Response of Snowmelt over the Mongolian Plateau to Air Temperature

NIU Jin¹, LIU Yahong², BAO Gang¹, YUAN Zhihui³, TONG Siqin¹, Chaobuga¹

¹College of Geographical Science, Inner Mongolia Normal University, Hohhot, Inner Mongolia, China
²Inner Mongolia Academy of Agricultural & Animal Husbandry Sciences, Hohhot, Inner Mongolia, China
³College of Water Conservancy and Civil Engineering, Inner Mongolia Agricultural University, Hohhot, Inner Mongolia, China

Abstract

Using MODIS snow product data, this study investigates the spatiotemporal variation characteristics of the snowmelt period over the Mongolian Plateau during the 2003–2022 hydrological years. The movement of the snowmelt line toward higher latitudes and its response to air temperature are tracked and analyzed at 15-day intervals. The results show that: (1) The proportion of snow-covered area to the total area of the Mongolian Plateau during the 2003–2022 hydrological years ranged from 55.59% to 87.61%, with the smallest snow cover in 2018 and the largest in 2009. Additionally, over the past 20 years, the snowmelt start time on the Mongolian Plateau exhibited a significant advancing trend at a rate of 0.18 days per decade (P<0.05), while the stable snow-cover area showed a delaying trend. (2) Spatially, snowmelt occurred significantly later in northern regions of the Mongolian Plateau compared to southern regions. Stable snow-cover areas were primarily concentrated in western Mongolia and northeastern Inner Mongolia, where snowmelt times were generally later. Approximately 64.9% of these areas showed an advancing trend in snowmelt, while regions with delaying trends were mainly distributed in the northwestern part of the study area. (3) Observational analysis at half-monthly scales from January during the winter season revealed that the movement of the snowmelt line demonstrated successive synchronicity with the -5°C and 0°C isotherms. Correlation coefficients between snowmelt line positions and temperature, except for the year 2018 (with the least snow cover), generally fell within the higher range of 0.72 to 0.98, indicating that temperature is a key factor influencing the position of the snowmelt line.

Keywords: snow end day; snow line; isotherm; temperature response; Mongolian Plateau

1 Introduction

Snow cover, as one of the most sensitive climate elements to global warming, plays a crucial role in regional and global water cycles and climate systems. Over the past century, Earth has experienced significant warming, particularly after 1990, with an accelerated warming rate. According to the IPCC's Sixth Assessment Report, global surface temperature has risen by 1.1°C by 2023 compared to pre-industrial levels, with all regions facing unprecedented climate system changes. The direct impact of rising temperatures is the reduction of snow cover area and changes in snow phenology, including snow onset day (SOD), snow end day (SED), and snow duration days (SDD). These changes affect agricultural irrigation districts that depend on snowmelt water, thereby influencing food production and regional socio-economic development. Additionally, increased warming leads to more freeze-thaw cycles and higher rain-to-snow ratios, causing reduced streamflow and habitat loss for certain species in high-latitude ecosystems. Snow end day serves as one of the most intuitive and critical indicators for measuring climate warming impacts on Earth's surface.

Research methods for snow phenology at regional or global scales can be categorized into two main types: ground-based observations using meteorological data and remote sensing retrieval methods using satellite observations. The former relies on long-term snow records from meteorological stations, offering high accuracy but limited spatial coverage and significant discontinuity. The latter obtains large-scale snow dynamic information through satellite remote sensing, providing advantages in broad spatial coverage and strong timeliness. With the public release of high spatiotemporal resolution satellite remote sensing data products such as MODIS, significant progress has been made in regional-scale seasonal snow phenology research, revealing spatiotemporal distribution patterns of snow phenology, snow water equivalent, and snow depth, as well as their relationships with climate change.

Previous studies have shown that snow cover duration in most Northern Hemisphere regions has decreased in recent years. For instance, research using MODIS snow products and snow water equivalent data to construct a comprehensive snow phenology matrix revealed that SED in high-latitude and high-altitude regions of the Northern Hemisphere showed an advancing trend during 2001–2014, while mid-latitude regions experienced relative delays. Studies on the Tibetan Plateau using passive microwave remote sensing snow depth data quantified the dependence of maximum snow depth and SED on altitude. Analysis of daily cloud-free snow area products in Northern Xinjiang found that snow cover area ratio and SOD significantly advanced from 1980 to 2019, though SDD showed no obvious change. Research using ERA5-Land data and multiple auxiliary datasets for random forest modeling to generate daily snow cover fraction data revealed a significant decreasing trend in normalized cumulative snow cover fraction. Investigations in the Yurungkax River Basin of the West Kunlun Mountains found that temperature was the main factor affecting snow area changes in spring and summer at low altitudes, while precipitation dominated changes at high altitudes in winter and spring.

The Mongolian Plateau is located deep inland, far from oceans, with limited water vapor transport, strong evaporation, and a typical arid to semi-arid climate, fostering ecosystems dominated by temperate grasslands and desert steppes. During the dry and rainless spring, as temperatures rise, surface snow melts extensively while the soil layer remains frozen, preventing infiltration. With low vegetation coverage and minimal evapotranspiration, snowmelt water flows into rivers, becoming an important source for local lakes and rivers. Studies indicate that lakes on the Mongolian Plateau have shown significant reductions in number and area, decreasing by 1443.92 km² overall. Therefore, studying snowmelt dynamics on the Mongolian Plateau is crucial for understanding spatiotemporal evolution trends of water resources and the convergence effects between temperature and snow cover under deepening climate change. Although recent studies have gradually focused on spatiotemporal changes in snow phenology and its relationship with climate variability, the latitudinal movement characteristics of snowmelt dynamics and its association and synchronicity with different isotherms remain unclear, making it difficult to visually present the process of snowmelt advancing toward higher latitudes and altitudes.

This study utilizes daily MODIS snow products MOD10C1 and MYD10C1, processed through maximum value composition and masking to obtain daily snow cover data for the study area. Combined with ERA5-Land reanalysis temperature data, we analyze spatiotemporal variation characteristics of snow distribution area and SED on the Mongolian Plateau, and examine the response process of the snowmelt line to air temperature at the pixel scale. The results enrich regional response studies to global climate change and provide references for local climate change adaptation and water resource management.

1.1 Study Area Overview

The Mongolian Plateau is situated in central Asia, located between 37°22′–53°20′N and 87°43′–126°04′E. Geomorphologically, it extends from the Mongolian Altai Mountains in the west to the Greater Khingan Mountains in the east, bounded by the Sayan and Khentii Mountains in the north, with the vast Gobi Desert to the south, demarcated by the Yin Mountains. The main area includes the entire territory of Mongolia and China's Inner Mongolia, covering approximately 2.75×10⁶ km². The plateau features high plains and mountainous terrain with an average elevation of 1580 m, increasing from east to west. It has a temperate continental climate with average annual precipitation of about 200 mm, transitioning from arid to semi-arid regions from west to east.

1.2 Data Sources

MODIS Snow Data. The data sources for this study are MODIS MOD10C1 and MYD10C1 snow products, both with a spatial resolution of 0.05°×0.05° and temporal resolution of one day. These two products observe the same region once in the morning and once in the afternoon within the same day, allowing synthesis of the two daily observations to reduce cloud impacts on snow detection. Both products contain four datasets: daily global snow extent map, daily snow map clear index, daily cloud obscuration percentage, and general QA of data in grid cell. The global daily snow extent data represents the percentage of snow cover in each 0.05°×0.05° pixel, ranging from 0–100, where 0 indicates no snow cover and 100 indicates complete snow cover. Considering the high-altitude characteristics of the Mongolian Plateau where snow typically begins falling after October 1 and largely melts by mid-May of the following year, this study defines a hydrological year from September 1 to August 31 of the following year. Daily global snow extent data from 2003–2022 are selected as snow phenology identification data (2000 data were excluded due to missing and discontinuous records). To reduce impacts from cloud cover, atmosphere, and other external factors, MOD10C1 and MYD10C1 data underwent maximum value composition and clipping along the Mongolian Plateau boundary to obtain final daily snow cover data.

Meteorological Data. Meteorological data are obtained from the ERA5-Land dataset produced by the European Centre for Medium-Range Weather Forecasts, available from 1981 to present with 0.1° spatial resolution. Compared to its predecessor ERA5 (0.25° monthly single-level data) and earlier climate research datasets (0.50°), ERA5-Land shows significant improvements in spatial resolution and accuracy. The dataset employs a tiled scheme for land surface exchange proven suitable for ground modeling. This study uses 2-meter air temperature data from ERA5-Land, processed through online data extraction, conversion to daily data, cropping, and Kelvin-to-Celsius conversion to obtain daily temperature spatial distribution data for analyzing snow response to temperature.

1.3 Methods

1.3.1 Snow Phenology Analysis. To reduce potential errors from clouds and thin, temporary snow that easily melts or blows away, and considering the actual spatial precipitation distribution on the Mongolian Plateau (with annual precipitation below 100 mm in the southwestern Gobi Desert, some areas even below 50 mm), this study records pixels with snow coverage ≥10% as snow-covered and sets values <10% as snow-free (0). Snow cover fraction for each year is calculated by counting snow pixels relative to total pixels in the study area. Following Yuan Zhihui et al., SED is defined as the last day with continuous snow cover for at least 15 days, SOD as the first day with continuous snow cover for at least 15 days, and SDD as the number of days between SOD and SED. The formulas are:

$$
SED = \max{j | S(j) \geq 4}
$$

$$
SOD = \min{j | S(j) \geq 4}
$$

$$
SDD = SED - SOD
$$

where i represents the day of the hydrological year (1–365/366), j represents the hydrological year (2003–2022), n is the number of snow days, and S is the snow cover status.

1.3.2 Trend Analysis. Linear regression analysis between SED and year is performed to indicate change trends. The regression slope b represents the change rate (interannual variation rate), with negative slopes indicating advancing trends and positive slopes indicating delaying trends. Significance is tested using F-test, with trends classified as: significantly delaying (b>0, P<0.05), significantly advancing (b<0, P<0.05), non-significantly delaying (b>0, P>0.05), and non-significantly advancing (b<0, P>0.05).

1.3.3 Synchronization Analysis Between Isotherms and Snowmelt Line. The snowmelt line is defined as the boundary between snow-covered and snow-free areas. To systematically analyze the response mechanism of snowmelt processes to temperature changes, this study examines dynamic coupling between the snowmelt line and isotherms. The time scale is set at half-month intervals to balance spatial fluctuation capture and temporal association. The snowmelt line is identified by sliding forward from the last day of the hydrological year in 15-day steps, recording the first occurrence of ≥4 days with snow cover as SOD. Isotherms are drawn from corresponding period temperature data, reflecting spatial heat distribution. This scale ensures effective characterization of regional heat gradients while capturing progressive changes in snowmelt boundaries under temperature accumulation effects.

2 Results

2.1 Temporal Variation Characteristics of SED

Figure 2 shows the proportion of snow-covered pixels relative to total plateau pixels from 2003–2022. Snow cover area exhibited large fluctuations, ranging from 55.59% to 87.61% of the total plateau area, with the minimum in 2018 and maximum in 2009. Overall, SED occurred primarily between early March and late August (days 155–235), showing a significant advancing trend at 0.18 days per decade (P<0.05). The earliest SED was day 184 in 2018, while the latest was day 203 in 2009.

2.2 Spatial Variation Characteristics of SED

The spatial distribution of multi-year average SED shows a clear pattern of earlier melt in the south and later melt in the north, concentrated in March–August. Southern regions of the Mongolian Plateau are climatically arid with low snowfall frequency, while northern regions receive more precipitation and have lower temperatures, resulting in later SED. Due to annual variations in snow cover extent, the years with maximum (2009) and minimum (2018) snow cover are selected for detailed analysis. In 2009, SED in areas north of 45°N was generally later, typically after day 195,主要分布在蒙古国蒙古阿尔泰山、萨彦岭、杭爱山及肯特山等高海拔山区和东部开阔的草原地区，以及降水量较丰富的内蒙古东北呼伦贝尔地区. In 2018, SED was earlier,主要分布在高原中部和南部区域, with no-snow area increasing by 44.4%.

Stable snow-cover areas, defined as regions with snow cover every year, are mainly distributed in high-latitude, high-altitude northwestern Mongolia and the Greater Khingan region of northeastern Inner Mongolia, with average SED of 217 days (range 193–238 days), about 27 days later than the overall plateau average. Trend analysis reveals that 64.9% of stable snow-cover areas showed an advancing trend, primarily in western Hulunbuir and northern Mongolia, while 35.1% showed a delaying trend, mainly in western Mongolia and northeastern Inner Mongolia.

2.3 Movement of Snowmelt Line and Isotherms

Given that snowmelt generally begins in spring, this study uses a half-monthly scale from January to May of the following year (2003–2022) to analyze dynamic movement of the snowmelt line and -5°C/0°C isotherms. The spatial distribution characteristics are shown in Figure 4.

From January 1–15, the snowmelt line appeared in southern plateau areas, with melting regions sporadically distributed in high-altitude western mountains (Altai and Hangay). The -5°C isotherm showed synchronous movement with the snowmelt line, while the 0°C isotherm did not appear. From January 16–31, melting regions remained small and scattered. The -5°C isotherm began showing synchronous trends with the snowmelt line, though distribution remained random due to low annual precipitation (<200 mm) and thin snowpack lacking regular patterns.

From February 1–15, synchronous movement between the -5°C isotherm and snowmelt line became more stable, with the isotherm at significantly higher latitudes. From February 16–28, large-scale snowmelt began in northern Mongolia as temperatures rose substantially, with both isotherms and snowmelt line showing highly synchronous movement. Regions with snow duration >80 days and annual precipitation >200 mm more readily exhibited this synchronization.

Figure 5 shows latitudinal and longitudinal movement trends. All three lines (snowmelt line, -5°C and 0°C isotherms) showed significant upward trends in latitude from February 1, moving northward at approximately 2.147° latitude per half-month for the -5°C isotherm. In longitude, both isotherms showed significant decreasing trends (westward movement), with the -5°C isotherm moving westward fastest at about 2.086° per half-month. The latitudinal movement of the -5°C isotherm more closely matched the snowmelt line than the 0°C isotherm.

Figure 6 shows the geodesic distance between the snowmelt line and adjacent isotherms. The nearest distance gradually decreased, particularly during early March to early April, indicating rapid temperature rise. After early April, distances stabilized at minimal values, showing the closest synchronization between snowmelt line and 0°C isotherm.

Figure 7 presents correlation analysis between half-monthly mean air temperature and snowmelt line latitude from 2003–2022. Excluding 2018 (the year with least snow cover), R² values ranged from 0.72–0.98, confirming temperature as a key factor controlling snowmelt line position. Higher average temperatures corresponded to higher latitudes of the snowmelt line, with the strongest correlation in 2009 (R²=0.9702) and weakest in 2018 (R²=0.6461). Years with R² between 0.7–0.8 corresponded to periods when snowmelt line and -5°C isotherm showed strongest synchronization.

3 Discussion

This study analyzed spatiotemporal variation characteristics of SED and the response relationship between snowmelt line and air temperature at half-monthly scales using MODIS daily snow products and ERA5-Land temperature data. The snowmelt process primarily follows the movement of the -5°C isotherm, though the critical melt temperature may involve multiple factors including geographic location shifts, latent heat accumulation, and differences between actual snow surface temperature and air temperature.

The spatial distribution of SED shows earlier melt in the south and later melt in the north, consistent with findings by Sun Hui et al. and Li Chenhao. This pattern results from combined effects of topography, latitude, and climate, particularly precipitation patterns. Water vapor from the Arctic and Pacific Oceans converges more easily in northern plateau regions, creating thicker snowpack. High latitudes receive limited solar radiation during snow cover periods, with high albedo further lowering surface temperatures. In contrast, central and southern Gobi and desert regions have limited water vapor transport, low precipitation, higher temperatures, and enhanced snow sublimation, resulting in earlier SED.

Stable snow-cover areas are concentrated in high-altitude northwestern Mongolia and the Greater Khingan region, with average SED of 217 days. In these areas, 64.9% showed advancing trends while 35.1% showed delaying trends, possibly related to accelerated snowmelt due to climate warming. Previous studies indicate snow phenology changes are influenced by temperature and precipitation, but research on how melt processes change with critical temperatures remains limited. This study reveals clear latitudinal variation characteristics of snowmelt trends following -5°C isotherm movement.

4 Conclusions

Using MODIS snow product data and ERA5-Land temperature reanalysis data, this study analyzed spatiotemporal distribution characteristics of snowmelt on the Mongolian Plateau and the response relationship between snowmelt line and air temperature at half-monthly scales. The main conclusions are:

1. Snow cover area and temporal trends: From 2003–2022, snow cover area fluctuated between 55.59%–87.61% of the total plateau area, with minimum coverage in 2018 and maximum in 2009. SED showed a significant advancing trend at 0.18 days per decade (P<0.05), while stable snow-cover areas advanced at 2.14 days per decade.

2. Spatial distribution patterns: SED exhibited a clear south-early, north-late pattern concentrated in March–August. In the year with maximum snow cover (2009), SED in areas north of 45°N was generally later (after day 195), while in the minimum snow year (2018), SED was earlier and no-snow area increased by 44.4%, mainly in central and southern plateau regions. Stable snow-cover areas are concentrated in high-altitude northwestern Mongolia and the Greater Khingan region, with average SED of 217 days (27 days later than the plateau average). In these areas, 64.9% showed advancing trends while 35.1% showed delaying trends.

3. Snowmelt line and isotherm synchronization: During winter seasons from 2003–2022, the northward movement of the snowmelt line showed high consistency with -5°C and 0°C isotherms. This reflects enhanced convergence effects from deepening climate change—continuous warming causes boundaries of different thermal zones (snowmelt line, isotherms) to couple spatially, demonstrating the协同响应 of regional climate systems to warming and providing direct evidence for systematic climate change impacts.

References

[1] Barnett T P, Adam J C, Lettenmaier D P. Potential impacts of a warming climate on water availability in snow dominated regions[J]. Nature, 2005, 438(7066): 303-309.

[2] Samset B H, Zhou C, Fuglestvedt J S, et al. Steady global surface warming from 1973 to 2022 but increased warming rate after 1990[J]. Communications Earth & Environment, 2023, 4(1): 400.

[3] Lee H, Calvin K, Dasgupta D, et al. IPCC, 2023: Climate Change 2023: Synthesis Report, Summary for Policymakers[R]. Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. IPCC, Geneva, Switzerland, 2023: 1-34.

[4] Hayhoe K, Wake C P, Huntington T G, et al. Past and future changes in climate and hydrological indicators in the US Northeast[J]. Climate Dynamics, 2007, 28: 381-407.

[5] Choi G, Robinson D A, Kang S. Changing northern hemisphere snow seasons[J]. Journal of Climate, 2010, 23(19): 5305-5310.

[6] Notaro M, Lorenz D, Hoving C, et al. Twenty first century projections of snowfall and winter severity across central-eastern North America[J]. Journal of Climate, 2014, 27(17): 6526-6550.

[7] Berghuijs W R, Woods R A, Hrachowitz M. A precipitation shift from snow towards rain leads to a decrease in streamflow[J]. Nature Climate Change, 2014, 4(7): 583-586.

[8] Niittynen P, Heikkinen R K, Luoto M. Snow cover is a neglected driver of Arctic biodiversity loss[J]. Nature Climate Change, 2018, 8(11): 997-1001.

[9] Dye D G. Variability and trends in the annual snow cover cycle in Northern Hemisphere land areas, 1972-2000[J]. Hydrological processes, 2002, 16(15): 3065-3077.

[10] Peng S, Piao S, Ciais P, et al. Change in snow phenology and its potential feedback to temperature in the Northern Hemisphere over the last three decades[J]. Environmental Research Letters, 2013, 8(1): 014008.

[11] Chen X, An S, Inouye D W, et al. Temperature and snowfall trigger alpine vegetation green-up on the world's roof[J]. Global Change Biology, 2015, 21(10): 3635-3646.

[12] Allchin M I, Déry S J. A spatio-temporal analysis of trends in Northern Hemisphere snow-dominated area and duration, 1971-2014[J]. Annals of Glaciology, 2017, 58(75pt1): 21-35.

[13] Chen X, Liang S, Cao Y, et al. Observed contrast changes in snow cover phenology in northern middle and high latitudes from 2001-2014[J]. Scientific Reports, 2015, 5(1): 16820.

[14] Qin Y, Hong C, Zhao H, et al. Snowmelt risk telecouplings for irrigated agriculture[J]. Nature Climate Change, 2022, 12(11): 1007-1015.

[15] Dai L, Che T. Spatiotemporal variability in snow cover from 1987 to 2011 in northern China[J]. Journal of Applied Remote Sensing, 2014, 8(1): 084693.

[16] Ma Q, Keyimu M, Li X, et al. Climate and elevation control snow depth and snow phenology on the Tibetan Plateau[J]. Journal of Hydrology, 2023, 617: 128938.

[17] Zhao Qin, Hao Xiaohua, He Dongcai, et al. The relationship between the temporal and spatial changes of snow cover and climate and vegetation in Northern Xinjiang from 1980 to 2019[J]. Remote Sensing Technology and Application, 2021, 36(6): 1247-1258.

[18] Xiao Xiongxin, Zhang Tingjun. Passive microwave remote sensing of snow depth and snow water equivalent: Overview[J]. Advances in Earth Science, 2018, 33(6): 590-605.

[19] Zhang Yin, Sun Congjian, Liu Geng, et al. Response of NDSI in the Tarim River Basin mountainous areas to climate change over the past 20 years[J]. Arid Zone Research, 2024, 41(10): 1639-1648.

[20] Li Hong, Li Zhongqin, Chen Puchen, et al. Spatio-temporal variation of snow cover in Altai Mountains of Xinjiang in recent 20 years and its influencing factors[J]. Arid Zone Research, 2023, 40(7): 1040-1051.

[21] Xiao X, Zhang T, Zhong X, et al. Support vector regression snow depth retrieval algorithm using passive microwave remote sensing data[J]. Remote Sensing of Environment, 2018, 210: 48-64.

[22] Wei Y, Li X, Gu L, et al. Significant decreasing trends in snow cover and duration in Northeast China during the past 40 years from 1980 to 2020[J]. Journal of Hydrology, 2023, 626: 130318.

[23] Yan Wei, Liu Jingshi, Luo Guangming, et al. Snow cover area changes in the Yurungkax River Basin of West Kunlun Mountains during 2000-2013 using MODIS data[J]. Progress in Geography, 2014, 33(3): 315-325.

[24] Jiang Kang, Bao Gang, Wulantuya, et al. Spatiotemporal changes of snow cover in Mongolian Plateau based on MODIS data[J]. Arid Land Geography, 2019, 42(4): 782-789.

[25] Zhang Gangdong. Spatio-temporal Differences of Vegetation Phenology and Snow Phenology in Snow Cover Area of Mongolian Plateau and Their Responses to Temperature[D]. Hohhot: Inner Mongolia Normal University, 2024.

[26] Yuan Zhihui. Analysis of Snow Dynamics and Their Influence on Spring Phenology in the Eurasian Steppe[D]. Hohhot: Inner Mongolia Normal University, 2024.

[27] Liu Zhonglin. Analysis of landscape ecological regions of the Mongolian Plateau[J]. Journal of Arid Land Resources and Environment, 1993, 7(Z1): 256-261.

[28] Gao Yanzhe. Analysis of Lake Area Changes in the Mongolian Plateau from 2000 to 2020[D]. Hohhot: Inner Mongolia Normal University, 2023.

[29] Sun Hui, Sa Chula, Meng Fanhao, et al. The spatiotemporal changes in snow cover of the Mongolian Plateau from 2000 to 2020 and analysis of its influencing factors[J]. Journal of Chifeng University (Natural Science Edition), 2022, 38(11): 1-6.

[30] Li Chenhao, Sa Chula, Liu Guixiang, et al. Spatiotemporal changes of snow cover and its response to climate changes in the Mongolian Plateau from 2000 to 2017[J]. Chinese Journal of Grassland, 2020, 42(2): 95-104.

[31] Hall D K, Riggs G A, Salomonson V V, et al. MODIS snow cover products[J]. Remote Sensing of Environment, 2002, 83(1-2): 181-194.

[32] Wang X, Xie H. New methods for studying the spatiotemporal variation of snow cover based on combination products of MODIS Terra and Aqua[J]. Journal of Hydrology, 2009, 371(1-4): 192-200.

[33] Hall D K, Kelly R E J, Riggs G A, et al. Assessment of the relative accuracy of hemispheric scale snow cover maps[J]. Annals of Glaciology, 2002, 34: 24-30.

[34] Chen X, Yang Y, Ma Y, et al. Distribution and attribution of terrestrial snow cover phenology changes over the Northern Hemisphere during 2001-2020[J]. Remote Sensing, 2021, 13(9): 1843.

[35] Harris I, Osborn T J, Jones P, et al. Version 4 of the CRU TS monthly high resolution gridded multivariate climate dataset[J]. Scientific Data, 2020, 7(1): 109.

[36] Wipfler E L, Metselaar K, Van Dam J C, et al. Seasonal evaluation of the land surface scheme HTESSEL against remote sensing derived energy fluxes of the Transdanubian region in Hungary[J]. Hydrology and Earth System Sciences, 2011, 15(4): 1257-1271.

[37] Balsamo G, A Beljaars, K Scipal, et al. A revised hydrology for the ECMWF model: Verification from field site to terrestrial water storage and impact in the integrated forecast system[J]. Hydrometeor, 2009, 10: 623-643.

[38] Kaur R, Kulkarni A V, Chaudhary B S. Using RESOURCESAT-1 data for determination of snow cover and snowline altitude, Baspa Basin, India[J]. Annals of Glaciology, 2010, 51(54): 9-13.

[39] Zhang Tingjun, Zhong Xinyue. Classification and regionalization of the seasonal snow cover across the Eurasian Continent[J]. Journal of Glaciology and Geocryology, 2014, 36(3): 481-490.

[40] Zhang Renhe, Zhang Ruonan, Zuo Zhiyan. An overview of wintertime snow cover characteristics over China and the impact of Eurasian snow cover on Chinese climate[J]. Journal of Applied Meteorogical Science, 2016, 27(5): 513-526.

[41] Huang Kunlin, Wu Guozhou, Xu Weixin, et al. Dynamic snowmelt process and its influencing factors in the eastern farmland region of Hulun Buir[J]. Arid Zone Research, 2024, 41(9): 1514-1526.

[42] Xue Hao, Yu Ruihong, Zhang Yanxia, et al. Temporal variation of runoff depth in the drainage basin of typical grassland area in Inner Mongolia: A case study of Xilin River Basin[J]. Science of Soil and Water Conservation, 2019, 17(2): 27-36.

[43] Li Peiji. Response of Xinjiang snow cover to climate warming[J]. Acta Meteorologica Sinica, 2001, 59(4): 491-501.