Runoff Characteristics and Their Response to Meteorological Condition in the Yarlung Zangbo River Basin: Spatial Heterogeneity Due to the Glacier Coverage Difference

Abstract

The Yarlung Zangbo River (YZR) is a sizeable highland river on the Tibetan Plateau, and its runoff process is crucial for understanding regional water resource features and related ecological patterns. However, the runoff characteristics of the YZR Basin (YZRB) remain unclear, especially how it would react to climate change. This study comprehensively analyzed the runoff characteristics of the entire YZRB based on a validated distributed hydrological model (SWAT) coupled with a glacier module (SWAT-glac), identified the runoff components, and explored the climate–discharge relationship, with a particular focus on the relationships between glacier runoff and changes in precipitation and air temperature. The results indicate that the SWAT-glac model, with localized glacier parameters, accurately simulates the runoff processes due to regional differences in meteorological conditions and uneven glacier distribution. Summer runoff dominates the basin, contributing 46.2% to 57.9% of the total, while spring runoff is notably higher in the downstream sections than in other areas. Runoff components vary significantly across river sections; precipitation is the primary contributor to basin-wide runoff (23.4–59.5%), while glacier runoff contribution can reach up to 54.8% in downstream areas. The study found that underlying surface conditions, particularly glacier coverage, significantly influence runoff responses to meteorological changes. The correlation between runoff and precipitation is stronger at stations where rainfall predominates, whereas runoff shows greater sensitivity to air temperature in glacier-covered areas. These findings enhance the understanding of runoff processes in the YZRB and offer valuable insights for the sustainable management of water resources in similar basins under climate change.

1. Introduction

The Yarlung Zangbo River (YZR) is the most important river on the Tibetan Plateau, with highly complex and critical runoff changes [1]. The runoff from the YZR is not only the primary water source for the Tibet region but also plays a crucial role in ensuring water security for the South Asian region [2,3]. However, due to the area’s high altitude and complex terrain, obtaining runoff data is more challenging than in other regions, leading to an inadequate understanding of the regional runoff characteristics in the YZR Basin (YZRB), especially in response to changing climate.

Changes in glacier meltwater are particularly important for understanding the runoff characteristics of the YZRB. The region is densely covered with glaciers and snow, and meltwater from these sources is a critical component of runoff in the YZRB [4,5]. Studies have shown that glacier meltwater is a major contributor to runoff in glacierized watersheds [6,7,8,9,10]. For instance, Sun et al. [11] quantified that the contribution of glacier meltwater to runoff in the YZRB reached 16%. Glacier meltwater not only provides an indispensable source of freshwater for human survival in cold and arid areas [12,13] but also determines critical environmental conditions for terrestrial and aquatic communities [14]. In summer, melting water can significantly regulate river water temperature [15] and maintain habitats for cold-water fish [16]. However, under the influence of ongoing climate change, glaciers are rapidly shrinking [14,17], posing severe effects on water resources and biodiversity [18]. Therefore, assessing the contribution of glacier runoff is vital to accurately evaluating runoff changes and the related ecological impact in glacierized catchments.

Currently, hydrological models are the most frequently used approach to research and project runoff processes, and the influence of glacier and snow melt should be included when simulating runoff over glacierized areas. Generally, research on the contribution of glacier runoff relies on hydrological models, including glacier components or hydrological models integrated with glacier modules. Zhao et al. [19] coupled a glacier module in the VIC (variable infiltration capacity) distributed hydrological model to simulate glacier runoff in the two main tributaries of the Aksu River. The HBV model is also commonly used for glacier melt simulation [20,21]. Lutz et al. [22] and Singh et al. [23] used the SPHY (spatial processes in hydrology) model to quantify the contribution of glacier runoff to total runoff in the Himalayas. Some researchers have integrated glacier modules into the SWAT to simulate the contribution of glacier meltwater to runoff in different regions. For instance, the SWAT-coupled glacier module in the Tianshan Mountains successfully captured the changing characteristics of runoff driven by glacier meltwater, demonstrating its simulation capability in complex cryosphere environments [24,25,26]. Similarly, in the Heihe River Basin, the SWAT model has been used to assess the impact of glacier melt on runoff and its future changes, providing reliable simulation results even in areas with limited observational data [27]. Moreover, the SWAT model accurately simulates the runoff changes in the Yarkant River Basin in the Tarim Basin and effectively quantifies the contribution of glacial meltwater to runoff [28]. Compared to these models, one advantage of the SWAT is that it requires fewer data inputs to achieve robust hydrological predictions [29,30]. Thus, these successful applications highlight the strong potential of the SWAT model for accurate hydrological process simulation, particularly in data-scarce regions such as the YZRB.

Recently, many researchers have focused on studying the runoff components of the YZRB, primarily at the Nuxia (NX) station, which represents the overall runoff from the upper-middle reach of the river [11,31,32,33,34]. However, there are currently significant differences in the contribution of runoff components (especially glacier meltwater) to the NX station; the proportion of glacier runoff ranges from 1.8% to 16%. Furthermore, Xuan et al. [35] used the SWAT model to analyze the contribution of runoff components (rainfall, snowmelt, and groundwater) at the NX station; the contributions of rainfall, snow, and baseflow were 52%, 19%, and 29%, respectively, but without considering glacier meltwater separately. Neglecting the contribution of glacier meltwater to the YZRB runoff would confound the characteristics of runoff changes and their underlying mechanisms. Overall, a comprehensive and detailed examination of the runoff components across the entire YZRB is still lacking.

Hence, based on the basin-wide observation dataset, this study aims to (1) couple a glacier module in the SWAT model, simulate historical runoff by locally processing glacier parameters, analyze the runoff characteristics, and explore the climate–discharge relationship, (2) identify the contribution of runoff components (rainfall, snowmelt, glacier, and groundwater) at nine stations, and (3) evaluate the response of glacier runoff to precipitation and temperature. The results of this study will help better identify the runoff characteristics under regional differences across the entire YZRB, providing a scientific basis for formulating water resource management strategies in the region. Moreover, the methodology and findings of this study can offer valuable insights and references for other glacierized basins worldwide, contributing to the broader understanding of hydrological processes in high-altitude, data-scarce regions.

2. Study Area and Data

2.1. Study Area

The YZR (82°1′–97°1′E, 27°5′–31°2′N) originates from the Jiemayangzong Glacier in the Himalayas and is an international river flowing through China, India, and Bangladesh (Figure 1). The YZR is one of the highest rivers in the world, with an average elevation exceeding 4000 m and an elevation range within the basin (757–7261 m). Its main stream is approximately 1770 km long, and the watershed covers around 0.24 million km², including a total glacier area of about 9470 km² (http://www.ncdc.ac.cn/portal/) (accessed on 9 July 2024), which accounts for 4% of the basin’s total area. The basin is influenced by the South Asian monsoon [36] and the unique plateau geographical environment, resulting in highly uneven precipitation distribution. According to data from the China Meteorological Administration, from 1980 to 2016, the mean annual precipitation at 16 stations in the basin was about 430 mm; the mean annual temperature was 5.2 °C.

2.2. Data and Analysis

The data used in this study include topography (Figure 1), soil type, land use type, meteorological data, runoff, and glaciers. Topography data are used to delineate subbasins and determine the outflow path. The land use type data calculate vegetation growth, water consumption, and surface runoff. The soil data provide the physical and chemical properties of soil in the watershed. The meteorological data include daily minimum and maximum temperature, precipitation, solar radiation, wind speed, and relative humidity, and are used for model calibration and validation and the study of historical hydrometeorological laws. The glacier dataset is derived from the Second Glacier Inventory in China [37]. It has been widely used in glaciated basins [19,28,38,39], and the glacier is defined as a single land use type for subsequent runoff simulation. Table S1 provides detailed information and sources of these data.

This study analyzed the determination coefficients (R²) between runoff and meteorological factors and tested the significance of these correlation coefficients. p < 0.01 indicates that the correlations are significant at 99%.

3. Model Development

3.1. Coupling Glacier Module with the SWAT Model

Considering the contribution of glacier meltwater to runoff in the YZRB, integrating a glacier module into the SWAT model is essential for distinguishing the runoff processes in this region. The Soil and Water Assessment Tool (SWAT) is a semi-distributed hydrological model [40] primarily used to simulate and predict water quality/quantity, soil erosion, sediment yield, and the impacts of land use changes [41], etc. This study used SWAT (2012 version) to simulate the monthly average runoff.

Figure 2 shows the schematic diagram of the glacier module coupled to SWAT. For glacier HRUs, the glacier area is dynamic, experiencing either advance (accumulation exceeds ablation) or retreat (accumulation is less than ablation). When snow covers the glacier surface, the snow will first melt and sublimate, while some snow transforms into ice due to accumulation. For glaciers not covered by snow, they will melt and sublimate directly. A small portion of the water from glacier melt will be refrozen.

A model based on glacier mass balance is often used to understand glacier advance and retreat. This model mainly consists of two components: (1) snow accumulation in the glacier accumulation zone and (2) ice ablation in the glacier ablation zone. In the model, the glacier mass balance equation can be written as [25]:

(1)${\displaystyle \frac{d{W}_g}{dt}}=-(1-f)M-S+F$

where W_g is the depth of water equivalent of ice in mm H₂O, M is the melt rate of ice in mm H₂O day⁻¹, and f is the proportion of meltwater refreezing. It has been noted that a large amount of infiltrated meltwater may be refrozen and retained in the glacier [25,42], with a value of 0.2 observed for Qiyi Glacier in the Qilian Mountains in northwest China [43]. S is sublimation rate of ice in mm H₂O day⁻¹, F is glacier accumulation rate in mm H₂O day⁻¹, and t is the time step in day.

The calculation of the parameters in the glacier module is as follows. The ice melt rate (M) is an essential parameter for evaluating glacier melt runoff, and it is commonly estimated using the degree-day model. This model represents the relationship between the amount of ice and snow melt and the cumulative daily temperature. The specific calculation formula is as follows [44]:

(2)$M=\left\{{}_{0\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\overline{T_{\mathit{air}}} \le { T}_{gmlt}}^{b \cdot (\overline{T_{\mathit{air}}}{-T}_{gmlt})\hspace{1em}\hspace{1em} \overline{T_{\mathit{air}}}{ > T}_{gmlt}}\right.$

where b is the degree-day factor for ice melt in mm day⁻¹ °C, $\overline{T_{\mathit{air}}}$ is the daily average air temperature, and T_gmlt is the threshold value for ice melt, both in °C.

The sinusoidal calculation method for the snowmelt degree-day factor was referenced to represent the seasonal change trend of the ice melt degree-day factor in the SWAT model [45]. The ice melt degree-day factor reaches its maximum value on 21 June and its minimum value on 21 December. Accordingly, the ice melt degree-day factor is calculated as [44]:

(3)$b={\displaystyle \frac{b_{gmlt6}+b_{gmlt12}}{2}}+{\displaystyle \frac{b_{gmlt6}-b_{gmlt12}}{2}}\cdot \sin \left[{\displaystyle \frac{2\pi }{365}}(t-81)\right]$

where b_gmlt6 is the glacier degree-day factor for June 21, and b_gmlt12 is the glacier degree-day factor for 21 December, both in mm day⁻¹ °C⁻¹, and t represents the sequential day of the year.

Glacier sublimation is an important component of regional hydrological processes and energy balance, the net radiation is the primary influencing factor for glacier sublimation. In this study, the glacier sublimation rate (S) is considered to be equal to the product of the glacier sublimation coefficient (s_gla) and potential evapotranspiration (ET_p) (calculated by the Penman-–Monteith equation, which is commonly used to calculate surface potential evapotranspiration) [45,46]. The calculation formulas for glacier sublimation rate and glacier sublimation coefficient are as follows:

(4)$S=s_{gla}\cdot E{T}_p$

(5)$s_{gla}={\displaystyle \frac{s_{gmlt6}+s_{gmlt12}}{2}}+{\displaystyle \frac{s_{gmlt6}-s_{gmlt12}}{2}}\cdot \sin \left[{\displaystyle \frac{2\pi }{365}}(t-81)\right]$

where s_gla is the sublimation/evaporation coefficient for ice or snow, ET_p is the potential evaporation in mm day⁻¹; It is assumed that ice/snow changes seasonally, s_gmlt6 is the sublimation factor for June 21 in mm H₂O day⁻¹ °C⁻¹, s_gmlt12 is the sublimation factor for December 21 in mm H₂O day⁻¹ °C⁻¹, and t is the day of year.

To capture the seasonal and gradual pattern of the accumulation. We adopted the glacier mass accumulation algorithm proposed by [25] for simulating the glacier processes in the Tianshan Mountains. The glacier accumulation rate (F) is considered to be the product of the water equivalent of snow and ice and the glacier accumulation coefficient, and the formula given below:

(6)$F=\beta \cdot W_{\mathrm{s}}$

where W_s is the water equivalent of snow over ice in mm and β is an accumulation coefficient that is assumed to be changing seasonally. The formula is given below:

(7)$\beta ={\beta }_0\left\{1+\sin \left[{\displaystyle \frac{2\pi }{365}}(t-81)\right]\right\}$

where β₀ is a basal accumulation coefficient. The accumulation rate (β) was maximum on June 21 and minimum on December 21.

According to Chen and Ohmura [47], using a volume–area scaling relationship based on the measured geometries of mountain glaciers worldwide can simulate the advance or retreat of glaciers. The specific formula is as follows:

(8)$A_{gla}={{\displaystyle \left(\frac{{{\displaystyle V}}_{gla}}{m}\right)}}^{1/n}$

where A_gla is the glacier surface area (km²), V_gla is the glacier volume (km³), and m and n are constants. Liu et al. [48] used m = 29.8, n = 0.7 in their study on glacier area changes in the Qilian Mountains of northwestern China.

The relationship between glacier volume, glacier area, and glacier mass water equivalent depth can be expressed as:

(9)${{\displaystyle V}}_{gla}={\displaystyle \frac{{{\displaystyle W}}_g\cdot {{\displaystyle A}}_{gla}}{{{\displaystyle \rho }}_i}}$

where W_g is the ice equivalent water depth and ρ_i is the density of ice, typically 900 kg/m³.

3.2. The SWAT-Glac Model Set-Up

The model divided the watershed into several subbasins, and each subbasin contains several hydrological response units (HRUs) with combinations of soil, land use, and slope. This study partitioned the YZRB into 149 subbasins and 3186 HRUs for runoff simulation (Figure 1). We established nine control stations to investigate the runoff characteristics of the entire YZRB. Following the model’s verification sequence from tributaries to the main stream and from upstream to downstream, the control stations are distributed as follows: the tributaries include Rikaze (RKZ) station on the Nianchu River, the Lhasa (LS) station on the Lhasa River, the Gengzhang (GZ) station on the Niyang River, and the Pailong (PL) station on the Palong Zangbu River. The main stream includes Lazi (LZ), Nugesha (NGS), Yangcun (YC), Nuxia (NX), and Dexing (DX) stations (Figure 1). Table S2 lists the information on the subbasins controlled by each station.

The reclassified soil and land use types in the YZRB are shown in Figure 3. LPi and CMi are the significant parts of the soil in all watersheds; FRST, PAST, and SWRN are the main components of land use. The slope of all watersheds is divided into 0–5, 5–10, 10–20, 20–30, and >30. Figure 4 shows that the glaciers are mainly distributed in the upstream and downstream areas in the YZRB, with the most glacier coverage at the PL, GZ, and LZ stations accounting for 56.9%, 12.1%, and 10.2% of the total glacier area, respectively. The runoff of the nine stations is mainly distributed in June–August (summer).

The model’s simulation period was from 2000 to 2016, and from 2000 to 2002 was the warm-up period. Table S3 shows the calibration and validation periods for runoff at each station.

3.3. Model Parameters

The model contains many parameters (>200), and the inclusion of all parameters may not remarkably affect the accuracy of simulation results but would increase computational demands [49]. Through literature investigation and consideration of parameter applicability to the study area, this study selected 40 parameters (Table S4) related to hydrological processes in the YZRB, including six parameters related to glacier melting. To analyze parameter sensitivity and choose a set of sensitive parameters covering most hydrological processes, we used the multivariate regression method (SUFI-2) [50,51].

3.4. Assessment Criteria

The SWAT-glac model was calibrated and validated for runoff using SWAT-CUP [52]. This study evaluated the model performance by using the coefficient of determination (R²), Nash-Sutcliffe efficiency (NSE), and percent bias (PBIAS). With R² and NSE values closer to 1 and PBIAS values closer to 0, it indicated a better model simulation performance. When R² ≥ 0.6, NSE ≥ 0.5, and PBIAS ≤ ±25%, it means good applicability of the model [53].

3.5. Runoff Components

In this study, the SWAT-glac model divided the total simulated runoff into four components: rainfall-induced runoff (PRE), snowmelt-induced runoff (SNO), glacier-induced runoff (GLA), and groundwater-induced runoff (GWQ). The input meteorological data provide the original precipitation, divided into rain and snow by the snowfall temperature parameter (SFTMP) in the SWAT-glac model. Based on this, the model includes initial volumes of rainfall and snowmelt, allowing for determining their respective proportions in the remaining flow [54]. The contribution of glacier melt to runoff can be obtained from the glacier module, while the contribution of groundwater to runoff can be directly obtained from the hydrological model output.

4. Results

4.1. Sensitivity Analysis Results

The sensitivity analysis of hydrological parameters shows that the p-value of sensitive parameters at all stations in the YZRB is less than 0.3 (Figure 5). The sensitivity ranking of parameters is determined based on their p-values, where smaller values indicate greater sensitivity. Among the nine stations, NGS and NX stations have the highest number of sensitive parameters, while PL station has the lowest number.

Across all stations, the sensitive parameters directly affecting the baseflow have the most significant proportion. When analyzing the sensitivity of parameters related to evapotranspiration, the maximum canopy storage (CANMX.hru) is a sensitive parameter at all stations except for the RKZ station. In the process of surface runoff, the parameter “CN2” (Curve number for moisture condition II) is highly sensitive in previous studies [55,56]. In this study, except for RKZ and PL stations, CN2.mgt is considered a highly sensitive parameter at the remaining seven stations.

Glaciers and snow are widely distributed in the YZRB, and the contribution of runoff from snow and glacier melt to total runoff must be addressed [31,57]. Therefore, the parameters related to snow and glaciers are considered sensitive parameters in this study.

4.2. Model Performance

Table 1 provides R², RMSE, and PBIAS evaluation indexes for comparing simulated and observed monthly runoff. Two time periods were extracted from the entire simulation period to compare runoff during the calibration and validation processes (Figure 6).

It is clear from Table 1 that when the model does not include the glacier module, some stations, such as RKZ, LZ, and NX, achieve decent NSE values exceeding 0.65 during both calibration and validation periods. However, the PBIAS values exceed 25%, and the NX station reaches 40.5% during validation. Particularly for regions with significant glacier coverage, such as the GZ station and PL station (Figure 4), the model without the glacier module underestimates runoff at these stations, especially for the simulation of runoff peaks (gray solid circles in Figure 6). The underestimation mainly occurs after April because rising temperatures cause snow and glaciers to melt, rapidly increasing runoff. In contrast, the rise in rainfall does not achieve the same effect. Due to the contribution of glacier runoff, the NSE and PBIAS values are significantly improved for the GZ and PL stations. For the NX and DX stations, the simulation performance of the model is also enhanced due to the supplementation of glacier meltwater from the Niyang River (GZ station) and the Parlung Zangbu River (PL station).

Table 1 shows that coupling the glacier module in the model profoundly improves model performance. This indicates that using a degree-day model to simulate glacier melt processes is reliable. Hence, this glacier model is suitable for simulating runoff in the YZRB.

4.3. Runoff and Corresponding Components

Figure 7 shows the monthly average simulated rainfall, snowmelt, glacier, and groundwater runoff at the nine stations from 2003 to 2016 and their annual contributions to runoff. It can be seen from Figure 7 that runoff from June to September accounts for over 60% of the total runoff in all subbasins of the YZRB. Due to the highest precipitation and temperature during these months, the runoff from glacier melt also reaches its peak. The runoff from snowmelt at the nine stations mainly occurs from April to October, but the months when the snowmelt runoff reaches its maximum value vary among all stations.

There is a noticeable difference in the contribution of the four components to runoff among the nine hydrological stations. Overall, during the study period, PRE contributes the most to runoff in the YZRB, accounting for 23.4% to 59.5% of the runoff at the nine stations, with the highest contribution at the NGS station (59.5%). SNO contributes 0.8% to 5.3% of the runoff at the nine stations, with the highest contribution at the LS station. GWQ contributes 12.4% to 40.3% of the runoff at the nine stations, with the highest contribution at the LZ station (40.3%). GLA contributes 4.5% to 54.8% of the runoff at the nine stations, with the highest contributions at the PL station (54.8%), whose glacier areas accounted for approximately 56.9% of the total glacier area in the YZRB (Figure 4). The results also show that, for the entire basin, rainfall contributes the most to runoff, accounting for 47.7%, while glacier meltwater contributes 22.8%. Particularly in the downstream (from NX to DX station), where extensive glacier distribution exists, the study indicates that glacier runoff can account for up to 36.1% of the total runoff in this region.

It is worth noting that rainfall, snowmelt, and glacier meltwater will infiltrate into the soil profile and contribute to groundwater, which means the contribution of groundwater is overestimated to some degree. In contrast, the contributions of the other three components are underestimated. In addition, the inconsistent calibration results of snow-related parameters will also lead to significant differences in the contribution of SNO to runoff.

4.4. Runoff Process Analysis

4.4.1. Variability in Regional Runoff

We utilized the calibrated and validated SWAT-glac model, simulated the monthly average runoff from 2003 to 2016 for nine stations in the YZRB, and explored the characteristics of river flow changes at nine stations (Table S5). For the four hydrological stations on the tributaries, the highest and lowest annual average simulated runoff depths were at PL station (160.9 mm) and RKZ station (3.9 mm), respectively. Of the five hydrological stations on the main stream, the DX station, which is the total water outlet of the study area, had the highest annual average runoff depth at 444.6 mm, while the lowest average yearly runoff depth was at the most upstream station, LZ station, with 17.6 mm.

From Table S5, the summer runoff in the YZRB is highest compared with other season runoff, with the highest being 57.9% at the YC station and the lowest being 46.2% at the RKZ station. The monthly average simulated runoff values at each station also show different distributions (Figure 8). For the four stations on the tributaries, the maximum monthly average runoff occurs in August at the RKZ station, while for the other three stations, it occurs in July. The maximum monthly average runoff occurs in August for the five stations on the main stream. With the rise in temperature and the increase in precipitation, there is a sharp increase in runoff at each station in April, with the maximum proportion of the monthly average runoff to the annual total runoff in April occurring at the PL station, reaching 7.0%.

In summary, runoff at each station exhibits seasonal and monthly differences, and the underlying characteristics of the basin, including glacier distribution and meteorological factors (temperature and precipitation), also showed prominent regional characteristics. Therefore, the relationship between meteorology and runoff should be further investigated.

4.4.2. Impacts of Precipitation on Runoff

As a fundamental meteorological variable, precipitation notably impacts hydrological processes. Generally, runoff is closely related to rainfall, for the four main stream stations (NGS, YC, NX, and DX) receive inflows from upstream sources. For instance, the amount of water from the Lazi station and Nianchu River affects the runoff at NGS station, while DX station is the total outlet of the study area. Therefore, in the analysis of the correlation between regional runoff and meteorology in this section, we primarily focused on five representative stations, four stations on the tributaries (from top to bottom: RKZ, LS, GZ, and PL), and the upstream LZ station. Figure 9 shows the correlation between precipitation and the simulated runoff at the five stations mentioned above on a monthly scale.

Overall, there is a positive correlation between precipitation and runoff at each station, and the correlation was significant. However, the R² for precipitation and runoff ranges from 0.32 to 0.80. The summer precipitation at Lhasa station (355.8 mm) accounts for approximately 76.2% of the annual precipitation total (466.8 mm). In comparison, the corresponding summer runoff at LS station (18.0 mm) accounts for approximately 55.0% of the average yearly runoff (32.7 mm) (Table S5). The R² between annual precipitation and runoff at this station is 0.80, with a significant correlation (p < 0.01). The summer precipitation at Bomi station (292.7 mm) accounts for approximately 34.1% of the annual precipitation total (857.4 mm), while the summer runoff (78.6 mm) accounts for approximately 48.9% of the average yearly runoff (160.9 mm). The R² between annual precipitation and runoff at this station is 0.32, with a significant correlation (p < 0.01). Additionally, the trends in precipitation and runoff at each station are inconsistent. For example, the precipitation at the Linzhi and Bomi meteorological stations exhibits multimodal patterns, while the runoff at GZ and PL stations peaks in July, with the phase of precipitation and runoff not consistent at the other three stations.

In addition to precipitation, meltwater from snow, glaciers, and frozen soil in the YZRB supplements runoff. Other factors, like air temperature, may influence the relationship between runoff and precipitation. Therefore, further analysis of the impact of temperature on runoff is warranted.

4.4.3. Impacts of Air Temperature on Runoff

Table S5 lists the statistical results of annual and seasonal average maximum and minimum temperatures. For maximum temperatures, the seasonal differences range from 1.9 °C to 2.7 °C. For minimum temperatures, the seasonal differences range from 3.0 °C to 7.3 °C, and the differences in minimum temperatures during autumn and winter seasons are higher than the annual differences.

To investigate the impact of temperature on runoff, we further analyzed the relationships between monthly average runoff and monthly average temperature at five stations. Figure 9 shows that the scatters could be well-fitted by linear regression. The maximum and minimum R² values for the relationship between minimum temperatures and runoff are 0.94 for Bomi station and 0.25 for Rikaze station, respectively, while the maximum and minimum R² values for the relationship between maximum temperatures and runoff are 0.90 for Bomi station and 0.17 for Rikaze station, respectively.

The results show that the maximum and minimum temperatures are generally positively correlated with runoff; higher temperatures typically lead to more significant runoff. Overall, the relationship between minimum temperatures and runoff is closer than that of maximum temperatures. Therefore, temperature is an essential factor that must be considered when studying the runoff characteristics in this region.

5. Discussions

5.1. Localized Model Parameters Suggest the Impact Glaciers on Runoff Processes

During the development of the SWAT-glac model, we noticed that the different glacier parameter setting strategy leads to massive differences in model performance in the YZRB. Table S6 presents the values of glacier parameters for two different settings (global and local) at each station. It can be observed that the differences between the global and local glacier parameter settings mainly occur in regions with abundant glacier coverage, such as at the GZ, PL, LZ, and DX stations (Figure 4). The differences primarily revolve around parameters b_gmlt6 (glacier melt factor on June 21) and b_gmlt12 (glacier melt factor on December 21), with the maximum and minimum values of these two factors being 6.5 and 5.5 mm °C⁻¹ day⁻¹ for the DX station, and 0.8 and 0.4 mm °C⁻¹ day⁻¹ for the LZ station, respectively. The runoff simulation in different glacierized areas can be calibrated by adjusting the glacier melt factor. Notably, the calibration value of this parameter shows considerable differences. Cui et al. [58] found that the annual degree-day factor of the east and west branches of Glacier No. 1 at the headwaters of Ürümqi River in China from 1983–2006 ranged from 1.12–14.30 mm·°C⁻¹·d⁻¹ and 0.98–13.84 mm·mm·°C⁻¹·d⁻¹, respectively. The glacier melt factor in Tian Shan on June 21 and December 21 ranged from 2.1–27.3 mm·°C⁻¹·d⁻¹ and 0.9–10.8 mm·°C⁻¹·d⁻¹, respectively [26]. The glacier melt degree-day factor of the Yarkant River Basin was 5.5 mm·°C⁻¹·d⁻¹ [28]. This indicates that there are significant differences in glacier-related parameters between regions.

For the YZRB, the significant spatial differences in underlying surface conditions make the regionalization of such parameters more critical. In studies on the glacier melt degree-day factor of the YZR, the calibration values within different basins vary between 4.5 and 10.97 mm·°C⁻¹·d⁻¹ [11,19,33]. Figure S1 compares monthly average simulated and observed runoff values for three stations (GZ, PL, and LZ) under different glacier parameter settings in this study. For GZ and PL stations, the model fails to adequately capture the peak runoff at these stations by setting glacier parameters globally; still, the model leads to better simulation of runoff peaks for these two stations by setting glacier parameters locally. For the LZ station, runoff simulation values substantially exceed observed values by setting glacier parameters globally; conversely, the simulated runoff values align well with the observed values by setting glacier parameters locally. Table 2 indicates substantial improvements in these stations’ three performance metrics (R², NSE, and PBIAS).

Hence, for regions with glacier coverage, localizing glacier parameters is effective for simulating runoff in that area, and it holds a reference value for runoff simulation in similar regions. Additionally, the configuration of this parameter suggests that even within the same basin, spatial differences in underlying surface conditions can significantly impact the runoff characteristics of different subbasins.

5.2. Runoff Components of YZRB Under the Impact of Different Glacier Coverage

Previous research on the runoff components of the YZRB primarily focuses on the NX station (the controlled area accounts for more than 80% of the total area of the YZRB), as shown in Table 3, and the results vary among different reports. Zhang et al. [5] used the VIC and degree-day model to simulate the runoff components in the YZRB, where glacier runoff contributed 11.6% to the total runoff. Su et al. [33], Sun and Su [4], Sun et al. [11], and Zhao et al. [19] also used the VIC model coupled with a glacier module to simulate the runoff components in the YZRB, with glacier runoff contributing between 5.5% and 16.0% to the total runoff. The SPHY model, which includes modules for glaciers and snowmelt, has also been utilized to identify runoff contributions. Lutz et al. [22] determined that glacier runoff contributed 16.0% to the runoff in the YZRB, whereas Khanal et al. [32] found a contribution of 1.8% using the same model. Furthermore, Xuan et al. [35] used the SWAT model to study the contributions of three components (rainfall, snowmelt, and groundwater) at the NX and GZ stations in the YZRB.

As the estimation of runoff components in YZR remains uncertain, this study established a model based on detailed data of the underlying surface across the entire basin. It used data from 16 ground meteorological stations to drive the model, resulting in satisfactory runoff simulation outcomes. The runoff component results generally fall within the range of previous results and provide a comprehensive understanding of runoff characteristics for the entire basin. In addition, the study also found that due to the widespread distribution of glaciers in the YZRB, glacier runoff is an essential component of the total runoff of the basin, and the contribution of glacier runoff varies greatly. This is consistent with the conclusions drawn from other glaciated basins, such as Engel et al. [7] found that the maximum contribution of glacier melt was 65% (in August) in the Alps; the glacier melt contributes to runoff of 35–48% in the Sari-Djaz catchment of the Tarim River [6]. Rets et al. [10] quantified that the contribution of glacier meltwater to runoff in the Baksan River of the Caucasus Mountains is no less than 30%. Zhang et al. [26] simulated 24 headwater catchments with varying glacier area ratios in the Eastern and Central Tianshan Mountains. They found that the contribution of glacier meltwater to runoff ranged from 3.5% to 67.5%. Moreover, the contribution of glacier runoff within the entire basin varies significantly, ranging from 4.5% to 54.8%. The period when glacier runoff contribution is the highest is relatively fixed, generally occurring from July to September [10,20], and significantly impacts physical characteristics such as river water temperature [59]. Sun et al. [60] found that the basin hydrological process is a critical factor in maintaining the balance and integrity of basin ecosystems, and the water temperature process plays an essential ecological role in controlling the life and habitat of aquatic species [61]. As a result, this unique runoff characteristic might be one of the driving forces behind the distinctive ecological patterns in the YZRB.

Table 3

Summary table of studies on runoff components in the YZRB. “-” indicates that this runoff component was not considered in the literature.

Region	Relevant Studies	Period	Model	Runoff Contribution (%)
				PRE	SNO	GLA	GWQ
NX	Zhang et al. [5]	1961–1999	VIC+DD	65.4	23	11.6	-
	Lutz et al. [22]	1998–2007	SPHY	59	9	16	16
	Su et al. [33]	1971–2000	VIC+DD	57.7	27.3	15	-
	Chen et al. [31]	2003–2014	CREST	79.5	10.6	9.9	-
	Zhao et al. [19]	1971–2010	VIC+DD	71.4	23.1	5.5	-
	Sun and Su [4]	1980–2000	VIC+DD	62.3	23.8	13.9	-
	Xuan et al. [35]	1993–1999	SWAT	52	19	-	29
	Khanal et al. [32]	1985–2014	SPHY	62.1	13.2	1.8	22.9
	Wang et al. [34]	1981–2019	WEB-DHM	-	16.6–22.3	3.5–7.2	-
	Sun et al. [11]	1971–2020	VIC+DD	63	24	13	-
	This Study	2003–2014	SWAT+DD	53.5	0.8	16.9	28.8
GZ	Xuan et al. [54]	2000–2011	SWAT	64	18	-	18
	This Study	2003–2008	SWAT+DD	27.7	3.8	51.4	17.2
PL	Ban et al. [62]	2004–2018	SPHY	10	25	45	20
	This Study	2011–2016	SWAT+DD	23.4	2.6	54.8	19.2

5.3. Uneven Distribution of Glaciers Changes the Response of Runoff to Meteorological Conditions

The impact of meteorological conditions on river runoff has been widely discussed. In the YZRB, there were essential spatial differences in the relationship between runoff precipitation and air temperature. For instance, although the precipitation during summer at the LS, GZ, and PL stations is nearly the same, the runoff during summer varies greatly. During the summer, the precipitation at the GZ station is even higher than at the PL station, yet the runoff is twice as high at the PL station (Table S5). Additionally, from the correlation results (Figure 9), the correlation between precipitation and runoff is notably weakened in regions of the YZRB where glaciers are more abundant. Still, the correlation between runoff and air temperature is more robust than at other stations. For example, at the Bomi station on the Parlung Tsangpo River, the R² between rainfall and runoff is only 0.32, but the R² between minimum temperature and runoff reaches 0.94. It indicates that variations in glacier distribution within the basin alter the response patterns of runoff to meteorological conditions.

Glaciers are extremely sensitive to meteorological conditions, especially air temperature and precipitation [25,63]. Against this background, to better understand the response mechanism of runoff processes to meteorological conditions, the response of glacier runoff to rainfall and air temperature in the YZRB was analyzed (Figure 10). Overall, a positive correlation exists between glacier runoff and air temperature, indicating that air temperature substantially impacts glaciers. However, spatial heterogeneity was shown in the correlation between glacier runoff and precipitation, as well as air temperature in different regions of the YZRB. In the upstream (LZ station) and midstream (RKZ and LS stations), where runoff is mainly supplied by precipitation, the maximum R² between glacier runoff and precipitation is observed at the LS station (R² = 0.78, p < 0.01). In contrast, for regions where glacier meltwater is the primary source of runoff, such as the GZ and PL stations, the correlation between glacier runoff and temperature is stronger than the correlation between glacier runoff and precipitation. For the PL station, in particular, the glacier runoff is enormously significant and positively correlated with air temperature (p < 0.01) with an R² value of 0.94, while the R² between glacier runoff and precipitation is only 0.32. The uneven distribution of glaciers within the basin leads to significantly different responses of glacier meltwater supply in different river sections during changing meteorological conditions. Chen et al. [55], He et al. [57], and Liu et al. [2] found that the temperature in the YZRB will exhibit a significant upward trend. Under the further influence of climate change, due to glaciers’ extreme sensitivity to temperature changes [63] and the uneven distribution of glaciers, the response of runoff processes could also behave more substantial spatial heterogeneity. Therefore, for basins with glacier distribution like the YZR to further manage their water resources and ecological health, it is crucial to focus on the spatial differences in the runoff processes under climate change.

6. Conclusions

This study assessed the runoff characteristics of the whole YZRB and its response pattern to meteorological variation. By coupling the SWAT model and a glacier module, the runoff processes of the YZRB were accurately simulated, and the runoff components were extracted from the modeled results as well. The summer runoff of YZRB is the largest compared with the runoff in other seasons, ranging from 46.2% to 57.9%; however, the proportion of spring runoff in the downstream river section is higher than in different sections. This is due to the significant regional differences in runoff components; while PRE is the dominant contributor to the runoff of the entire basin (23.4~59.5%), the supply of GLA to runoff in the downstream section can reach up to 54.8%. These runoff characteristics are highly related to the spatial differences in underlying surface conditions within the basin, especially the glacier coverage, and that is why the glacier module parameters should be localized to achieve better model performance. By analyzing the correlation between the runoff and meteorological variables, it was found that the underlying surface could strongly influence the response of runoff processes to meteorological conditions. In general, the above findings can enhance our understanding of the runoff process in the YZRB, as well as provide a valuable reference for sustainable management of water resources in similar river basins under climate change.

Author Contributions

Conceptualization, Y.D. and M.C.; methodology, L.Z., Y.D. and M.C.; software, L.Z., Y.T. (Yi Tan) and G.B.; validation, L.Z.; data curation, X.W. and Y.D.; writing—original draft preparation, L.Z.; writing—review and editing, L.Z., Y.D. and M.C.; supervision, Y.T. (Youcai Tuo); funding acquisition, R.A. All authors have read and agreed to the published version of the manuscript.

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 author.

Conflicts of Interest

The authors declare no conflicts of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1. The distribution of hydrological stations in the Yarlung Zangbo River Basin and the subbasins controlled by each station.

Figure 2. Schematic diagram of glacier module. P represents precipitation, SF represents snowfall, Tav is the daily average temperature, SFTMP is the critical temperature at which snowfall occurs. S is the sublimation rate of ice/snow, M is the melt rate of ice/snow, and F represents the turnover rate of snow to ice. f is the refreezing proportion after ice melting and W represents the water equivalent of ice/snow.

Figure 3. The spatial distributions of soil type (a) and land use (b) over the study area.

Figure 4. Contribution of seasonal runoff at each station. The four boxes in each row represent the contribution of seasonal runoff at the nine stations in the study area.

Figure 5. Monthly timestep parameter sensitivity assessment for nine stations in the YZRB. Yellow indicates higher sensitivity of the parameter, green corresponds to lower sensitivity, and black patches denote parameters considered insensitive at that station and not considered in subsequent model tuning processes.

Figure 6. Monthly observed and simulated runoff trends from 2003 to 2016 for the nine stations in the YZRB. The slope coefficient is provided for both observed runoff (in red) and simulated runoff using the SWAT-glac model (in green). The dashed blue line separates the runoff calibration and validation time periods.

Figure 7. Monthly average simulated rainfall, snowmelt, glacier, and groundwater runoff at each station in the YZRB from 2003 to 2016, along with their contributions to annual runoff.

Figure 8. Distribution of simulated average monthly runoff at each station. The solid black line represents the error bars of average monthly runoff.

Figure 9. Bivariate scatterplot matrix of simulated runoff versus precipitation, maximum and minimum temperatures for the five meteorological stations. The five stations are divided into two graphs to illustrate the content conveyed by the graphs: the left graph compares the Lazi, Rikaze, and Lhasa stations, while the right graph compares the Linzhi and Bomi stations.

Figure 10. Correlation between monthly average simulated glacier runoff and monthly total precipitation and monthly average air temperature at each station. Black circles represent the correlation between glacier runoff and precipitation, while red circles represent the correlation between glacier runoff and monthly average air temperature.

Table 1

The calibration and validation results of monthly runoff in the YZRB. In the column of SWAT-glac, “×” means that the glacier module is not included in the runoff simulation, while “✓“ means the runoff simulation included the glacier module.

HydrologicalStations	SWAT-Glac	Calibration			Validation
		R2	NSE	PBIAS (%)	R2	NSE	PBIAS (%)
RKZ	×	0.81	0.73	28.0	0.89	0.66	21.8
	✓	0.81	0.77	20.2	0.90	0.73	15.5
LS	×	0.79	0.69	17.2	0.70	0.60	18.3
	✓	0.77	0.75	−1.6	0.70	0.68	2.8
GZ	×	0.87	0.43	51.1	0.49	0.16	51.4
	✓	0.89	0.87	−9.6	0.92	0.89	−9.6
PL	×	0.69	−0.11	62.5	0.43	0.01	52.3
	✓	0.89	0.85	−17.9	0.79	0.73	−24.7
LZ	×	0.93	0.81	29.3	0.89	0.62	34.6
	✓	0.95	0.95	3.9	0.94	0.88	17.5
NGS	×	0.98	0.96	14.4	0.77	0.66	19.1
	✓	0.97	0.97	5.9	0.72	0.70	10.7
YC	×	0.91	0.89	17.4	0.96	0.93	19.2
	✓	0.91	0.91	5.1	0.96	0.96	3.9
NX	×	0.94	0.70	40.2	0.91	0.81	40.5
	✓	0.96	0.92	16.6	0.97	0.95	17.4
DX	×	0.90	0.13	59.9	0.87	0.22	54.4
	✓	0.90	0.83	16.7	0.90	0.88	9.7

Table 2

Statistical table of monthly average runoff simulation performance under different glacier parameter settings at each station.

Stations	Parameter SettingMethods	Calibration			Validation
		R2	NSE	PBIAS (%)	R2	NSE	PBIAS (%)
GZ	Global	0.89	0.67	25.9	0.93	0.69	25.1
	Local	0.89	0.87	−6.6	0.91	0.88	−13.5
PL	Global	0.86	0.82	0.7	0.78	0.73	−4.0
	Local	0.89	0.86	−15.6	0.82	0.76	−24.1
LZ	Global	0.67	0.61	−20.6	0.83	0.73	−14.7
	Local	0.95	0.95	1.3	0.93	0.88	16.8

Supplementary Materials

The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/rs16244646/s1, Table S1: Description of input data used in setting up a SWAT model; Table S2: Information on the subbasins controlled by each hydrological station; Table S3: The calibration and validation periods for runoff at each station. (2010) means that the runoff for 2010 is missing at the LS station; Table S4: Descriptions and initial ranges of the parameters used for the SWAT-glac model cali-bration. The fourth column indicates the type of modification to be made to the existing param-eter value: “r” means multiplying the original value by the adjustment factor (add 1 to a given value in the range), and “v” means replacing the original value with a value in the range; Table S5: Statistical results of temperature (°C), precipitation (P), and simulated river discharge (R) at nine stations in the YZRB. Parentheses in the Stations column denote the meteorological station near the hydrological station, such as the meteorological station near the RKZ hydrological station which is the Rikaze station; Table S6: Values of glacier parameters under two different settings at each station. “Global” in-dicates uniform glacier parameter values across the entire basin, while “Local” signifies different glacier parameter values set at various stations. Figure S1: Violin plots of monthly average simulated and measured runoff at three stations (GZ, PL, LZ) under different glacier parameter settings. The red represents observed monthly average runoff; the green represents simulated results under global glacier parameterization; the yellow represents simulated results under local glacier parameterization.

References

1. Zhu, Y.; Sang, Y.F.; Wang, B.; Lutz, A.; Hu, S.; Chen, D.; Singh, V.P. Heterogeneity in Spatiotemporal Variability of High Mountain Asia’s Runoff and Its Underlying Mechanisms. Water Resour. Res.; 2023; 59, e2022WR032721. [DOI: https://dx.doi.org/10.1029/2022WR032721]

2. Liu, S.; Yao, Y.; Kuang, X.; Zheng, C. A preliminary investigation on the climate-discharge relationship in the upper region of the Yarlung Zangbo River basin. J. Hydrol.; 2021; 603, 127066. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2021.127066]

3. Pritchard, H.D. Asia’s shrinking glaciers protect large populations from drought stress. Nature; 2019; 569, pp. 649-654. [DOI: https://dx.doi.org/10.1038/s41586-019-1240-1]

4. Sun, H.; Su, F. Precipitation correction and reconstruction for streamflow simulation based on 262 rain gauges in the upper Brahmaputra of southern Tibetan Plateau. J. Hydrol.; 2020; 590, 125484. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2020.125484]

5. Zhang, L.; Su, F.; Yang, D.; Hao, Z.; Tong, K. Discharge regime and simulation for the upstream of major rivers over Tibetan Plateau. J. Geophys. Res. Atmos.; 2013; 118, pp. 8500-8518. [DOI: https://dx.doi.org/10.1002/jgrd.50665]

6. Duethmann, D.; Bolch, T.; Farinotti, D.; Kriegel, D.; Vorogushyn, S.; Merz, B.; Pieczonka, T.; Jiang, T.; Su, B.; Güntner, A. Attribution of streamflow trends in snow and glacier melt-dominated catchments of the Tarim River, Central Asia. Water Resour. Res.; 2015; 51, pp. 4727-4750. [DOI: https://dx.doi.org/10.1002/2014WR016716]

7. Engel, M.; Penna, D.; Bertoldi, G.; Dell’Agnese, A.; Soulsby, C.; Comiti, F. Identifying run-off contributions during melt-induced run-off events in a glacierized alpine catchment. Hydrol. Process.; 2015; 30, pp. 343-364. [DOI: https://dx.doi.org/10.1002/hyp.10577]

8. Pu, T.; Qin, D.; Kang, S.; Niu, H.; He, Y.; Wang, S. Water isotopes and hydrograph separation in different glacial catchments in the southeast margin of the Tibetan Plateau. Hydrol. Process.; 2017; 31, pp. 3810-3826. [DOI: https://dx.doi.org/10.1002/hyp.11293]

9. Rodriguez, M.; Ohlanders, N.; Pellicciotti, F.; Williams, M.W.; McPhee, J. Estimating runoff from a glacierized catchment using natural tracers in the semi-arid Andes cordillera. Hydrol. Process.; 2016; 30, pp. 3609-3626. [DOI: https://dx.doi.org/10.1002/hyp.10973]

10. Rets, E.; Khomiakova, V.; Kornilova, E.; Ekaykin, A.; Kozachek, A.; Mikhalenko, V. How and when glacial runoff is important: Tracing dynamics of meltwater and rainfall contribution to river runoff from headwaters to lowland in the Caucasus Mountains. Sci. Total Environ.; 2024; 927, 172201. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2024.172201]

11. Sun, H.; Yao, T.; Su, F.; Yang, W.; Huang, G.; Chen, D. Regional difference in runoff regimes and changes in the Yarlung Zangbo river basin. Hydrol. Earth Syst. Sci. Discuss.; 2023; 2023, pp. 1-53.

12. Immerzeel, W.W.; Lutz, A.F.; Andrade, M.; Bahl, A.; Biemans, H.; Bolch, T.; Hyde, S.; Brumby, S.; Davies, B.J.; Elmore, A.C. et al. Importance and vulnerability of the world’s water towers. Nature; 2020; 577, pp. 364-369. [DOI: https://dx.doi.org/10.1038/s41586-019-1822-y] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31816624]

13. Su, B.; Xiao, C.; Chen, D.; Ying, X.; Huang, Y.; Guo, R.; Zhao, H.; Chen, A.; Che, Y. Mismatch between the population and meltwater changes creates opportunities and risks for global glacier-fed basins. Sci. Bull.; 2022; 67, pp. 9-12. [DOI: https://dx.doi.org/10.1016/j.scib.2021.07.027] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36545965]

14. Huss, M.; Hock, R. Global-scale hydrological response to future glacier mass loss. Nat. Clim. Chang.; 2018; 8, pp. 135-140. [DOI: https://dx.doi.org/10.1038/s41558-017-0049-x]

15. Moore, R.D. Stream Temperature Patterns in British Columbia, Canada, Based on Routine Spot Measurements. Can. Water Resour. J./Rev. Can. Des Ressour. Hydr.; 2006; 31, pp. 41-56. [DOI: https://dx.doi.org/10.4296/cwrj3101041]

16. Stahl, K.; Moore, R.D. Influence of watershed glacier coverage on summer streamflow in British Columbia, Canada. Water Resour. Res.; 2006; 42, W06201. [DOI: https://dx.doi.org/10.1029/2006WR005022]

17. Wufu, A.; Chen, Y.; Yang, S.; Lou, H.; Wang, P.; Li, C.; Wang, J.; Ma, L. Changes in Glacial Meltwater Runoff and Its Response to Climate Change in the Tianshan Region Detected Using Unmanned Aerial Vehicles (UAVs) and Satellite Remote Sensing. Water; 2021; 13, 1753. [DOI: https://dx.doi.org/10.3390/w13131753]

18. Lee, J.R.; Raymond, B.; Bracegirdle, T.J.; Chades, I.; Fuller, R.A.; Shaw, J.D.; Terauds, A. Climate change drives expansion of Antarctic ice-free habitat. Nature; 2017; 547, pp. 49-54. [DOI: https://dx.doi.org/10.1038/nature22996]

19. Zhao, Q.; Ding, Y.; Wang, J.; Gao, H.; Zhang, S.; Zhao, C.; Xu, J.; Han, H.; Shangguan, D. Projecting climate change impacts on hydrological processes on the Tibetan Plateau with model calibration against the glacier inventory data and observed streamflow. J. Hydrol.; 2019; 573, pp. 60-81. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2019.03.043]

20. Jost, G.; Moore, R.D.; Menounos, B.; Wheate, R. Quantifying the contribution of glacier runoff to streamflow in the upper Columbia River Basin, Canada. Hydrol. Earth Syst. Sci.; 2012; 16, pp. 849-860. [DOI: https://dx.doi.org/10.5194/hess-16-849-2012]

21. Rodriguez Suarez, J.A.; Diaz-Fierros, F.; Perez, R.; Soto, B. Assessing the influence of afforestation with Eucalyptus globulus on hydrological response from a small catchment in northwestern Spain using the HBV hydrological model. Hydrol. Process.; 2013; 28, pp. 5561-5572. [DOI: https://dx.doi.org/10.1002/hyp.10061]

22. Lutz, A.F.; Immerzeel, W.W.; Shrestha, A.B.; Bierkens, M.F.P. Consistent increase in High Asia’s runoff due to increasing glacier melt and precipitation. Nat. Clim. Chang.; 2014; 4, pp. 587-592. [DOI: https://dx.doi.org/10.1038/nclimate2237]

23. Singh, V.; Jain, S.K.; Shukla, S. Glacier change and glacier runoff variation in the Himalayan Baspa river basin. J. Hydrol.; 2021; 593, 125918. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2020.125918]

24. Gan, R.; Luo, Y.; Zuo, Q.; Sun, L. Effects of projected climate change on the glacier and runoff generation in the Naryn River Basin, Central Asia. J. Hydrol.; 2015; 523, pp. 240-251. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2015.01.057]

25. Luo, Y.; Arnold, J.; Liu, S.; Wang, X.; Chen, X. Inclusion of glacier processes for distributed hydrological modeling at basin scale with application to a watershed in Tianshan Mountains, northwest China. J. Hydrol.; 2013; 477, pp. 72-85. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2012.11.005]

26. Zhang, Y.; Luo, Y.; Sun, L.; Liu, S.; Chen, X.; Wang, X. Using glacier area ratio to quantify effects of melt water on runoff. J. Hydrol.; 2016; 538, pp. 269-277. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2016.04.026]

27. Wu, F.; Zhan, J.; Wang, Z.; Zhang, Q. Streamflow variation due to glacier melting and climate change in upstream Heihe River Basin, Northwest China. Phys. Chem. Earth Parts A/B/C; 2015; 79–82, pp. 11-19. [DOI: https://dx.doi.org/10.1016/j.pce.2014.08.002]

28. Yin, Z.; Feng, Q.; Liu, S.; Zou, S.; Li, J.; Yang, L.; Deo, R. The Spatial and Temporal Contribution of Glacier Runoff to Watershed Discharge in the Yarkant River Basin, Northwest China. Water; 2017; 9, 159. [DOI: https://dx.doi.org/10.3390/w9030159]

29. Devia, G.K.; Ganasri, B.P.; Dwarakish, G.S. A Review on Hydrological Models. Aquat. Procedia; 2015; 4, pp. 1001-1007. [DOI: https://dx.doi.org/10.1016/j.aqpro.2015.02.126]

30. Nyeko, M. Hydrologic Modelling of Data Scarce Basin with SWAT Model: Capabilities and Limitations. Water Resour. Manag.; 2014; 29, pp. 81-94. [DOI: https://dx.doi.org/10.1007/s11269-014-0828-3]

31. Chen, X.; Long, D.; Hong, Y.; Zeng, C.; Yan, D. Improved modeling of snow and glacier melting by a progressive two-stage calibration strategy with GRACE and multisource data: How snow and glacier meltwater contributes to the runoff of the Upper Brahmaputra River basin?. Water Resour. Res.; 2017; 53, pp. 2431-2466. [DOI: https://dx.doi.org/10.1002/2016WR019656]

32. Khanal, S.; Lutz, A.F.; Kraaijenbrink, P.D.A.; van den Hurk, B.; Yao, T.; Immerzeel, W.W. Variable 21st Century Climate Change Response for Rivers in High Mountain Asia at Seasonal to Decadal Time Scales. Water Resour. Res.; 2021; 57, e2020WR029266. [DOI: https://dx.doi.org/10.1029/2020WR029266]

33. Su, F.; Zhang, L.; Ou, T.; Chen, D.; Yao, T.; Tong, K.; Qi, Y. Hydrological response to future climate changes for the major upstream river basins in the Tibetan Plateau. Glob. Planet. Chang.; 2016; 136, pp. 82-95. [DOI: https://dx.doi.org/10.1016/j.gloplacha.2015.10.012]

34. Wang, Y.; Wang, L.; Zhou, J.; Yao, T.; Yang, W.; Zhong, X.; Liu, R.; Hu, Z.; Luo, L.; Ye, Q. et al. Vanishing Glaciers at Southeast Tibetan Plateau Have Not Offset the Declining Runoff at Yarlung Zangbo. Geophys. Res. Lett.; 2021; 48, e2021GL094651. [DOI: https://dx.doi.org/10.1029/2021GL094651]

35. Xuan, W.; Xu, Y.-P.; Fu, Q.; Booij, M.J.; Zhang, X.; Pan, S. Hydrological responses to climate change in Yarlung Zangbo River basin, Southwest China. J. Hydrol.; 2021; 597, 125761. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2020.125761]

36. Li, C.; Hao, J.; Zhang, G.; Fang, H.; Wang, Y.; Lu, H. Runoff variations affected by climate change and human activities in Yarlung Zangbo River, southeastern Tibetan Plateau. CATENA; 2023; 230, 107184. [DOI: https://dx.doi.org/10.1016/j.catena.2023.107184]

37. Guo, W.; Liu, S.; Xu, J.; Wu, L.; Shangguan, D.; Yao, X.; Wei, J.; Bao, W.; Yu, P.; Liu, Q. et al. The second Chinese glacier inventory: Data, methods and results. J. Glaciol.; 2017; 61, pp. 357-372. [DOI: https://dx.doi.org/10.3189/2015JoG14J209]

38. Feng, Y.; Jiang, K.; Kuang, X.; Yao, Y.; Liang, S.; Yu, K.; Liu, J.; Zheng, C. The dual role of meltwater in buffering river runoff in the Yarlung Zangbo Basin, Tibetan Plateau. J. Hydrol. Reg. Stud.; 2024; 54, 101857. [DOI: https://dx.doi.org/10.1016/j.ejrh.2024.101857]

39. Ji, H.; Peng, D.; Gu, Y.; Luo, X.; Pang, B.; Zhu, Z. Snowmelt Runoff in the Yarlung Zangbo River Basin and Runoff Change in the Future. Remote Sens.; 2022; 15, 55. [DOI: https://dx.doi.org/10.3390/rs15010055]

40. Arnold, J.G.; Srinivasan, R.; Muttiah, R.S.; Williams, J.R. Large Area Hydrologic Modeling and Assessment Part I: Model Development1. JAWRA J. Am. Water Resour. Assoc.; 2007; 34, pp. 73-89. [DOI: https://dx.doi.org/10.1111/j.1752-1688.1998.tb05961.x]

41. Chen, L.; Chen, S.; Li, S.; Shen, Z. Temporal and spatial scaling effects of parameter sensitivity in relation to non-point source pollution simulation. J. Hydrol.; 2019; 571, pp. 36-49. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2019.01.045]

42. Scott Munro, D. A revised Canadian perspective: Progress in glacier hydrology. Hydrol. Process.; 2005; 19, pp. 231-245. [DOI: https://dx.doi.org/10.1002/hyp.5769]

43. Jiang, X.; Wang, N.; He, J.; Wu, X.; Song, G. A distributed surface energy and mass balance model and its application to a mountain glacier in China. Chin. Sci. Bull.; 2010; 55, pp. 2079-2087. [DOI: https://dx.doi.org/10.1007/s11434-010-3068-9]

44. Hock, R.; Jacobsen, P.; Braun, A. Modelling the Response of Mountain Glacier Discharge to Climate Warming. Global Change and Mountain Regions: An Overview of Current Knowledge; Huber, U.M.; Reasoner, M.A.; Bugmann, H. Springer: Berlin/Heidelberg, Germany, 2005; pp. 243-252.

45. Neitch, S.L.; Arnold, J.G.; Kiniry, J.R.; Williams, J.R.; King, K.W. Soil Water Assessment Tool Theoretical Document; version 2000 Grassland, Soil and Water Research Laboratory, Agricultural Research Service: Temple, TX, USA, 2002.

46. Allen, R.G.; Pereira, L.S.; Raes, D.; Smith, M. Crop Evapotranspiration—Guidelines for Computing Crop Water Requirements; Irrigation and Drain, Paper No. 56. FAO: Rome, Italy, 1998.

47. Chen, J.; Ohmura, A. Estimation of alpine glacier water resources and their change since the 1870s. Hydrology in Mountainous Regions—I. Hydrologic Measurements, the Water Cycle; IAHS: Wallingford, CT, USA, 1990; Volume 193, pp. 127-135.

48. Liu, S.; Sun, W.; Shen, Y.; Li, G. Glacier changes since the Little Ice Age maximum in the western Qilian Shan, northwest China, and consequences of glacier runoff for water supply. J. Glaciol.; 2003; 49, pp. 117-124.

49. Li, M.; Di, Z.; Duan, Q. Effect of sensitivity analysis on parameter optimization: Case study based on streamflow simulations using the SWAT model in China. J. Hydrol.; 2021; 603, 126896. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2021.126896]

50. Abbaspour, K.C.; Johnson, C.A.; van Genuchten, M.T. Estimating uncertain flow and transport parameters using a sequential uncertainty fitting procedure. J. Vadose Zone; 2004; 3, pp. 1340-1352. [DOI: https://dx.doi.org/10.2136/vzj2004.1340]

51. Abbaspour, K.C.; Yang, J.; Maximov, I.; Siber, R.; Bogner, K.; Mieleitner, J.; Zobrist, J.; Srinivasan, R. Modelling hydrology and water quality in the pre-alpine/alpine Thur watershed using SWAT. J. Hydrol.; 2007; 333, pp. 413-430. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2006.09.014]

52. Abbaspour, K.C.; Rouholahnejad, E.; Vaghefi, S.; Srinivasan, R.; Yang, H.; Kløve, B. A continental-scale hydrology and water quality model for Europe: Calibration and uncertainty of a high-resolution large-scale SWAT model. J. Hydrol.; 2015; 524, pp. 733-752. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2015.03.027]

53. Moriasi, D.N.; Arnold, J.G.; Van Liew, M.W.; Bingner, R.L.; Harmel, R.D.; Veith, T.L. Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Trans. ASABE; 2007; 50, pp. 885-900. [DOI: https://dx.doi.org/10.13031/2013.23153]

54. Xuan, W.; Fu, Q.; Qin, G.; Zhu, C.; Pan, S.; Xu, Y.-P. Hydrological Simulation and Runoff Component Analysis over a Cold Mountainous River Basin in Southwest China. Water; 2018; 10, 1705. [DOI: https://dx.doi.org/10.3390/w10111705]

55. Chen, Z.; Zhu, R.; Yin, Z.; Feng, Q.; Yang, L.; Wang, L.; Lu, R.; Fang, C. Hydrological response to future climate change in a mountainous watershed in the Northeast of Tibetan Plateau. J. Hydrol. Reg. Stud.; 2022; 44, 101256. [DOI: https://dx.doi.org/10.1016/j.ejrh.2022.101256]

56. Wu, H.; Chen, B. Evaluating uncertainty estimates in distributed hydrological modeling for the Wenjing River watershed in China by GLUE, SUFI-2, and ParaSol methods. Ecol. Eng.; 2015; 76, pp. 110-121. [DOI: https://dx.doi.org/10.1016/j.ecoleng.2014.05.014]

57. He, Q.; Kuang, X.; Chen, J.; Hao, Y.; Feng, Y.; Wu, P.; Zheng, C. Glacier retreat and its impact on groundwater system evolution in the Yarlung Zangbo source region, Tibetan Plateau. J. Hydrol. Reg. Stud.; 2023; 47, 101368. [DOI: https://dx.doi.org/10.1016/j.ejrh.2023.101368]

58. Cui, Y.; Ye, B.; Wang, J.; Liu, Y. Influence of degree-day factor variation on the mass balance of Glacier No. 1 at the headwaters of Ürümqi River, China. J. Earth Sci.; 2013; 24, pp. 1008-1022. [DOI: https://dx.doi.org/10.1007/s12583-013-0394-2]

59. Du, X.; Silwal, G.; Faramarzi, M. Investigating the impacts of glacier melt on stream temperature in a cold-region watershed: Coupling a glacier melt model with a hydrological model. J. Hydrol.; 2022; 605, 127303. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2021.127303]

60. Sun, G.; Hallema, D.; Asbjornsen, H. Ecohydrological processes and ecosystem services in the Anthropocene: A review. Ecol. Process.; 2017; 6, 35. [DOI: https://dx.doi.org/10.1186/s13717-017-0104-6]

61. Mohsenstandi, O.; Stefan, H.G.; Erickson, T.R. A nonlinear regression model for weekly stream temperatures. Water Resour. Res.; 1998; 34, pp. 2685-2692.

62. Ban, C.; Xu, Z.; Zuo, D.; Li, P.; Wang, J.; Huang, G.; Chen, D. Hydrological simulation and runoff components analysis in the high cold alpine region: Case study in the upper reaches of the Parlung Zangbo watershed of the Yarlung Zangbo River basin. J. Beijing Norm. Univ.; 2023; 59, pp. 85-93.

63. Rounce, D.R.; Hock, R.; Maussion, F.; Hugonnet, R.; Kochtitzky, W.; Huss, M.; Berthier, E.; Brinkerhoff, D.; Compagno, L.; Copland, L. et al. Global glacier change in the 21st century: Every increase in temperature matters. Science; 2023; 379, pp. 78-83. [DOI: https://dx.doi.org/10.1126/science.abo1324]