What effects have snow cover and climate change had on water flows, rivers, water storage, and groundwater? What solutions are there to prevent damage caused by climate change?

As a significant component of the cryosphere, snow cover plays a crucial role in modulating atmospheric circulation and regional hydrological equilibrium. Therefore, studying the dynamics of snow cover and its response to climate change is of great significance for regional water resource management and disaster prevention. In this study, reanalysis climate datasets and a new MODIS snow cover extent product over China were used to analyze the characteristics of climate change and spatiotemporal variations in snow cover in the Keriya River Basin (KRB). Furthermore, the effects of climate factors on snow cover and their coupling effects on runoff were quantitatively evaluated by adopting partial least squares regression (PLSR) method and structural equation modeling (SEM), respectively. Our findings demonstrated the following: (1) Air temperature and precipitation of KRB showed a significant increase at rates of 0.24 ◦C/decade and 14.21 mm/decade, respectively, while the wind speed did not change significantly. (2) The snow cover frequency (SCF) in the KRB presented the distribution characteristics of “low in the north and high in the south”. The intra-annual variation of snow cover percentage (SCP) of KRB displayed a single peak (in winter), double peaks (in spring and autumn), and stability (SCP > 75%), whose boundary elevations were 4000 m and 6000 m, respectively. The annual, summer, and winter SCP in the KRB declined, while the spring and autumn SCP experienced a trend showing an insignificant increase during the hydrological years of 2001–2020. Additionally, both the annual and seasonal SCF (except autumn) will be further increased in more than 50% of the KRB, according to estimates. (3) Annual and winter SCF were controlled by precipitation, of which the former showed a mainly negative response, while the latter showed a mainly positive response, accounting for 43.1% and 76.16% of the KRB, respectively. Air temperature controlled SCF changes in 45% of regions in spring, summer, and autumn, mainly showing negative effects. Wind speed contributed to SCF changes in the range of 11.23% to 26.54% across annual and seasonal scales. (4) Climate factors and snow cover mainly affect annual runoff through direct influences, and the total effect was as follows: precipitation (0.609) > air temperature (−0.122) > SCP (0.09).Snow cover, as an indispensable constituent of the cryosphere, plays a highly important role within the global and regional climate system, exerting an essential impact on surface radiation balance, energy balance, and hydrological partitioning [1–4]. Concurrent with the relentless warming of the global climate, the Northern Hemisphere is experiencing a pronounced decline in snow cover [5,6], which profoundly impacts the regional water cycle [7,8]. Inland rivers situated in the arid and semi-arid regions of northwest China predominantly rely on the runoff derived from the thawing of adjacent alpine mountain snowpacks [9]. Changes in the total and seasonal runoff distributions caused by snow cover changes will affect social and economic development in the middle and lower reaches [10–12]. Therefore, studying the influence of snow cover dynamics and climate change on runoff to inform decisions about the scientific management and planning of water resources in Northwest arid areas has become imperative [13,14]. The conventional method for monitoring snow cover involves the collection of snow cover data from meteorological stations or observation sites. However, this approach suffers from spatial inhomogeneity due to the uneven distribution of the stations, primarily concentrated in low-altitude regions [15–18]. In remote alpine areas, the complex terrain and challenging environmental conditions impede field monitoring efforts [19]. Nevertheless, recent advancements in remote sensing technology have unlocked new opportunities for snow cover research. Remote sensing data offer the advantages of wide broad coverage, frequent updating, and high spatial resolution, which compensates for the limitations of ground-based monitoring data [20,21]. Currently, various remote sensing data are leveraged for snow cover research, including the moderate resolution imaging spectroradiometer (MODIS), advanced very high resolution radiometer (AVHRR), scanning multichannel microwave radiometer (SMMR), and other related products [22–24]. Among these products, MODIS snow cover products have emerged as the mainstream data of remote sensing snow cover products due to their high spatiotemporal resolutions. Additionally, they have been widely adopted in the study of snow cover changes across regions of varying scales [25–27]. For instance, Zou et al. [28] employed MOD10A2 and MYD10A2 snow cover products to explore the variation of snow cover in Northern Xinjiang, Qinghai-Tibet Plateau, and Northeast China. Their study revealed an insignificant increasing trend in snow cover areas and snow depths from 2001 to 2020. Thapa et al. [29] combined three different 8 day composite snow products, including MOD10A2, MYD10A2, and MOYDGL06, to analyze the variation trend of snow cover in the Karakoram region. They discovered a negligible decline in snow cover areas from 2003 to 2018. However, despite the widespread utilization of MODIS data in snow cover research, the accuracy of MODIS data can be affected by factors such as cloud cover and land cover. Therefore, a daily cloud-gap-filled MODIS snow cover extent product produced by Hao et al. [30], which comprehensively considered the impact of land cover and cloud on original MODIS snow cover data, was selected as the main data source for the present study. A detailed explanation of these data can be seen in Section 2.2.1. Snow cover, as an exceedingly responsive component of the cryosphere, is profoundly influenced by climate change and is regarded as a vital indicator of global climate change [31]. Unveiling the response of snow cover to climate change has emerged as a primary focus within snow cover research, garnering extensive attention from the scholarly community. For instance, Du et al. [32] conducted an analysis of the relationship between snow cover frequency (SCF) and climate factors in the Qilian Mountains from 2000 to 2020. They suggested that the SCF was dominated by precipitation rather than air temperature, with precipitation playing a positive role. However, Hussain et al. [33] investigated the impact of climate factors on snow cover area within the Gilgit River Basin from 2001 to 2015 and reported a negative correlation between snow cover area and air temperature, while precipitation exhibited no evident relationship. In recent years, the focus of scholars has predominantly centered around the influence of air temperature and precipitation on snow cover, often neglecting the potential effect of wind speed. Although certain researchers have discussed the potential impact of wind speed on snow cover [34–36], the majority of these discussions lie within the realm of qualitative studies, with few quantitative analyses conducted on the impact of wind speed on snow cover [37]. Moreover, regarding research methods, most researchers mainly applied the Pearson correlation method to investigate the influence of climate factors on snow cover. However, this kind of method ignores the interactions among various climate factors, which subsequently affects the accuracy of the results derived from the analysis. In contrast, partial least squares regression (PLSR) represents a new multivariate statistical data analysis method capable of eliminating the multiple correlations among independent variables [38], which enables one to clarify the degree of influence of different climate factors on snow cover change. The Keriya River, a typical inland river in the arid region, stands as the largest river and primary water source for the Yutian Oasis. The alpine mountains contribute crucial meltwater, which serves as the principal supply for the Keriya River and represents a valuable resource for the survival and development of the downstream regions [39]. Although most of the glaciers within the Keriya River Basin (KRB) exhibited stability, there existed a slight trend of total area reduction [40]. In contrast, the snow cover area experienced a trend depicting a slight increase [41]. From 1957 to 2017, the KRB witnessed an increase in runoff depth at a rate of 4.27 mm/decade, which was mainly affected by air temperature and precipitation [41,42]. Recent studies concerning the KRB only focus on the changes in snow cover or runoff, with few conducting quantitative and systematic analyses regarding the relationship between climate, snow cover, and runoff. Structural equation modeling (SEM) can comprehensively analyze the relationship between various variables, allowing one to quantify the direct and indirect effects of climate and snow cover on runoff. The research objectives of this study were as follows: (1) to reveal the variation characteristics of climate and snow cover in KRB based on reanalysis climate datasets and the new MODIS snow cover extent product over China, (2) to analyze the influence of different climate factors (air temperature, precipitation, and wind speed) on SCF at the pixel scale by adopting PLSR method, and (3) to discuss the interplay between snow cover, climate factors, and their collective influence on runoff through the application of structural equation modeling (SEM). The results of this paper could facilitate a better understanding the spatiotemporal variation of snow cover in the KRB and its influencing mechanism and clarify the regional water cycle process. Furthermore, this study is of great significance to the utilization and management of water resources in the context of climate change.

. Conclusions In this study, reanalysis climate datasets were used to analyze the characteristics of climate change in the KRB. The spatiotemporal distributions of snow cover in the KRB during 20 hydrological years from 2001 to 2020 were analyzed utilizing a new MODIS snow cover extent product over China. The response of snow cover change to climate factors was evaluated using PLSR. Furthermore, the study delved into the impact of snow cover and climate factors on annual runoff by employing SEM. The findings from the present study can be summarized as follows: (1) There was a significant increase in air temperature and precipitation, with rates of 0.24 ◦C/decade and 14.21 mm/decade, and the mutation year occurred in 1996 and 1986, respectively. However, wind speed did not change significantly. (2) In terms of spatial distribution, the SCF in the KRB presented “low in the north and high in the south” distribution characteristics. The SCP in the KRB demonstrated an elevation-dependent increase, with the highest values observed in the north aspect and in the 10–15 degrees slope zone. Regarding the intra-annual variation, the SCP within the KRB demonstrated distinctive patterns, including a single peak in winter, double peaks in both spring and autumn, and a consistent high value (SCP > 75%) with turning elevations of 4000 m and 6000 m, respectively. Moreover, the peak SCP values showed a delayed trend with increasing elevation. In terms of temporal change, the SCP in the KRB decreased annually and in summer and winter; however, it increased in spring and autumn between 2001 and 2020. More than 50% of the KRB experienced a decreasing trend for annual, spring, summer, and winter SCF, whereas 38.24% of the areas showed an increasing trend in autumn. In addition to autumn, annual and seasonal SCF is estimated to show an upward trend in the future, accounting for more than 50% of the KRB. (3) The annual SCF was mainly negatively affected by precipitation, while in winter, it was mainly positively affected by precipitation, accounting for 43.1% and 76.16% of the area, respectively. The spring, summer, and autumn SCF changes in more than 45% of KRB were controlled by air temperature, exerting a predominantly negative influence. Annually and during spring, the impact of wind speed on SCF was mainly positive; however, it negative in summer, autumn, and winter, with the area controlled by wind speed ranging from 11.23% to 26.54%. (4) The total effect of climate factors and SCP on the annual runoff in the KRB was as follows: precipitation (0.609) > air temperature (−0.122) > SCP (0.09). Climate factors and SCP mainly exerted a direct effect on the changes in annual runoff.