasing frequency and intensity, leading to significant controversy over the dominant water stress factors in ecosystems. Here, we analyzed the relative effects of SM and VPD on the Gross Primary Productivity (GPP) during non-drought and drought periods and how drought events alter the dominant water stress of the ecosystems using daily observations from 89 flux sites at the Northern Hemisphere. Our results showed that 84 of 89 sites and 81 of 89 sites were significantly affected by water stress during both non-drought and drought periods, with 43 % and 34 % of sites remaining dominant water stress from the high VPD and low SM during non-drought and drought conditions, respectively. In addition, 19 sites have experienced significant dominant water stress shifts when drought events occurred, including forests (47 %), grasslands (37 %), and shrubs (16 %). The normalized GPP by radiation and temperature also emphasizes that high VPD has a stronger control effect on vegetation than low SM and confirmed that drought events could alter the dominant water stress factors in forests and grasslands. Our results might help to clarify the relative impacts of the SM and VPD on the vegetation, as well as the driving forces of drought events on the shifting role of SM and VPD. The driving mechanism for this shifting varies depending on the biomes and plant functional types Furthermore, our findings could provide a more accurate prediction of the vegetation productivity in different ecosystems, and a deeper understanding of the relationship between the vegetation and the climate. [ABSTRACT FROM AUTHOR]

• We analyzed daily air temperature and soil moisture time series of 54 forest sites. • Moister soils supported cooler understory microclimates. • Temperature buffering was more sensitive to soil moisture in dry soils and on sunny slopes. • Forests can lose their temperature buffering function when soils become drier. Forest canopies can buffer the understory against temperature extremes, often creating cooler microclimates during warm summer days compared to temperatures outside the forest. The buffering of maximum temperatures in the understory results from a combination of canopy shading and air cooling through soil water evaporation and plant transpiration. Therefore, buffering capacity of forests depends on canopy cover and soil moisture content, which are increasingly affected by more frequent and severe canopy disturbances and soil droughts. The extent to which this buffering will be maintained in future conditions is unclear due to the lack of understanding about the relationship between soil moisture and air temperature buffering in interaction with canopy cover and topographic settings. We explored how soil moisture variability affects temperature offsets between outside and inside the forest on a daily basis, using temperature and soil moisture data from 54 sites in temperate broadleaf forests in Central Europe over four climatically different summer seasons. Daily maximum temperatures in forest understories were on average 2 °C cooler than outside temperatures. The buffering of understory temperatures was more effective when soil moisture was higher, and the offsets were more sensitive to soil moisture on sites with drier soils and on sun-exposed slopes with high topographic heat load. Based on these results, the soil–water limitation to forest temperature buffering will become more prevalent under future warmer conditions and will likely lead to changes in understory communities. Thus, our results highlight the urgent need to include soil moisture in models and predictions of forest microclimate, understory biodiversity and tree regeneration, to provide a more precise estimate of the effects of climate change. [Display omitted] [ABSTRACT FROM AUTHOR]

• We analysed eddy covariance data from 12 towers across 7 farms in New Zealand. • Climate effects were isolated from grazing or harvesting using a novel framework. • VPD and air temperature controlled NEE deficit, not soil moisture. Climate change is exposing agricultural systems to more frequent weather extremes, threatening production and challenging the sustainability of current management practices. However, isolating the climatic drivers of CO 2 exchange in managed systems can be complicated by the periodic removal of aboveground biomass through grazing or harvesting, abruptly shifting the potential for plant uptake and associated ecosystem biogeochemistry. We used eddy covariance datasets from managed grasslands across New Zealand to determine the relative effects of temperature, vapour pressure deficit, and soil moisture on net CO 2 uptake. Using time since grazing or harvest as a proxy for above-ground biomass, we accounted for the shifting potential CO 2 exchange associated with these management actions. We then fit light response curves to the outer envelope of data to estimate maximal potential NEE, and subsequently calculate NEE deficit for every measured half hour. Using this analytical framework, we combined data from seven locations making up sixteen individual flux datasets comprising >600 site months and found that high air temperature and vapour pressure deficit were stronger controls over NEE deficit than soil moisture. These results were also consistent with a controlled leaf-level experiment demonstrating increasing photosynthesis deficit as VPD increased under constant temperature. The dominance of atmospheric effects over soil moisture in determining grassland response to dry conditions has important implications because it may be more difficult to mitigate these responses with typical management approaches (e.g., irrigation). We found these patterns were consistent across a diverse range of managed grassland operations, but subtle differences suggest that expanding the approach to a wider array of climatic regions and soil types could provide additional insights into how we might avoid the worst effects of future drying on managed systems. [ABSTRACT FROM AUTHOR]

• Temperate and alpine grasslands were weak carbon sinks. • Elevated water consistently increased carbon fluxes in temperate and alpine grasslands. • Increased temperature reduced carbon fluxes in temperate grasslands and increased carbon fluxes in alpine grasslands. Temperate grassland (TG) and alpine grassland (AG) are two distinct grassland types with different climatic backgrounds and adaptation patterns. Little is known about the universal and divergent responses of TGs and AGs to climate change, particularly with regard to the carbon (C) fluxes. The large-scale responses of the C fluxes in different grasslands to climate change remain unclear, with only a few comparative studies of distinct responses of the C fluxes in TGs and AGs to climate change hinders our understanding of this subject. In this study, we conducted a large-scale transect study across the Mongolian, Loess, and Tibetan Plateaus in China to reveal the similarities and differences in the responses of the C fluxes in TGs and AGs to climate change. We used C flux data measured by eddy covariance from ChinaFLUX and published literature, covering the period from 2003 to 2020. The results showed that TGs and AGs in China were weak C sinks. Net ecosystem productivity in TGs and AGs exhibited opposite trends with increasing latitude, longitude, and altitude. Elevated water (mean annual precipitation and soil water content) consistently increased the C fluxes. Conversely, increased temperature (mean annual temperature and mean annual soil temperature) reduced the C fluxes in TGs and increased the C fluxes in AGs. The water factors promoted the C fluxes in both TGs and AGs by improving the leaf area indices. Conversely, the temperature factors had a strong direct, negative effect on the C fluxes in TGs and a weak direct, positive effect on the C fluxes in AGs, emphasizing the divergent patterns of the C fluxes in TGs and AGs and their responses to climate change. Overall, this study enhances our knowledge on C sinks and various grassland hydrothermal sensitivities. [ABSTRACT FROM AUTHOR]

• Seasonal ET was strongly correlated with g s in an urban forest in Beijing. • Seasonal SWC imposed stronger constraints on g s than VPD. • Annual ET had a weak correlation with PPT amount in the same calendar year. • PPT from the previous year was responsible for the higher spring ET relative to PPT. With urbanization rapidly increasing, evapotranspiration (ET) in urban forests plays an increasingly important role in urban hydrology and climate. However, large uncertainty remains regarding the regulating factors of ET in urban area. Using the eddy-covariance technique, we investigated the temporal variations of ET in an urban forest park in Beijing for 2012–2022. Daily ET was close to zero during winter and reached 3–6 mm day−1 in summer. Daily ET increased with vapor pressure deficit (VPD) and soil water content (SWC). When binning daily data into 10 bins according to percentiles of either SWC or VPD , surface conductance (g s) increased with SWC for all VPD bins, but showed weak responses to VPD for SWC bins. Monthly ET increased linearly with normalized difference vegetation index, showed a strong correlation with g s , and exhibited saturated responses to increasing monthly precipitation (PPT). Annual ET ranged from 326 to 566 mm (with the eleven-year mean of 428.66 ± 83.83 mm; mean±SD), and varied from 55 % to 148 % of annual PPT. Aggregated ET over the eleven years (4715 mm) accounted for 82 % of the aggregated PPT. Soil water replenishment through PPT from mid-summer to late autumn of the previous year was responsible for the generally higher monthly ET in spring relative to PPT , resulting in a weak correlation between annual ET and PPT in the same calendar year (R 2= 0.28). The PPT in the early growing season partly promoted annual ET through its effect on summer g s. Our results suggest that biotic factors and PPT seasonality played an essential role in regulating ET at the seasonal and interannual scales, respectively. Soil dryness imposed additional constraints on g s at the finer timescale (i.e., days). Urban land-surface models should adequately address the soil moisture carryover effect to improve simulations. [ABSTRACT FROM AUTHOR]

• A conductance-based ET partitioning method was proposed in a tree-grass ecosystem. • The partitioned ET components were comparable to sap flux and micro-lysimeter measurements. • Transpiration from tree (T Tree) and grass (T Grass) accounted for 42 % and 32 % ET , respectively. • Trade-off between T Tree / ET and T Grass / ET depended on tree canopy change and atmospheric evaporation demand. Partitioning of evapotranspiration (ET) into its components (vegetation transpiration (T), soil evaporation (E Soil) and evaporation from intercepted water (E i)) in agroforestry systems can provide valuable assessments of plant water use. However, it is still challenging to quantify the contributions of different vegetation elements to T , especially when the underlying surface is complex, consisting of trees and heterogeneous grass. Based on the water-carbon coupling theory and flux conservation hypothesis, an ET partitioning method was proposed to partition total ET into T from the tree (T Tree) and understory grass (T Grass), and E Soil in a humid region tree-grass orchard ecosystem during three growing seasons (from mid-March to end-October) of 2018–2020. The results demonstrated that the partitioned T Tree and understory evapotranspiration (ET un = T Grrass + E Soil) agreed well with the values measured by the heat-pulse sap flow sensors (R2 of 0.744, and RMSE of 0.379 mm d–1) and micro-lysimeters (R2 of 0.650 and RMSE of 0.545 mm d–1). The derived underlying transpiration (T Gs = T Tree + T Grass) was strongly consistent with Zhou's underlying water use efficiency approach (R2 of 0.939 and RMSE of 0.382 mm d–1). The estimated T Gs was 536, 433 and 383 mm during the three growing seasons, resulting in a higher T Gs / ET ratio of 0.74 ± 0.03 (±standard deviation). Among them, T Grass contributed to 0.32 ± 0.05 ET , while the contribution was 0.42 ± 0.02 for T Tree. E soil and residual evaporation (including E i and estimated error) accounted for a small ET proportion, with the value of 0.13 ± 0.02 and 0.14 ± 0.01, respectively. The partitioned ET components (T Tree / ET, T Gs / ET and E Soil / ET) were well controlled by variations of atmospheric evaporation demand and vegetation factors, but were little affected by soil water content. Increased tree leaf area limited T Grass due to shading effects, however, T Grass tended to contribute more transpiration water under a higher atmospheric evaporation environment. The findings of this study enhance the knowledge about the contributions of individual ET components on total ET in a complex tree-grass ecosystem, and provide further insights on plant-plant interactions in terms of water use strategies and orchard water management. [ABSTRACT FROM AUTHOR]

• There is an optimal rainfall for cowpea beyond which yield declined. • Cowpea yield response to water deficit and excess depends on soil type. • Projected heavy rainfalls would lead to excessive soil water even in dryland. • Excessive soil water poses as much of a threat as drought to legume production. • The need for countermeasures for both soil water deficit and excess is underscored. The intensification of the hydrological cycle has increased heavy rainfall and drought events in a changing climate. However, compared to drought, the impacts of heavy rainfall on crop production are under-studied. Using field experimental data and a calibrated crop model CYGMA, we showed that excessive soil water associated with heavy rainfall events is having a detrimental effect on cowpea yields, even in the dry environments of West Africa where cowpea is an important, protein-rich cash crop. Cowpea yields are susceptible to heavy rainfall in areas with poorly drained soils, and to drought in soils that have a low water-retention capacity. The crop model captured of the main characteristics of the observed development, growth, and yield, as well as the characteristics of root-zone soil water contents and how they vary by soil type. The analysis of d4PDF factual and counterfactual climate model simulations revealed that heavy rainfall events associated with anthropogenic climate change have increased in recent decades, and that they are projected to increase in future. Further, changes in seasonal rainfall and the number of dry days would be largely absent from CMIP6 climate projections by mid-century. Reductions in cowpea yields due to excessive soil water is projected to become more frequent, and the potential damage in a 1-in-100 extremely wet year would be comparable to the damage currently experienced in droughts, irrespective of soil types. Simulations of the projected damage due to drought show that the situation will be similar to current levels, with drought remaining a major climate hazard. However, excessive soil water is projected to be a serious threat to food security in the region. Our findings indicate that, even in dry environments, cropping systems need to be implemented in order to reduce the susceptibility of soils to both drought and excessive soil water. [ABSTRACT FROM AUTHOR]

• Precipitation was repackaged to simulate 4-level of precipitation variability. • Extra-extreme precipitation variability scenario potentially reduced N 2 O emissions. • Effects of precipitation variability on N 2 O fluxes had high interannual variations. • Soil, microbes, functional genes, and plants explained 61% variation in N 2 O fluxes. High temporal precipitation variability, characterized by less frequent but larger-magnitude precipitation events, is increasing. However, how precipitation variability affects N 2 O emissions and the underlying mechanisms remain unclear, especially at multiple-year scales. We conducted a 3-year manipulative experiment in which the same long-term mean growing season precipitation total was repackaged into events of inversely varying magnitude and frequency to simulate four levels of precipitation variability (extra-extreme, extreme, medium, and normal) in a semiarid grassland. Cumulative N 2 O emissions was the smallest under the extra-extreme precipitation variability scenario (6 very large rainfall events), 26% less than emission under the other three treatments (10, 16, and 24 rainfall events) over the three growing seasons. However, this difference was almost entirely due to three sampling events in the third year. Plant community (biomass and biodiversity), soil abiotic properties (water, dissolved organic carbon and pH), soil microbial biomass (carbon, nitrogen and the ratio), and soil functional genes (archaeal and bacterial amoA, nirS, nirK, narG , and nosZ) explained 61% of the variation in N 2 O emissions in response to the precipitation variability. Structural equation modelling indicated that the precipitation variability had direct positive effects, and indirect negative effects via soil abiotic properties, on soil functional genes that ultimately had positive effects on N 2 O emissions. The reduction in N 2 O emission in the extra-extreme precipitation variability scenario in the third year was due to low levels of soil water content, soil pH, aboveground biomass, and microbial biomass carbon simultaneously. Our results suggest that at the multi-year timescale, semiarid grasslands may have negative feedbacks to future precipitation regimes with higher variability. [Display omitted] [ABSTRACT FROM AUTHOR]

• The systematic parameter sensitivity analysis and model calibration approach has been implemented into BEPS to calibrate parameters over 10 ecosystems of China. • The constraint of soil hydrological processes together with the physiological processes promotes better simulation of GPP over arid regions. • Higher productivity over the southern regions of China has been detected. Terrestrial ecosystems are the largest sink for carbon, and their ecosystem gross primary productivity (GPP) regulates variations in atmospheric carbon dioxide (CO 2) concentrations. Current process-based ecosystem models used for estimating GPP are subject to large uncertainties due to poorly constrained parameter values. In this study, we implemented a global sensitivity analysis (GSA) on parameters in the Boreal Ecosystem Productivity Simulator (BEPS) considering the parameters' second-order impacts. We also applied the generalized likelihood estimation (GLUE) method, which is flexible for a multi-parameter calibration, to optimize the GPP simulation by BEPS for 10 sites covering 7 plant functional types (PFT) over China. Our optimized results significantly reduced the uncertainty of the simulated GPP over all the sites by 17 % to 82 % and showed that the GPP is sensitive to not only the photosynthesis-related parameters but also the parameters related to the soil water uptake as well as to the energy balance. The optimized GPP across South China showed that the mix forest, shrub, and grass have a higher GPP and are more controlled by the soil water availability. This study showed that the GLUE method together with the GSA scheme could constrain the ecosystem model well when simulating GPP across multiple ecosystems and provide a reasonable estimate of the spatial and temporal distribution of the ecosystem GPP over China. We call for more observations from more sites, as well as data on plant traits, to be collected in China in order to better constrain ecosystem carbon cycle modeling and understand its response to climate change. [ABSTRACT FROM AUTHOR]

• Soil moisture decreased in most areas before 2000, while it showed significantly increasing trend after 2000. • Soil moisture variation induced by vegetation greening was limited compared to climate change. • Surface soil moisture had positive responses to vegetation greening in semiarid subarea. • Vegetation greening exacerbated root zone soil moisture consumption in semiarid and dry subhumid subareas. Drylands are particularly sensitive to climate change, and soil water availability is critical for dryland ecosystems. Climate change and large-scale ecological restoration have led to significant vegetation greening and soil moisture (SM) variations in the drylands. However, the influencing mechanisms of climate change and vegetation greening on SM variability were not well understood. In this study, the spatiotemporal variations of climate variables, vegetation cover, surface SM (SM_surf) and root-zone SM (SM_root) along aridity gradient in the drylands of northern China during 1981–2018 were investigated. The random forest model and partial least-squares structural equation modelling (PLS-SEM) analysis were employed to quantify the contributions of climate change and vegetation greening to SM variability and identify the influencing pathways. The climate change showed temporal trends of wetting to drying before 2000 and drying to more wetting in 2000–2018, and vegetation greening was significant in 2000–2018. The variations of SM displayed obvious east-negative and west-positive trends. The temporal trend of SM deficit decreased during 1981–2018 in most regions, and SM showed a significant increasing trend in all regions of different aridity gradient after 2000. The correlations of climate factors and vegetation cover with SM became stronger in the area with lower aridity. The vegetation greening-induced SM change was limited compared to climate change. The climate factors can directly impact SM availability, and also indirectly influence SM availability through vegetation (e.g., precipitation→LAI→SM_surf). The SM_surf had positive responses to vegetation greening, especially in semiarid subarea. In contrast, vegetation greening exacerbated SM_root consumption in semiarid and dry subhumid subareas, which became intensive after 2000. This study fills the knowledge gap in the underlying reasons for the variations of both surface and root zone SM in water-limited areas and reveals the potential risks of SM depletion, particularly in dry subhumid areas. [ABSTRACT FROM AUTHOR]

• We quantified the sensitivity of seven grassland models to climate change factors. • Consistency amongst the models was variable and depended on the climate change factor. • Multi-model averaging ensured a realistic response to the studied factors. • The representation of hydrology in grassland models needs to be improved. Grassland models often yield more uncertain outputs than arable crop models due to more complex interactions and the largely undocumented sensitivity of grassland models to environmental factors. The aim of the present study was to assess the impact of single-factor changes in temperature, precipitation, and atmospheric [CO 2 ] on simulated soil water content (SWC), actual evapotranspiration (ET), gross primary production (GPP) and yield biomass, and also to link the sensitivity analysis with experimental results. We employed an unprecedented multi-model framework consisting of seven grassland models at nine sites with different environmental characteristics in Europe and Israel, with two management options at three sites. For warming/cooling and wetting/drying, models showed general consistency in the direction of SWC and ET changes, but less agreement regarding GPP and biomass changes. The simulated responses consistently revealed an overall positive effect of CO 2 enrichment on GPP and biomass, while the direction of change differed for SWC and ET. Comparing with single-factor experimental manipulations, SWC simulations slightly underestimated the observed effect of warming, while the overall mean model sensitivity for biomass (+7.5%) closely matched the mean response observed with 1–2 °C warming (+6.6%). The models exhibited lower sensitivity of SWC to wetting or drying compared to the experiments. The overall mean sensitivity of biomass to drying was -4.3%, contrasting with the mean experimental effect size of -9.6%, which proved to be more realistic than the mean wetting effect (+3.2%, against +38.9% in the field trials). The simulated sensitivity of SWC to CO 2 enrichment was markedly underestimated, while the biomass response (+12.0%) closely matched the observations (+17.5%). Although the multi-model averaging did not manifestly improve the realism of the simulations, it ensured a realistic response in the direction of change to varying conditions. The results suggest a paradigm shift in grassland modelling meaning that the usual practice of model optimisation/validation needs to be complemented by a sensitivity analysis following the approach presented. The results also highlight the importance of model improvements, especially in terms of soil hydrology representation, a key environmental driver of grassland functioning. [ABSTRACT FROM AUTHOR]

• New soybean model integrated with a two-dimensional root growth module was evaluated • Improved model successfully reproduced the long term SoyFACE experiment data • Current and future CO 2 effects on soybean yields in the US Midwest were quantified • CO 2 effect on yield was highest in northern counties while it declined in southern and western counties Crop simulation models are indispensable tools that facilitate studies to assess climate impacts and adaptation responses, but have not been adequately tested in terms of accurately predicting crop growth and water stress responses to high carbon dioxide concentrations [CO 2 ]. The soybean model, GLYCIM, previously modified with a coupled leaf-level gas exchange – energy balance model, was integrated with a two-dimensional convective-diffusive root growth module which linked soil and leaf water potentials with the regulation of stomatal conductance. We evaluated the accuracy of this modified GLYCIM using experimental data from a Free-Air CO 2 Enrichment (FACE) site, SoyFACE, and used the calibrated model to quantify current and projected elevated CO 2 effects on soybean yields in the United States (US) Midwest. The model accurately matched SoyFACE observed data including canopy temperature (RMSE, 2.26 -2.42 °C; IA, 0.89-0.90), photosynthesis rate (RMSE, 6.33 -12.88 µmol m−2 s−1; IA, 0.82-0.92), seasonal LAI (RMSE, 0.96 -1.08 cm2 cm−2; IA, 0.91-0.93), biomass (RMSE, 2.06 -2.34 Mg ha−1; IA, 0.87-0.90), soil water contents (RMSE, 5.42 -7.82 %; IA, 0.80-0.85) and root length (RMSE, 2.12 -5.64 m plant−1; IA, 0.70-0.86). GLYCIM also reproduced the observed declining response of yield to CO 2 fertilization with increasing drought. Using an ensemble of five climate model projections, GLYCIM showed that the CO 2 fertilization effect on yield was highest in northern counties and declined in southern and the western counties by late 21st century in response to increased drought characterized by higher daily maximum temperature and decreased rainfall. This study highlights the need to continue to test and improve crop models and use geospatially referenced adaptive strategies to capitalize the benefits from elevated [CO 2 ] to better counteract increasing drought. [ABSTRACT FROM AUTHOR]