1. Introduction
Glaciers are one of the crucial components of the cryosphere and are a dynamic and scarce natural resource. Mountain glaciers are widely distributed in middle and high latitudes, with high sensitivity and feedback to climate change. They are known as natural “recorders” and “early warners” [
1]. Glacier meltwater has an obvious recharge and regulating function for many river runoffs and is known as an alpine “solid water reservoir”. China is one of the countries with the most mountain glaciers, but it also features some of the world’s desert areas and is considered a water-poor country. Glaciers and their meltwater are essential freshwater resources in the arid region of northwest China. They are the “lifeline” for the health of the regional ecological environment and the sustainable socio-economic development of the oasis region [
2]. Xinjiang ranks second in China after Tibet in terms of total area and number of glacier resources and first in terms of ice reserves [
1]. Glaciers play an essential role in the composition of water resources in Xinjiang, and their variation has a profound impact on Xinjiang and Central Asia.
Under the dual influence of climate change and human activities, the processes and effects of regional ecohydrology, resources and environment, and natural disasters caused by glacier structure and scale changes become more and more apparent [
3]. Since the middle of the 20th century, the global mean surface temperature (GMST) has continued to rise, and the precipitation in the 30–60° N zone has increased significantly [
4]. During this period, northwest China’s average annual temperature rise reached 0.37 °C/10a. Under this climate background, glacier resources in Xinjiang have continued to shrink and become thinner [
5]. The high mountains of Xinjiang, including Tian Shan, Altai Shan, Kunlun Shan, Karakorum Shan, and Pamir Plateau, provide a vast space for developing surviving glaciers. Due to the diversity of hydrothermal conditions and complex terrains in the southern, northern and central mountain systems (groups) with large latitude spans, glaciers vary in form and size.
After the Research Team of Alpine and Ice Utilization of the Chinese Academy of Sciences was established in 1958, glaciologists such as Shi, Y.F. began to investigate and conduct scientific research on glaciers in the Tianshan mountains [
6]. So far, research on glacier changes in the Tianshan mountains has achieved remarkable results. Chen, H. [
7], Zhao, Q. [
8], Wang, S. [
9] and Xing, W. [
10] studied the temporal and spatial variation of the glacier area in the Tianshan mountains and its response to climate and showed that the glacier area in Tianshan mountains in China has shrunk by 11.5–18.4% in the most recent 50 years, and there is a great difference in the rate of decline in the eastern and western parts of the Tianshan mountains before and after 2000. Meanwhile, field observations and in-depth studies were made on the glacier area, mass balance, and glacier hydrometeorology of a single typical glacier in the Tianshan mountains, such as Tianshan No. 1 glacier [
11], Hasilegen Glacier 51 [
12], Qingbingtan Glacier 72 [
13] and Hami Miaergou flat-topped glacier [
14]. A large number of studies have also been carried out on the changes of glacier meltwater runoff in many inland river basins, such as the Manas river basin [
15], the Aksu river basin [
16] and the Kuitun river basin [
17,
18]. With the development and application of 3S technology, glacier research’s spatial and temporal limitations by traditional means have been gradually broken. The “high spatial and temporal resolution” remote sensing monitoring of glaciers has entered a new era. Studies on glacier changes in Xinjiang also involve the Altai mountains in the north and the Pamir Plateau, Karakoram, Kunlun, and Altun mountains in the south (referred to as the Pakakuna mountain group). For example, Lv, H. [
19], Huai, B. [
20], Xu, C. [
21], Lv, M. [
22], Ke, L. [
23], Yu, X. [
24] and Liu, S. [
25] have analyzed the glacier changes and their causes in the Alta mountains, Saurshmusi Island, the East Pamir-West Kunlun region and the Altun mountains. Comparatively, studies on glaciers in Xinjiang other than the Tianshan mountains are limited, especially the glaciers and their ice reserve changes in the Pakakuna mountain group. This is particularly important and urgent for the hydrological processes and water resources security of the Tarim basin, where glacial meltwater runoff recharge accounts for up to 43.3% [
26].
In Xinjiang, 98% of surface water resources come from mountainous areas, and the “triadic” (rainfall, snow, glacier) flow-production model is a common feature of many watersheds in the region [
27]. Glacier meltwater has a relatively stable replenishment and regulation effect on river runoff, and climate change has a significant impact on glacier ablation/accumulation. The process of glacier change and mass balance is of great importance to the basin’s ecohydrology and water resources use. In the past, researchers have investigated a lot of glacier changes in most mountains and river basins in the study area. However, it is necessary to analyze the distribution and variation of glacier resources in mountain systems, watersheds, and prefectures and investigate further the response characteristics and sensitivity of glaciers to hydrothermal conditions and topography. This study analyzes the characteristics of glacier changes in Xinjiang over the past 50 years based on Chinese glacier inventory, combined with GIS spatial analysis techniques. The responses of glacier changes to topographic features and hydrothermal conditions are discussed to provide references for research and practice on ecological security and water resources utilization in Xinjiang.
2. Study Area
Xinjiang (34°22′~49°10′ N, 73°40′~96°23′ E) is located at the northwest border of China and is the largest administrative province in China (
Figure 1). Located deep in the hinterland of the Eurasia continent, the high mountains and vast basins form a unique landform pattern of three mountains sandwiched by two basins. Coupled with the arid climate characteristics, the hydrological and water resources cycle process is more complex. The mountains are broad and have a wide range of elevations, providing a special topography for the development of mountain glaciers. Glacier resources in Xinjiang are mainly distributed in the Altai mountains in the north, the Tian Shan mountains in the central part of the province, and the Pamir, Karakorum, and Kunlun mountains in the south. It belongs to the Irtysh river watershed, Junggar basin, and Tarim basin, surrounded by high mountains. Xinjiang is located in the northwest desert area, where there is little rain and intense evaporation. Surrounded by high mountains where it is difficult for marine moisture to enter, it forms a distinct temperate continental climate. The average summer temperature can reach 18.24 °C, and the average annual precipitation is about 130 mm. However, the temperature and precipitation vary significantly from place to place, with the temperature in the south being higher than that in the north, and the precipitation is the opposite. The unique topographic advantages and climatic conditions provide conditions for the development of glaciers. Xinjiang’s total surface water resources are about 7.93 × 10
10 m
2, ranking only 12th in China, and glacial meltwater is the primary recharge source for many river runoffs [
1].
3. Data and Methods
The data in this study mainly include the Chinese Glacier Inventory, topographic and geomorphological data, and meteorological data (
Table 1). The National Cryosphere Desert Data Center provided the data for the two periods of glacier inventory, and the cataloging methods are described in the literature [
1,
28]. For the first glacier inventory, topographic data such as elevation and orientation were extracted from single glaciers and assigned to each glacier; DEM and 1:1,000,000 geomorphological data were analyzed using zonal statistics of the elevation range and the topographic relief size of different mountain glaciers. Monthly precipitation and temperature data of the stations from 1961 to 2010 were selected as climate change indicators. Geostatistical analysis was carried out on the original data, and the ordinary kriging method with the lowest error was fixed for spatial interpolation. In order to facilitate comparative analysis, meteorological stations located in and around the mountains are divided into three regions: north, middle, and south. The annual or summer average temperature and precipitation increase or decrease in each mountain system (group) are calculated. The trends of their time series can identify the response of glacier changes to the climate.
3.1. Glacier Area Change and Ice Reserves Estimation
Glacier area change is the most direct and specific means to assess changes in glacier resources. Due to the significant difference in the time interval between the first and second glacier inventory in Xinjiang, this study establishes the correspondence between the number of glaciers in the two glacier inventories by considering the number of omissions in the first glacier inventory and the number of increases and decreases in the second glacier inventory due to splitting or merging according to Li, L. [
4]. Sun, M. [
29] calculated the glacier area change rate and relative rates in different regions of the study area. Ice storage is an essential indicator for assessing glaciers and their changes and is a vital driving parameter for constructing glacier hydrological models [
30]. Only a few glaciers in the world are currently available with relatively accurate thickness and volume data. Most ice reserves are still estimated mainly by indirect and fast empirical formulas, which are also commonly used to estimate glacier ice reserves in larger spatial scales [
31]. Most of the current calculations of ice reserves use the volume–area empirical formula.
where:
V is ice reserves (km
3);
A is glacier area (km
2),
c and gamma are empirical coefficients. In this paper, the coefficients proposed by Radić et al. [
32], Grinsted [
33], and Liu et al. [
34] are used to calculate the glacier reserves in Xinjiang, and the average value of the three methods is used as the result.
3.2. Response Analysis of Glacier Change to the Topography and Climate
Glacier variability is not only directly affected by hydrothermal conditions but also obviously constrained by topographic conditions. Geodetectors can test both the spatial heterogeneity of a single variable and the possible causal relationship between two variables by trying the coupling of their spatial distribution [
35]. To explore the contribution of topographic factors and meteorological factors to the change of the glacier area, Geodetector software was applied. For each factor, the data were divided into five levels using the natural breaks (jenks). The Geodetector model expression is:
where
L is the stratification of the independent and dependent variables,
and
are the variances of the independent variable topographic or meteorological factors and the dependent variable glacier area change. The
q takes values in the range of [0, 1], and an enormous value of
q indicates a more substantial explanatory power of the factors in glacier change. In the extreme case, a
q value of 1 indicates that the factors control the spatial distribution of glacier change entirely, and a
q value of 0 indicates no relationship between them.
In meteorological data, temperature and precipitation data processed by spatial interpolation are used to calculate air temperature and precipitation trend rates. The average temperature tendency, average precipitation tendency, summer temperature, and summer precipitation tendency are extracted for a single glacier. This paper uses the slope analysis method to analyze the spatial interannual variation trend of temperature and precipitation in Xinjiang. The calculation formula is as follows:
In the formula, is the interannual rate of change; n is the number of years from 1960 to 2010, and this study takes 50. The b is the time series, from 1960 to 2010 in turns 1–51. The is the temperature or precipitation in the i-th year. < 0 and > 0 indicate increasing and decreasing with time, respectively, during the study period. The larger the positive value, the faster the temperature or precipitation rises, and the smaller the negative value, the quicker it decreases.
Terrain factors include elevation, slope, aspect, and topographic relief. Topographic relief is the difference between the highest and lowest altitude, an essential index for classifying geomorphic types. Based on DEM data, using the neighborhood analysis method in GIS, the topographic relief of a single glacier is extracted under a 3 × 3 rectangular analysis window.