The fragile social-ecological systems of Central Asia's arid regions (ACA) are critically dependent on evapotranspiration (ET) dynamics, which govern water and energy cycles in this precipitation-limited region experiencing increasing climatic variability and water stress.
Methods
Using the Global Land Evaporation Amsterdam Model (GLEAM) for historical analysis (1985-2014) and an optimized ensemble of five CMIP6 models (selected via DISO metric) under four SSP scenarios (SSP1-2.6 to SSP5-8.5) for future projections (2021-2100), this study provides a comprehensive assessment of ET trends and climate linkages.
Results
Key results show: (1) Historical ET increased significantly (0.43 mm/a by GLEAM; 0.98 mm/a by CMIP6 MME, p < 0.05); (2) Future projections reveal strongest ET increases under SSP5-8.5 (0.79 mm/a annually), with >85% of ACA (notably northern Kazakhstan and Kunlun Mountains) showing upward trends, though SSP1-2.6 causes growing-season declines in 48% of regions; (3) ET exhibits strong positive correlations with temperature (especially in Kyrgyzstan/Tajikistan/southern Xinjiang) and precipitation (r = 0.52 under SSP1-2.6).
Discussion
These findings underscore ET’s growing hydrological significance in ACA, highlighting the urgent need for adaptive water management strategies to address climate change impacts on agricultural, ecosystem sustainability and zoonics from one health perspective.
arid regions of central asiaCMIP6DISOevapotranspirationfuture dynamical variationsThe author(s) declared that financial support was received for this work and/or its publication. This study was supported by the National Key R&D Program of China (Grant No. 2025YFE0104200), the National Natural Science Foundation of P.R. China (Grant No. 42230708, 42361144887).section-at-acceptanceDrylandsIntroduction
Evapotranspiration (ET) is a fundamental component of the terrestrial water, energy, and carbon cycles, regulating land–atmosphere interactions and exerting strong controls on vegetation dynamics, hydrological processes, and climate variability at regional to global scales.
At the global scale, observation-based datasets and land data assimilation products consistently indicate an overall increase in terrestrial ET during recent decades. Using Global Land Data Assimilation System (GLDAS) data, Liu et al. (2022) reported a global ET increase of approximately 40.3 mm per decade between 2002 and 2020. Climate model simulations further suggest that ET will continue to rise in many regions under future warming scenarios, particularly under high-emission pathways (Zhao et al., 2022). However, pronounced regional heterogeneity exists, reflecting differences in climate regimes, land surface conditions, and water availability.
In arid and semi-arid regions, ET responses to climate change are especially complex. Central Asia, one of the world’s largest contiguous arid regions, is characterized by scarce precipitation, strong continentality, and fragile social–ecological systems. Recent studies have documented spatially heterogeneous ET trends across the region. Hao et al. (2023) found that approximately 17% of Central Asia experienced significant ET declines, while 16% showed significant increases, indicating strong spatial contrasts. Similar heterogeneity has been reported in adjacent regions, such as western China and the Tibetan Plateau, where ET trends exhibit distinct east–west gradients linked to climatic and topographic controls (Han et al., 2021). Despite these efforts, uncertainties remain substantial due to sparse ground observations and discrepancies among ET products, particularly in arid environments.
Temperature and precipitation are the dominant climatic drivers of ET variability, but their influences differ markedly between humid and arid regions. Rising temperature generally enhances atmospheric evaporative demand, which can increase ET when sufficient soil moisture is available. In moisture-limited systems such as arid Central Asia, however, warming may intensify evaporative demand without a corresponding increase in water supply, leading to soil moisture depletion and suppression of actual ET (Kim et al., 2021). Precipitation, by contrast, directly controls water availability and often serves as the primary constraint on ET in drylands. Changes in precipitation amount, seasonality, and intensity can therefore substantially alter ET magnitude and timing, with cascading effects on runoff generation and ecosystem functioning.
Climate model projections based on the Coupled Model Intercomparison Project Phase 6 (CMIP6) suggest continued warming and a tendency toward increased precipitation in parts of Central Asia during the 21st century (Hua et al., 2022; Cao et al., 2023; Su et al., 2024). These changes are expected to accelerate regional ET and intensify the hydrological cycle, potentially modifying runoff recharge processes and water-resource availability (Gao et al., 2022). Nevertheless, CMIP6-based ET projections for Central Asia remain limited, and large inter-model uncertainties persist, particularly under different Shared Socioeconomic Pathway (SSP) scenarios.
Several critical knowledge gaps therefore remain. First, many existing studies rely on single ET products or raw CMIP outputs, despite known uncertainties of both observations and models in arid climates. Second, systematic evaluation and selection of optimal climate models tailored to Central Asia are still lacking. Third, most studies focus either on historical or future periods, with few providing an integrated, multi-timescale assessment spanning historical, near-term, mid-century, and late-century periods. Fourth, the spatial–temporal heterogeneity of ET responses across seasons, sub-regions, and elevation gradients under different SSP scenarios has not been comprehensively quantified. Finally, the relative contributions of temperature and precipitation to future ET changes in arid Central Asia remain insufficiently constrained.
To address these gaps, this study integrates GLEAM evapotranspiration data with outputs from 23 CMIP6 Earth System Models and applies the Distance between Indices of Simulation and Observation (DISO) metric to identify optimal models for arid Central Asia. Using the selected model ensemble, we (i) evaluate historical ET variability during 1985–2014; (ii) project future ET changes under four SSP scenarios (SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5) across multiple future periods; (iii) examine spatial and seasonal heterogeneity of ET responses across sub-regions and elevation zones; and (iv) quantify the relationships between ET and key climatic drivers, particularly temperature and precipitation. This integrated, model-evaluated assessment provides new insights into hydroclimatic processes in arid Central Asia and supports climate adaptation and sustainable water-resource management in this climate-sensitive region.
Materials and methodsStudy area
The arid regions of Central Asia (ACA) encompasses five Central Asian countries (Kazakhstan, Uzbekistan, Turkmenistan, Kyrgyzstan and Tajikistan) and the arid regions of northwestern China, located between 35°–53°N and 40°–112°E (Supplementary Figure S1), and is one of the largest non-zonal arid regions in the world (Chen et al., 2013; Guo et al., 2021; Guo et al., 2019). The region is characterised by geomorphological types that exhibit an alpine basin structure (Xu et al., 2015). The climate of the study area is predominantly influenced by the westerly circulation and low-latitude circulation system, which is characterised as arid and semi-arid (Hu et al., 2021). The mean annual temperature is approximately 16.7 °C, with an annual precipitation of approximately 131.2 mm (Liu et al., 2023).
Data
The Global Land Evaporation Amsterdam Model (GLEAM) uses a multi-source weighted ensemble precipitation dataset (MSWEP), which provides a distinct advantage in precipitation forcing compared to other evapotranspiration products (Bai and Liu, 2018). Moreover, the efficacy of GLEAM has been augmented through the integration of soil moisture constraints into its intrinsic algorithmic structure and the incorporation of data from microwave remote sensing products (Pan et al., 2020). These factors collectively contribute to the reliability and accuracy of the GLEAM evapotranspiration estimation model. Accordingly, this paper employs the actual evapotranspiration data (ETa) of GLEAM V3.6a (hereinafter referred to as GLEAM, https://www.gleam.eu/) as the evaluation data for CMIP6 evapotranspiration. This data was developed jointly by hydrologists and remote sensing experts from Wageningen University in the Netherlands, with a spatial resolution of 0.25 ° × 0.25 ° (Martens et al., 2017).
The Coupled Model Intercomparison Project (CMIP) was originally conceived as a means of facilitating a comparison of the performance of global coupled climate models. CMIP6 is of particular note for comprising the largest number of models, the most comprehensive scientific experiments, and the most extensive simulation datasets in over 2 decades since CMIP was initiated (Zhou et al., 2019). This paper employs monthly data from 23 Earth System Models (ESMs) included in the CMIP6 model (Supplementary Table S1, https://esgf-node.llnl.gov/projects/cmip6/). The datasets encompasses variables related to evaporation including sublimation and transpiration (evspsbl, [kgm-2s-1]), new surface air temperature (tas, [K]) and precipitation (pr, [kgm-2s-1]). For analysis consistency, all variables were converted to standardized daily units: evspsbl and pr were converted to mm/day (1 kg m-2 s-1 = 86,400 mm/day, accounting for seconds-to-day integration), while tas was converted from Kelvin to degrees Celsius (°C) Seasonal and annual totals (mm/season or mm/year) were derived by accumulating daily values.
The future simulation period is 2021–2,100, and four scenarios (SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5) have been selected for analysis. Different scenarios represent different combinations of shared socioeconomic pathways and radiative forcings, with the reference time period being 1985-2014 (Supplementary Table S2) (Meng et al., 2022; O'Neill et al., 2016; Zhang et al., 2019).
Methods
This study is concerned with the analysis of linear trends, abrupt change characteristics and spatial variation characteristics of ET across a range of time scales in the ACA. The corresponding analysis methods at the different time scales are the linear least squares method and the Mann-Kendall (MK) test. The seasons under consideration in this study are spring (MAM, March to May), summer (JJA, June to August), autumn (SON, September to November), winter (DJF, December to February) and the growing season (April to October). Given the disparate applicability of CMIP6 model data across diverse geographical regions, the data resolution was initially standardised to 0.25 ° × 0.25 ° through the utilisation of the bilinear interpolation method.
Distance between Indices of Simulation and Observation (DISO) is applied to evaluate the overall performance of CMIP 6 models and obtain the five optimal model. DISO has been widely used in diverse research areas (Hu et al., 2022; Zhou et al., 2019; Hu et al., 2021). The details of MK test and DISO can be found in the Text S1.
ResultsSpatial and temporal variations of ET during the historical period of 1985–2014
The history characteristics of ET are obtained from two datasets: GLEAM, and MME data derived from five optimal CMIP 6 models. The optimal CMIP 6 models are selected by the DISO against GLEAM. The evaluating results of CMIP 6 models are found in the Text S2.
In the following sections, we focus on the temporal and spatial variations of ET from GLEAM and MME during 1985–2014. ETa represents the ET from GLEAM data.
Temporal variations of ET during the historical period of 1985–2014
Between 1985 and 2014, ETa and MME exhibited an increase (Supplementary Table S3; Supplementary Figure S2a). A statistically significant increase was observed in both ETa (0.43 mm/a) and MME (0.98 mm/a; p < 0.05).
The MK trend analysis (Supplementary Table S4; Supplementary Figure S3a,b) indicates that the years of shift for ETa were 2002, 2004 and 2014, while the shift point year for MME was 2002. During the growing season (Supplementary Table S3; Supplementary Figure S2b), the mean trend change rate of ETa and MME was 0.19 mm/a and 0.78 mm/a, respectively. The shuift years for ETa and MME during the growing season are consistent with the annual shift point years.
In general, ETa and MME exhibited an upward trajectory across all seasons. However, the growth rates exhibited variability between time scales (Supplementary Table S3; Supplementary Figure S2c–f). The highest rate of increase for ETa was observed in spring, at 0.2 mm/a, with slower growth rates of 0.06 mm/a and 0.05 mm/a in autumn and winter, respectively. A rate of 0.11 mm/a was observed during the summer period. MME demonstrated considerable growth, with rates of 0.43 mm/a and 0.32 mm/a in spring and summer, respectively, and 0.14 mm/a and 0.12 mm/a in autumn and winter. The MK shift trend (Supplementary Figure S3c–l) indicates that ETa was frequently mutated in all seasons. MME exhibited comparable patterns, with the exception of summer, during which the shift point year was 2000. The majority of shifts occurred between the years 2003 and 2011.
Spatial variations of ET during the history period of 1985–2014
The spatial distribution characteristics of ET in Central Asia between 1985 and 2014 were determined using the linear least squares method, as illustrated in Figures 1a,b. In general, ETa and MME demonstrated an increase across the majority of the ACA. Areas exhibiting a notable rise in ETa constituted 24% of the total area and were concentrated in the Junggar Basin and the Tianshan Mountains. The southern TianShan Mountains exhibited the rapid rate of change. The area in which ETa exhibited a decline constituted 4% of the total area and was situated in the northern Caspian coastal lowlands of Kazakhstan and the southern periphery of the Moincombe Desert. However, there was a 76% increase in MME, which was primarily concentrated in northern and southern Kazakhstan. During the growing season (Figures 1c,d), the spatial distribution of ETa and MME exhibited a similar pattern to those of the annual spatial distribution, with ET increasing in the majority of areas.
Spatial trends and significance tests (p < 0.05, dotted areas) of ET in the ACA from 1985 to 2014. (a,b) show annual changes in GLEAM (Eta) and CMIP6 (MME), respectively; (c,d) show growing season changes; (e,f) show spring changes; (g,h) show summer changes; (i,j) show autumn changes; and (k,l) show winter changes.
Twelve-panel scientific figure displaying spatial maps of a geographic region, arranged in six rows and two columns, each panel labeled from a to l. Panels visualize blue-to-red color gradients corresponding to a vertical color bar scale ranging from negative two to positive two, indicating different climate or environmental data values across the region. Western and mountainous areas show varied blue or red shades in some panels, suggesting significant spatial variability, while several panels depict mostly red or brown tones. Latitude and longitude axes are displayed, with consistent map outlines across all panels. No specific data sources or variable names are indicated.
Regional seasonal variations in ET were observed to manifest differently in ETa and MME (Figures 1e–l). The most significant increase in ETa occurred during the winter period with a maximum of 26%, followed by spring and then summer and autumn. The most notable declines were observed in the Moinkum and Saryesik Atyrau deserts in Kazakhstan during the spring season. The summer decrease was primarily concentrated in the southern edge of the Moinkum Desert in Kazakhstan, the northern edge of the lowlands along the Caspian Sea coast, and the northern part of the Turgai Plateau. With regard to the MME, the season exhibiting the most pronounced increase is spring, followed by winter, with comparatively minor increases observed in summer and autumn. The regions exhibiting a decline in summer are predominantly situated in central Xinjiang, while those displaying a reduction in autumn are primarily located in northwestern Kazakhstan.
Future changes of ET against the historical period (1985–2014)
Figure 2 illustrates that the increase in ET was not significant in the near, mid, long and late term in the SSP1-2.6 scenario. However, at the late-term of the SSP5-8.5 scenario, the alterations in ET were particularly noteworthy. The expanded area was situated primarily in the vicinity of the Tianshan and Kunlun Mountains, while the diminished area was predominantly located in the southeastern region of Turkmenistan.
Changes in annual ET in the ACA in the near, mid, long and late term relative to the historical period (1985–2014) under different SSP scenarios. (a–d) show the changes of SSP1-2.6 (near-term (2021–2040), mid-term (2041–2060), long-term (2061–2080) and late-term (2081–2,100)). (e–h) show the changes in SSP2-4.5; (i–l) show the changes in SSP3-7.0; (m–p) show the changes for SSP5-8.5.
Grid of sixteen color-coded maps labeled (a) through (p) shows climate or spatial data for a region spanning approximately 35 to 55 degrees north latitude and 50 to 95 degrees east longitude. Each panel displays variations in values, primarily in shades of red and blue, with a color bar on the right ranging from minus ten to one hundred twenty. Main features include consistent regional patterns and some localized blue and red highlights, suggesting temporal or scenario-based comparisons across the panels.Temporal characteristics of ET in the future under different scenarios
The temporal changes in scenarios SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5 were observed during the study period, which spanned from 2021 to 2,100. Figure 3A and Table 1 illustrate that future annual ET in the ACA across the four scenarios will exhibit considerable increases at rates of 0.42 mm/a, 0.71 mm/a and 0.79 mm/a in SSP2-4.5, SSP3-7.0 and SSP5-8.5, respectively (p < 0.05). With regard to the SSP1-2.6 scenario, the average rate of trend change was determined to be 0.06 mm/a.
The trends of ET in the ACA from 2021 to 2,100 under different SSP scenarios. (A–F) show the annual, growing season, spring, summer, autumn and winter, respectively. The blue line indicates SSP1-2.6; the green line indicates SSP2-4.5; the yellow line indicates SSP3-7.0; the red line indicates SSP5-8.5.
Six line graphs show projected changes in evapotranspiration (ET) in millimeters from 2020 to 2100 under four SSP climate scenarios, with panels for annual (ANN), growing season (GRO), spring (MAM), summer (JJA), fall (SON), and winter (DJF). Each panel compares blue, green, yellow, and red lines representing SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5, respectively. Higher SSP numbers generally show greater increases in ET over time, especially in panels ANN, MAM, SON, and DJF.
The rates and significance of ET in the ACA under different scenarios from 2021 to 2,100.
Time
SSP1-2.6
SSP2-4.5
SSP3-7.0
SSP5-8.5
ANN
0.06
0.42*
0.71*
0.79*
GRO
0.02
0.26*
0.44*
0.4*
MAM
0.02
0.19*
0.35*
0.37*
JJA
−0.01
0.09*
0.13*
0.11*
SON
0.02*
0.07*
0.1*
0.11*
DJF
0.02*
0.08*
0.12*
0.2*
* (p < 0.05).
During the growing season (Figure 3B; Table 1), the trends in ET under the four scenarios were similar to those observed between annual periods. An increasing trend was under the SSP1-2.6 scenario, while a significant increasing trend was observed under the SSP2-4.5, SSP3-7.0 and SSP5-8.5 scenarios (p < 0.05).
Figures 3C–F and Table 1 illustrate the characteristics of changes in ET across the four seasons. Overall, the four scenarios show a general trend of increasing ET. The spring ET evinces a resemblance to the annual ET, exhibiting a pronounced upward trajectory across the SSP2-4.5, SSP3-7.0, and SSP5-8.5 scenarios (p < 0.05). In contrast, the ET in summer exhibits a declining trend under the SSP1-2.6 scenario, while under the SSP2-4.5, SSP3-7.0 and SSP5-8.5 scenarios, it displays a notable upward trajectory (p < 0.05). The ET in autumn and winter exhibits a pronounced upward trend under the four scenarios. Moreover, the increasing trend of the ET is more pronounced with the increase in scenario, with the most pronounced upward trajectory evident under the SSP5-8.5 scenario.
Spatial characteristics of ET in the future under different scenarios
The linear least squares method was employed to analyse the trends in ET in the period between 2021 and 2,100 under different scenarios. The MK test was employed to ascertain the statistical significance of these trends at the 0.05 level of significance. In order to facilitate a more detailed analysis and comparison of the spatial distribution of ET across different future periods, this study employs the 1985-2014 as the reference period. The difference is calculated by subtracting the average of the base period data from the average of the 20-year data of different periods: near-term (2021–2040), mid-term (2041–2060), long-term (2061–2080) and late-term (2081–2,100).
Figure 4 illustrates that, on an annual basis, the SSP1-2.6 scenario exhibits considerable regional disparities in ET when compared to the other three scenarios. In the SSP1-2.6 scenario, the ET demonstrated a decrease in central and western Kazakhstan, whereas in the other three scenarios, it exhibited an increase. In the SSP2-4.5, SSP3-7.0 and SSP5-8.5 scenarios, the ET increased significantly in more than 85% of the regions, mainly in northern Kazakhstan, the border between Kyrgyzstan and Tajikistan and the Kunlun Mountains in Xinjiang.
Spatial variation trends and significance tests (p < 0.05, dotted areas) of annual ET in the ACA from 2021 to 2,100 under different scenarios. (a–d) show SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5, respectively.
Four-panel map graphic showing slope changes in millimeters per annum across a geographic region. Panel (a) contains blue and light brown areas indicating decreasing and slightly increasing slopes, respectively. Panels (b), (c), and (d) show widespread gradients of darker red shades, indicating increasing slopes, with limited areas of negative trend, mapped against latitude and longitude gridlines. A vertical color bar legend on the right represents slope values ranging from negative 0.15 in blue to positive 1.95 in red.
During the growing season (Figure 5), in the SSP1-2.6 scenario, the ET reduction in the ACA was 48%, with the majority of this occurring in Kazakhstan, particularly in the north-west. In the SSP5-8.5 scenario, the proportion of ET reduction area decreases to 15%, with the significant reduction area located mainly on the border between Uzbekistan and Tajikistan. Conversely, the significant area increase is concentrated in the Kunlun Mountains.
Spatial variation trends and significance tests (p < 0.05, dotted areas) of growing season ET in the ACA from 2021 to 2,100 under different SSP scenarios. (a–d) show SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5, respectively.
Four-panel map graphic presents spatial data across Central Asia using a color scale from blue to red, with each panel labeled a, b, c, and d. Blue hues represent lower values while red indicates higher values, with a key scale on the right ranging from negative zero point three to one point nine five. Panels show varying regional patterns, suggesting differences by scenario, season, or variable, with most red appearing in the southern and eastern regions in panels b, c, and d, and more blue tones concentrated in western and northern areas, especially in panel a. Borders and latitudinal and longitudinal lines are included for geographic reference.
Figure 6 illustrates the spatial distribution of seasonal ET under four different scenarios. Overall, there were notable alterations in the spring and summer seasons as the scenarios increased. For the spring season (Figures 6a–d), the SSP1-2.6 scenario indicated a decrease in ET in southern Kazakhstan and the Tarim Basin in Xinjiang. Conversely, the SSP5-8.5 scenario indicated an increase in ET.
Spatial variation trends and significance tests (p < 0.05, dotted areas) of seasonal ET in the ACA from 2021 to 2,100 under different SSP scenarios (a–d) show changes in the ACA in spring (SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5); (e–h) show changes in summer; (i–l) show changes in autumn; (m–p) show changes in winter.
Sixteen-panel grid of color-coded contour maps showing slope trends in millimeters per year across a geographic region, with latitude and longitude axes. Panels are labeled a through p. A red-blue diverging scale is provided, indicating positive slopes in red up to zero point eight and negative slopes in blue down to negative zero point two. Individual maps depict spatial variability and distribution patterns, highlighting areas of increase and decrease in slope values.
During the summer season (Figures 6e–h), under SSP1-2.6 scenario, the area with reduced ET accounts for 62%, mainly distributed in Kazakhstan and northern Xinjiang, among which the reduction in northern Kazakhstan is most significant. In contrast, the SSP5-8.5 scenario exhibited a comparatively modest decline in ET, with a reduction of 29%. The observed decrease was concentrated in northwestern Kazakhstan and western Tajikistan, while significant increases were observed in eastern Tajikistan and the Kunlun Mountains.
In autumn (Figures 6i–l), the area of increase in the ET under the SSP1-2.6 scenario accounts for about 74%, while this proportion increases to 84% under the SSP5-8.5 scenario. The regions exhibiting the most pronounced alterations are situated in the vicinity of the Aral Sea and the Kunlun Mountains. During the winter season (Figures 6m–p), as the scenario gradually increases, the area of Kazakhstan with an increase in ET gradually expands from south to north.
The relationship between air temperature, precipitation and ET
Climate change represents a significant environmental factor that influences regional hydrothermal distribution. The most evident factors are temperature and precipitation. The expression of air temperature is a determining factor in the occurrence of weather change (Lin et al., 2023), while precipitation affects surface runoff and determines the rate of regional water circulation (Ye et al., 2018). Accordingly, temperature and precipitation were identified as the primary climate variables influencing ET in the ACA and were thus selected for correlation analysis.
Relationship between ET and temperature
Table 2 illustrates the correlation coefficients of ET with air temperature and precipitation for different scenarios. The projected correlation coefficient between ET and air temperature will reach approximately 0.7 during the period from 2021 to 2,100 in the SSP1-2.6 and SSP2-4.5 scenarios, and 0.66 in the SSP3-7.0 and SSP5-8.5 scenarios. Figure 7 illustrates the spatial distribution of the correlation coefficients between ET and temperature from 2021 to 2,100 in the four scenarios, as well as the area identified as passing the 95% significance test. As illustrated in Figures 7a–d, a notable positive correlation coefficient of 0.5–0.7 is evident in Kyrgyzstan, Tajikistan and southern Xinjiang. Furthermore, a pronounced negative correlation with air temperature is observed in Uzbekistan and Turkmenistan, with a correlation coefficient of approximately −0.6.
Correlation coefficients between temperature, precipitation and ET in different scenarios for the period 2021-2,100 (p < 0.05).
Period
Variable
SSP1-2.6
SSP2-4.5
SSP3-7.0
SSP5-8.5
2021–2,100
E-T
0.71
0.70
0.67
0.65
E-P
0.52
0.46
0.46
0.43
Spatial variation of correlation coefficients between ET and temperature in the ACA from 2021 to 2,100 under different SSP scenarios. (a–d) represent SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5.
Set of four color-coded contour maps labeled (a), (b), (c), and (d), each depicting regional data across a geographic area with a consistent color legend ranging from negative 0.5 (blue) to positive 0.75 (yellow). Latitude, longitude, and value scales are marked; spatial patterns and distributions are similar but show slight variations across panels.The relationship between ET and precipitation
As illustrated in Table 2 and Figures 8a–d, the correlation coefficients between ET and precipitation in the 2021-2,100 period can reach 0.52 in the SSP1-2.6 scenarios and approximately 0.45 in the SSP2-4.5, SSP3-7.0 and SSP5-8.5 scenarios. In four different scenarios, the future changes in ET and precipitation in regions such as the arid regions of northwestern China and Turkmenistan exhibit a notable positive correlation, with correlation coefficients ranging from 0.4 to 0.6. In eastern Tajikistan and northwestern Turkmenistan, ET and precipitation exhibit a negative correlation, with a correlation coefficient of approximately −0.5. In general, changes in ET are more closely related to changes in air temperature than to precipitation. It can be observed that temperature changes exert a greater influence on ET than precipitation changes.
Spatial variation of correlation coefficients between ET and precipitation in the ACA from 2021 to 2,100 under different SSP scenarios. (a–d) represent SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5, respectively.
Grouped data visualization showing four color-coded contour maps labeled a, b, c, and d, each depicting spatial data over a region with longitude from forty-five to one hundred-five degrees east and latitude from thirty-five to fifty-five degrees north, using a color bar scale from negative zero point five to zero point seven five.DiscussionNext-generation ET monitoring with NISAR soil moisture data
The upcoming NISAR mission will provide 200-m resolution soil moisture products that are expected to transform evapotranspiration (ET) monitoring in Central Asia (Lal et al., 2025). Validation results, with an unbiased root mean square error (ubRMSE) of less than 0.06 m3/m3, indicate consistent performance across varied landscapes, thereby enabling greater precision in ET assessments. These data will facilitate improved identification of moisture-limited and energy-limited ET regimes, reduction of bias in land surface models through data assimilation, and validation of Coupled Model Intercomparison Project Phase 6 (CMIP6) ET projections. The high-resolution moisture maps will support applications such as precision irrigation management, drought early warning systems, and ecosystem health monitoring, all of which are critical for sustainable water resource management in this arid region. Integrating these data with thermal datasets (for example, ECOSTRESS) and optical datasets (such as Sentinel-2) is anticipated to further enhance ET partitioning at the field scale.
Uncertainty analysis of ET in the future under different scenarios
Figure 9 presents the spatial distribution of evapotranspiration (ET) uncertainty across arid Central Asia (ACA) from 2021 to 2,100 under four SSP scenarios (SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5). Here, uncertainty is quantified as the inter-model standard deviation of annual mean ET derived from the selected CMIP6 model ensemble. This metric reflects the degree of agreement among models, with higher values indicating greater uncertainty and lower confidence in ET projections. ET uncertainty exhibits a clear increasing trend over time, with substantially larger standard deviations under higher-emission scenarios. Pronounced spatial heterogeneity is evident, and the magnitude of uncertainty varies considerably among regions, reaching values exceeding 50 mm under high-emission conditions.
Spatial variation of ET standard deviation in the ACA from 2021 to 2,100 under different SSP scenarios. (a–d) show the standard deviation for SSP1-2.6 in the near-term (2021–2040), mid-term (2041–2060), long-term (2061–2080), and late-term (2081–2,100), respectively. (e–h) show the standard deviation for SSP2-4.5; (i–l) represent SSP3-7.0; (m–p) represent SSP5-8.5.
Sixteen-panel graphic displays maps labeled from a to p, showing spatial distributions of standard deviation in millimeters for a certain variable across a region between 45°N–55°N latitude and 45°E–105°E longitude. Color bar on the right ranges from 5 millimeters (dark blue) to 50 millimeters (dark red), indicating lower to higher standard deviation. Each map uses the same color scale to depict spatial variability, with concentrated red spots indicating areas of highest variability. Panels are arranged four by four, facilitating comparison between scenarios.
Under the low-emission SSP1-2.6 scenario (Figures 9a–d), ET uncertainty remains relatively low throughout the projection period. Standard deviation values are generally below 10 mm across most of the ACA, indicating strong inter-model agreement and stable ET projections. The eastern plains, including the Tarim Basin, exhibit minimal variability, whereas mountainous regions such as the Tianshan and Pamir ranges show moderately higher uncertainty, with standard deviations of approximately 15–20 mm. These results suggest that under stringent emission mitigation, ET behaves as a relatively predictable hydrological variable with limited spatial variability.
In the intermediate-emission SSP2-4.5 scenario (Figures 9e–h), ET uncertainty increases gradually over time. The most notable increases occur in central mountainous regions during the mid-term (2041–2060) and long-term (2061–2080) periods. Although uncertainty remains moderate, the upward trend indicates increasing divergence among model projections as radiative forcing intensifies.
The SSP3-7.0 scenario shows a pronounced escalation in ET uncertainty, particularly during the long-term and late-term periods (2061–2,100; Figures 9i–l). Standard deviation values reach 35–40 mm in northern Kazakhstan and southeastern Uzbekistan, while elevated uncertainty expands into southern and northeastern Kazakhstan. This spatial expansion suggests that regions previously characterized by relatively stable ET conditions become increasingly sensitive under intermediate–high emission pathways.
The highest ET uncertainty occurs under the SSP5-8.5 scenario (Figures 9m–p), especially during the late-term period (2081–2,100). Standard deviation values exceed 50 mm in northern Kazakhstan and southern Turkmenistan, indicating substantial inter-model disagreement. The widespread and elevated uncertainty under this scenario highlights the strong influence of intensified warming on ET projections, particularly in regions with complex terrain and strong climatic gradients.
From a regional perspective, mountainous areas in central and western ACA consistently exhibit higher uncertainty than the eastern plains across all scenarios and time periods. This contrast is most pronounced under SSP3-7.0 and SSP5-8.5 (Figures 9i–p), where standard deviation values reach up to 50 mm. The enhanced uncertainty in mountainous regions likely reflects their heightened sensitivity to temperature and precipitation variability, as well as differences in model representations of topography, snow processes, and land–atmosphere interactions. In contrast, the eastern plains, such as the Tarim Basin, display comparatively low variability, with standard deviations generally below 15 mm under SSP1-2.6 and SSP2-4.5 (Figures 9a–h), indicating a more stable hydrological response under lower-emission scenarios.
In summary, future ET uncertainty in ACA increases markedly with both time and emission intensity, with the largest uncertainties occurring under high-emission scenarios and in regions with complex terrain. Conversely, lower-emission pathways are associated with more stable and reliable ET projections. The pronounced spatial heterogeneity of ET uncertainty underscores the importance of emission mitigation for reducing hydrological uncertainty and highlights critical regions where adaptive water-resource management will be most necessary in arid Central Asia.
Impacts of climate factors on spatial and temporal ET variability
This study investigated the spatial and temporal variability of future evapotranspiration (ET) in Central Asia using projections from 23 CMIP6 climate models, with a particular focus on its relationships with temperature and precipitation under different SSP scenarios. The aim was to clarify how climate change may alter ET dynamics in this arid region and to provide scientific support for future water-resource management.
Overall, both annual and seasonal ET are projected to increase under most SSP scenarios, with the strongest enhancement occurring under the high-emission SSP5-8.5 scenario. The only exception is summer ET under SSP1-2.6, where increases are weak or absent. These trends are primarily driven by rising temperatures and changes in precipitation regimes across Central Asia. Elevated temperatures enhance atmospheric evaporative demand, while precipitation modulates the availability of soil moisture necessary to sustain ET. Similar responses have been reported in Mediterranean and other dryland regions, where warming has increased ET by 15%–25%, often exacerbating regional water stress (Avanzi et al., 2020).
Despite the overall positive relationship between ET and temperature or precipitation, anomalous correlations are evident in certain regions and seasons. In moisture-limited environments, higher temperatures do not necessarily lead to increased ET. Instead, warming can intensify soil moisture depletion, resulting in reduced actual ET and even negative correlations with temperature. This phenomenon is particularly apparent during summer, when precipitation is limited and vegetation experiences water stress. Under such conditions, ET becomes constrained by soil moisture availability rather than energy supply, a characteristic feature of arid and semi-arid systems.
Negative or weak correlations between ET and precipitation are also observed in specific contexts. For example, short-duration or highly variable precipitation events may not effectively replenish soil moisture, especially in areas with high evaporative demand or shallow soils. In mountainous regions, snowfall-dominated precipitation and delayed snowmelt can further decouple seasonal precipitation totals from ET responses. These mechanisms explain why increased precipitation does not always translate into higher ET and highlight the complexity of land–atmosphere interactions in Central Asia.
The spatial heterogeneity of ET responses in this study is consistent with findings from other arid and high-altitude regions. Globally, seasonal ET variability of approximately 10%–15% has been closely linked to precipitation and terrestrial water storage changes (Xiong and Abhishek, 2023). On the Tibetan Plateau, ET is primarily precipitation-controlled, ranging from about 200 mm in arid western areas to more than 500 mm in the southeastern region (Ma and Zhang, 2022). Central Asia exhibits comparable spatial contrasts, particularly under SSP5-8.5, where ET increases of 5%–10% are projected in several sub-regions, further complicating water-resource management (Yang et al., 2020).
Impact of ET on zoonics over central asia from one health perspective
In addition to hydrological and climatic implications, changes in evapotranspiration (ET) can exert significant influence on ecological systems and human–animal health interactions (San-José et al., 2023). The One Health framework underscores the interconnectedness of human, animal, and environmental health, emphasizing that zoonotic diseases arise from complex interactions among these systems rather than from isolated factors (Waltner-Toews, 2017). In arid and semi-arid regions such as Xinjiang and Central Asia, environmental processes are fundamental in shaping host habitats, pathogen survival conditions, and exposure pathways, thereby making environmental drivers essential for understanding zoonotic disease ecology (Giraudoux et al., 2013; Zhang et al., 2024).
Evapotranspiration integrates land surface water availability, vegetation dynamics, and climatic conditions, thereby reflecting variations in eco-hydrological processes at regional scales. In Xinjiang and Central Asia, spatial differences in ET often correspond to ecosystem types and habitat conditions, which may indirectly influence zoonotic disease dynamics by affecting intermediate hosts and pathogen viability (Yuan and Bai, 2018; Zou et al., 2020). For instance, ET-driven vegetation productivity can determine the distribution and activity patterns of rodent hosts, potentially altering contact rates between hosts and humans or livestock (Zhang et al., 2017; Marston et al., 2023). ET is also associated with near-surface temperature and humidity, both of which influence the persistence of certain pathogens in the environment. Climatic variables such as temperature and humidity have been shown to affect the transmission patterns of multiple zoonotic diseases (Denissen et al., 2020; Zhang and Wang, 2021). Nevertheless, ET should not be regarded as a simple causal factor; its effects are more likely realized through interactions with other meteorological drivers, representing the combined impacts of multiple environmental processes (Wang and Zheng, 2022; Cui et al., 2024).
Ongoing climate change is expected to further alter water and energy budgets in Xinjiang and Central Asia, which may in turn modify regional ET patterns (Fallah et al., 2024). These changes could influence ecosystem structure, host habitat suitability, and pathogen survival conditions, thereby reshaping landscape-level zoonotic disease risk (Gibb et al., 2020). Previous reviews have highlighted the significance of environmental change in modifying host, vector, and pathogen dynamics, with global warming and climatic variations associated with shifts in the distribution and incidence of zoonotic diseases (Bartlow et al., 2019; Caminade et al., 2019). Within the One Health framework, future research should incorporate ET into multi-factor analytical models, integrating high-resolution environmental data, host distribution, and epidemiological records to more accurately assess disease risk patterns under different climate scenarios. Scenario-based modeling that includes environmental indicators such as ET may enhance understanding of the interactions between environmental change and host–pathogen systems, supporting more comprehensive risk assessment and coordinated prevention strategies.
Conclusion
Historical and future changes in evapotranspiration (ET) across Arid Central Asia (ACA) were quantified using GLEAM observations and five skill-selected CMIP6 models under four emission scenarios. Historical increases in ET during 1985–2014, most pronounced in spring and concentrated near the Kazakhstan–Xinjiang border and major inland basins, indicate rising atmospheric water demand in key agricultural regions. Projections indicate that ET will continue to rise throughout the twenty-first century, especially under SSP5-8.5, which suggests declining surface-water availability, increased irrigation requirements, and heightened drought risk where precipitation increases do not compensate for warming-driven losses. Spatial heterogeneity is evident, with decreasing growing-season ET in northwestern Kazakhstan, increases along the southern Kunlun Mountains, and wide spread winter enhancement, signaling emerging shifts in seasonal water stress, runoff generation, and snowmelt contributions. Temperature is identified as the primary driver of ET across all scenarios, while the weakening coupling between ET and precipitation under stronger warming indicates that atmospheric demand will increasingly govern land-surface water loss, intensifying seasonal deficits even in areas with moderate rainfall increases. Collectively, these findings suggest that intensifying and seasonally redistributed ET will likely reshape basin-scale water budgets across ACA, highlighting the necessity for adaptive irrigation scheduling, reservoir management, and drought-risk mitigation strategies to maintain water security in this highly vulnerable arid region.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.
Author contributions
ZZ: Data curation, Writing – review and editing, Writing – original draft, Methodology, Investigation. GY: Writing – review and editing, Visualization, Investigation. XN: Writing – review and editing, Validation. QZ: Formal analysis, Writing – review and editing. NS: Writing – review and editing, Validation. YL: Writing – review and editing, Resources, Supervision. ZH: Funding acquisition, Writing – review and editing.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fenvs.2026.1757853/full#supplementary-material
ReferencesAvanziF.RungeeJ.MaurerT.BalesR.MaQ.GlaserS. (2020). Climate elasticity of evapotranspiration shifts the water balance of mediterranean climates during multi-year droughts. Hydrology Earth Syst. Sci.24 (9), 4317–4337. 10.5194/hess-24-4317-2020BaiP.LiuX. (2018). Intercomparison and evaluation of three global high-resolution evapotranspiration products across China. J. Hydrology566, 743–755. 10.1016/j.jhydrol.2018.09.065BartlowA. W.ManoreC.XuC.KaufeldK. A.Del ValleS.ZiemannA. (2019). Forecasting zoonotic infectious disease response to climate change: mosquito vectors and a changing environment. Veterinary Sciences6 (2), 40. 10.3390/vetsci602004031064099CaminadeC.McIntyreK. M.JonesA. E. (2019). Impact of recent and future climate change on vector‐borne diseases. Ann. N. Y. Acad. Sci.1436 (1), 157–173. 10.1111/nyas.1395030120891CaoL.XuC.SuoN.SongL.LeiX. (2023). Future dry-wet climatic characteristics and drought trends over arid central Asia. Front. Earth Sci.11, 1102633. 10.3389/feart.2023.1102633ChenX.JiangF.WangY.LiY.HuR. (2013). Eco-geographical pattern of arid region in central Asia. Arid Zone Res.30 (3), 385–390. 10.13866/j.azr.2013.03.001CuiL.ZhongL.MengJ.AnJ.ZhangC.LiY. (2024). Spatiotemporal evolution features of the 2022 compound hot and drought event over the yangtze river basin. Remote Sens.16 (8), 1367. 10.3390/rs16081367DenissenJ. M.TeulingA. J.ReichsteinM.OrthR. (2020). Critical soil moisture derived from satellite observations over Europe. J. Geophys. Res. Atmos.125 (6), e2019JD031672. 10.1029/2019JD031672FallahB.DidovetsI.RostamiM.HamidiM. (2024). Climate change impacts on central Asia: trends, extremes and future projections. Int. J. Climatol.44 (10), 3191–3213. 10.1002/joc.8519GaoJ.ZhaoY.YaoJ.WangM. (2022). Spatiotemporal evolution of atmospheric moisture cycling elements in arid regions of central Asia under the background of climate change. Arid. Zone Research/Ganhanqu Yanjiu39 (5), 1371–1384. 10.13866/j.azr.2022.05.04GibbR.FranklinosL. H.ReddingD. W.JonesK. E. (2020). Ecosystem perspectives are needed to manage zoonotic risks in a changing climate. BMJ371, m3389. 10.1136/bmj.m338933187958GiraudouxP.RaoulF.AfonsoE. V. E.ZiadinovI.YangY.LiL. I. (2013). Transmission ecosystems of Echinococcus multilocularis in China and central Asia. Parasitology140 (13), 1655–1666. 10.1017/S003118201300064423734823GuoH.BaoA.LiuT.NdayisabaF.JiangL.ZhengG. (2019). Determining variable weights for an optimal scaled drought condition index (OSDCI): evaluation in central Asia. Remote Sens. Environ.231, 111220. 10.1016/j.rse.2019.111220GuoH.BaoA.ChenT.ZhengG.WangY.JiangL. (2021). Assessment of CMIP6 in simulating precipitation over arid central Asia. Atmos. Res.252, 105451. 10.1016/j.atmosres.2021.105451HanC.MaY.WangB.ZhongL.MaW.ChenX. (2021). Long term variations in actual evapotranspiration over the Tibetan Plateau. Earth Syst. Sci. Data Discuss.2021, 1–32. 10.5194/essd-13-3513-2021HaoX.FanX.ZhaoZ.ZhangJ. (2023). Spatiotemporal patterns of evapotranspiration in central Asia from 2000 to 2020. Remote Sens.15 (4), 1150. 10.3390/rs15041150HuZ.ZhangZ.SangY. F.QianJ.FengW.ChenX. (2021). Temporal and spatial variations in the terrestrial water storage across central Asia based on multiple satellite datasets and global hydrological models. J. Hydrology596, 126013. 10.1016/j.jhydrol.2021.126013HuZ. Y.ChenD. L.ChenX. (2022). CCHZ-DISO: a timely new assessment system for data quality or model performance from Da Dao Zhi Jian. Geophysical Research Letters49 (23), e2022GL100681. 10.1029/2022GL100681HuaL.ZhaoT.ZhongL. (2022). Future changes in drought over central Asia under CMIP6 forcing scenarios. J. Hydrology Regional Stud.43, 101191. 10.1016/j.ejrh.2022.101191KimD.HaK. J.YeoJ. H. (2021). New drought projections over east Asia using evapotranspiration deficits from the CMIP6 warming scenarios. Earth's Future9 (6), e2020EF001697. 10.1029/2020EF001697LalP.SinghG.DasN. N.LohmanR. B. (2025). Validation of the NISAR multi-scale soil moisture retrieval algorithm across various spatial resolutions and landcovers using the ALOS-2 SAR data. J. Remote Sens.5, 0729. 10.34133/remotesensing.0729LinJ.ZhouG.ZhangQ.LvX.FengF.ZhaoG. (2023). Variations of air temperature differences at different temporal scales over the Tibetan Plateau since 1961 and their possible causes. Int. J. Climatol.43 (13), 5866–5883. 10.1002/joc.8177LiuY.JiangQ.WangQ.JinY.YueQ.YuJ. (2022). The divergence between potential and actual evapotranspiration: an insight from climate, water, and vegetation change. Sci. Total Environ.807, 150648. 10.1016/j.scitotenv.2021.15064834619219LiuL.PengJ.LiG.GuanJ.HanW.JuX. (2023). Effects of drought and climate factors on vegetation dynamics in central Asia from 1982 to 2020. J. Environ. Manag.328, 116997. 10.1016/j.jenvman.2022.11699736516706MaN.ZhangY. (2022). Increasing Tibetan Plateau terrestrial evapotranspiration primarily driven by precipitation. Agric. For. Meteorology317, 108887. 10.1016/j.agrformet.2022.108887MarstonC.RaoulF.RowlandC.QuéréJ. P.FengX.LinR. (2023). Mapping small mammal optimal habitats using satellite-derived proxy variables and species distribution models. PLoS One18 (8), e0289209. 10.1371/journal.pone.028920937590218MartensB.MirallesD. G.LievensH.Van Der SchalieR.De JeuR. A.Fernández-PrietoD. (2017). GLEAM v3: satellite-based land evaporation and root-zone soil moisture. Geosci. Model Dev.10 (5), 1903–1925. 10.5194/gmd-10-1903-2017MengY.DuanK.ShangW.LiS.XingL.ShiP. (2022). Spatiotemporal variation of land surface temperature over the Tibetan Plateau from 1961 to 2100 based on CMIP6 model data. J. Glaciol. Geocryol.44 (1), 24–33. 10.7522/j.issn.1000-0240.2022.0017O'NeillB. C.TebaldiC.Van VuurenD. P.EyringV.FriedlingsteinP.HurttG. (2016). The scenario model intercomparison project (ScenarioMIP) for CMIP6. Geosci. Model Dev.9 (9), 3461–3482. 10.5194/gmd-9-3461-2016PanS.PanN.TianH.FriedlingsteinP.SitchS.ShiH. (2020). Evaluation of global terrestrial evapotranspiration using state-of-the-art approaches in remote sensing, machine learning and land surface modeling. Hydrology Earth Syst. Sci.24 (3), 1485–1509. 10.5194/hess-24-1485-2020San-JoséA.MayorP.CarvalhoB.El BizriH. R.AntunesA. P.Antunez CorreaM. (2023). Climate determines transmission hotspots of polycystic echinococcosis, a life-threatening zoonotic disease, across pan-amazonia. Proc. Natl. Acad. Sci.120 (33), e2302661120. 10.1073/pnas.230266112037549288SuF.LiuY.ChenL.OrozbaevR.TanL. (2024). Impact of climate change on food security in the Central Asian countries. Sci. China Earth Sci.67 (1), 268–280. 10.1007/s11430-022-1198-4Waltner-ToewsD. (2017). Zoonoses, one health and complexity: wicked problems and constructive conflict. Philosophical Trans. R. Soc. B Biol. Sci.372 (1725), 20160171. 10.1098/rstb.2016.017128584179WangH.ZhengJ. (2022). Assessing the effects of surface conditions on potential evapotranspiration in a humid subtropical region of China. Front. Clim.4, 813787. 10.3389/fclim.2022.813787XiongJ.AbhishekXu L.ChandanpurkarH. A.FamigliettiJ. S.ZhangC.GhiggiG. (2023). ET-WB: Water-balance-based estimations of terrestrial evaporation over global land and major global basins. Earth Syst. Sci. Data15 (10), 4571–4597. 10.5194/essd-15-4571-2023XuL.DuL.YaoH.WangH. (2015). Temporal and spatial variation characteristics and trends of precipitation in arid regions of central Asia. Arid Land Resour. Environ.29 (11), 121–127. 10.13448/j.cnki.jalre.2015.373YangL.FengQ.AdamowskiJ. F.YinZ.WenX.WuM. (2020). Spatio-temporal variation of reference evapotranspiration in northwest China based on CORDEX-EA. Atmos. Res.238, 104868. 10.1016/j.atmosres.2020.104868YeH.ZhangT.YiG.LiJ.BieX.LiuD. (2018). Spatio-temporal characteristics of evapotranspiration and its relationship with climate factors in the source region of the yellow river from 2000 to 2014. Acta Geogr. Sin.73 (11), 2117–2134. 10.11821/dlxb201811006YuanX.BaiJ. (2018). Future projected changes in local evapotranspiration coupled with temperature and precipitation variation. Sustainability10 (9), 3281. 10.3390/su10093281ZhangH.WangL. (2021). Analysis of the variation in potential evapotranspiration and surface wet conditions in the hancang river basin, China. Sci. Rep.11 (1), 8607. 10.1038/s41598-021-88162-233883673ZhangH.SunJ.XiongJ. (2017). Spatial‐temporal patterns and controls of evapotranspiration across the Tibetan Plateau (2000–2012). Adv. Meteorology2017 (1), 7082606–7082612. 10.1155/2017/7082606ZhangL.ChenX.XinX. (2019). Short commentary on CMIP6 scenario model intercomparison project (ScenarioMIP). Adv. Clim. Change Res.15 (5), 519. 10.12006/j.issn.1673-1719.2019.082ZhangY.WuJ.AdiliS.WangS.ZhangH.ShiG. (2024). Prevalence and spatial distribution characteristics of human echinococcosis: a county-level modeling study in southern Xinjiang, China. Heliyon10 (7), e28812. 10.1016/j.heliyon.2024.e2881238596126ZhaoM.LiuY.KoningsA. G. (2022). Evapotranspiration frequently increases during droughts. Nat. Clim. Change12 (11), 1024–1030. 10.1038/s41558-022-01505-3ZhouT.ZouL.ChenX. (2019). Commentary on the coupled model intercomparison project phase 6 (CMIP6). Clim. Change Res.15 (5), 445. 10.12006/j.issn.1673-1719.2019.193ZouJ.DingJ.WelpM.HuangS.LiuB. (2020). Assessing the response of ecosystem water use efficiency to drought during and after drought events across central Asia. Sensors20 (3), 581. 10.3390/s2003058131973086
Edited by:Jianhua Xiao, Chinese Academy of Sciences (CAS), China
Reviewed by:Preet Lal, University of Maryland, United States
Yanfang Sang, Institute of Geographic Sciences and Natural Resources (CAS), China