Ice, Snow and Environment Over High Asia Zone II Versions EN2 Vol 4 (4) 2019
A dataset of glacier length changes in the Chinese Altay Mountains (1959 – 2016)
: 2019 - 04 - 07
: 2019 - 07 - 02
: 2019 - 04 - 28
: 2019 - 10 - 22
1428 8 0
Abstract & Keywords
Abstract: Located in the northernmost part of Xinjiang Uygur Autonomous Region, the Altay Mountains host glaciers in China’s highest latitude. Over the past decades, glaciers in this region dramatically receded under the impact of global warming. In this study, we used GIS technology to extract the centerlines of the glaciers in different periods from data sources including the First Chinese Glacier Inventory (FCGI), the Second Chinese Glacier Inventory (SCGI), the 2016 Landsat OLI images and SRTM DEM (V4.1) data, among others, through which to obtain major geometrical parameters of the glaciers, including their maximum and mean lengths. This dataset consists of two parts: first, vector data of the glacier boundaries in 2016; second, vector data of the glacier centerlines in 1959, 2006 and 2016, respectively. Reflecting glacial changes in the Chinese Altay Mountains from 1959 to 2016, this dataset has important values for studies on regional glacier changes, climate changes, and so forth.
Keywords: glacier length; glacier centerline; glacier inventory; Altay Mountains
Dataset Profile
English titleA dataset of glacier length changes in the Chinese Altay Mountains (1959 – 2016)
Data authorsJin Huian, Yao Xiaojun, Zhang Dahong
Data corresponding authorYao Xiaojun (
Time range1959 – 2016
Geographical scope45°47′N – 49°10′N, 85°27′E – 91°01′E
Spatial resolution30 mData volume866 KB
Data formatESRI Shapefile file (compressed as *.zip)
Data service system<>
Sources of fundingNational Natural Science Foundation of China (Grant No. 41561016, 41861013 & 41801052); Earlier Career Research Promotion Program of the Northwest Normal University (Grant No. NWNU-LKQN-14-4).
Dataset compositionThe dataset consists of the following two subsets:
“”, which stores vector data of the glacier boundary over the Chinese Altay Mountains in 2016;
“”, which stores vector data of the glacier centerlines in 1959, 2006 and 2016, respectively.
1.   Introduction
As a major component of the cryosphere,1 glacier is a strong indicator and forecaster of climate change.2-4 Often measured as the maximum length of the glacier axis, glacier length is not only an important parameter to characterize the geometry of the glacier measured,5 but also a key element for climate change reconstruction, glacier-ice reserve assessment, glacier dynamics modeling, and glacial change prediction. While glacier length plays an important role in the study of glacial changes.6 traditional methods relying on manual labor for glacier length extraction have been disadvantaged by its low work efficiency and verification difficulty. With the development of GIS technology, it is now possible to apply automatic or semi-automatic approaches to data extraction. 7 For example, glacier length can be extracted by the glacier mainstream line method8-9 or the glacier centerline method.5,7,10 The mainstream line method measures glacier length by measuring the boundary of the catchment area of the glacier using hydrological analysis, 11 while the centerline method measures glacier length by measuring its centerline from the highest to the lowest point of the glacier based on its boundary.
Altay Mountains is located in central Asia and the northernmost point of Xinjiang Uygur Autonomous Region. It stretches about 2000 km from northwest to southeast, diagonally across the borders of China, Kazakhstan, Russia, and Mongolia. The Chinese Altay Mountains is located south of the mountains’ central ridge, north of Haba, Buerjin, Altay, Fuyun, and Qinghe, across a geographical range between 45°47′N – 49°10′N, 85°27′E – 91°01′E, totaling up to a length of 500 km. The highest point – Youyi Peak (4374 m) – is located at the source of the Kanas River upstream of Buerjin River at the junction of China and Mongolia,12 forming a high mountain junction with Kuitun Peak (4101 m) in the north, which is a concentrated area for modern glacier development in the Chinese Altay Mountains.13 The area is affected by the Arctic Ocean current in winter and spring, and replenished by westerly circulation in summer. With abundant rainfall, its precipitation decreases from northwest to southeast.12,14 Glacial meltwater over the mountains, especially seasonal snow cover, accounts for 45% – 50% of the water sources that feed local rivers.13 Therefore, research on glacial changes is of great significance not only for understanding climate change in the region, but also for rationally utilizing water resources in the downstream areas. While a retreating trend has been witnessed in the overall glacial area over the Chinese Altay Mountains from 1959 to 2009,14-16 there is a lack of systematic knowledge about the extent of the glacier retreat, especially changes at the terminus of glacier. This study presents a collective endeavor to build a dataset of glacier length changes in the Chinese Altay Mountains from 1959 to 2016. Using the glacier centerline extraction method, 6 we drew data sources from the First Chinese Glacier Inventory (FCGI), the Second Chinese Glacier Inventory (SCGI), and the 2016 glacier vector data. The dataset can be used to support research on glaciers, climate changes and water resources of the region.
2.   Data collection and processing
2.1   Data sources
The dataset drew sources from FCGI, SCGI, Landsat OLI remote sensing images and Digital Elevation Model (DEM) data. Specifically, among the data we used, the 1:100,000 aerial photogrammetric maps (1959) were from FCGI; the Landsat TM/ETM+ remote sensing images and 1: 100,000 topographic maps were from SCGI; the 2016 glacier vector data were from Landsat OLI remote sensing images of the three scenes provided by the United States Geological Survey (USGS) (; the SRTM DEM data, with a spatial resolution of 90 m, were downloaded from Geospatial Data Cloud ( (Table 1).
Table 1   List of data sources used for building this dataset
NoSourceIDTimeResolution (m)
3SRTM DEMV4.1200090
4Landsat OLILC81440262016224LGN002016-08-1130
5Landsat OLILC81420272016226LGN002016-08-1330
6Landsat OLILC81430262016249LGN012016-09-0530
2.2   Data processing
Our collection and processing of the glacial vector data abided by the methods and specifications as per SCGI.17 The vector boundaries of the glaciers were extracted from Landsat OLI remote sensing images, through a combined use of band ratios and manual interpretation, which were then validated and modified by reference to topographic maps and Google Earth. The modified glacier boundaries were then segmented by the ridgeline automatic extraction method to obtain the glacier vector data.18-19 The glacier length was acquired by the centerline extraction method as proposed by Yao et al.6 Firstly, determine the highest and lowest points along the boundary of each glacier, based on which the glacier boundary was divided into independent yet connected line segments. Then, divide the plane into multiple regions by calculating the Euclidean distance, and the common edge would be the glacier centerline. Finally, calculate the length of each glacier based on the glacier centerline obtained therefrom, and identify the elevation of the highest and lowest points along the glacier boundary. The data processing flow is shown in Figure 1. For a higher work efficiency, we used ArcPy packet to write program scripts and converted them into .tbx files that could be read and edited by ArcGIS.

Figure 1   Flowchart illustrating glacier centerline extraction and glacier inventory generation
Affected by topographical factors, the Altay Mountains are dominated by smaller-scale single-basin and single-exit glaciers. According to the SCGI, 94% of the glaciers in the region are smaller than 2.0 km2,14 and there are only a small number of large-scale composite-basin and single-exit glaciers, such as the Kanas Glacier. In the case of the latter, the length of each tributary was calculated to obtain the mean length of all tributaries and the length of the longest tributary, respectively.
3.   Sample description
3.1   Data graphics sample
Single-basin and single-exit glacier
Figure 2 shows the changing centerline of a single-basin and single-exit glacier (GLIMS code: G087717E49104N) in different periods. Overall, the glacial area decreased from 0.41 km2 in 1959 to 0.35 km2 in 2016. Accordingly, the glacial length shrank from 1262 m in 1959, 1224 m in 2006 to 1219 m in 2016, with a total of 43 m (−3.41%) shrinkage.

Figure 2   Highest point, lowest point and centerline of a single-basin and single-exit glacier
Composite-basin and single-exit glacier
Figure 3 shows a receding trend of the Kanas Glacier from 1959 to 2016, with a total area reduction of 6.31 km2. In the meantime, the elevation of the terminus rose from 2409 m to 2547 m. The length of each tributary decreased, and that of the longest (tributary No. 5) decreased by 1663 m. Their mean length reduced from 9326 m in 1959 to 7947 m in 2016, by about 14.8%.

Figure 3   Highest point, lowest point and centerline of a composite-basin and single-exit glacier
3.2   Data attribute description
The data table is composed of 12 fields (Table 2), which designate the coordinate parameters, geometric parameters and auxiliary information of each glacier. Among them, identification code (ID) is the unique identifier assigned to each glacier used for identification purposes throughout data extraction and processing. Glacier code (GLIMS_ID) consists of 14 numbers and characters, abiding by the coding and formatting rule as per SCGI. Watershed code (CGI_ID) is drawn from FCGI, representing the sequence of clockwise rotation of glaciers in the drainage basin from the outlet. Mean length (Mean_Len) designates the mean length of all tributaries of a glacier while maximum length (Max_Len) denotes the length of the longest tributary of a glacier (the two are the same for single-basin and single-exit glaciers). Centerline count (Count) refers to the number of centerlines of each glacier – The number is 1 for all single-basin and single-exit glaciers, whereas for composite-basin and single-exit glaciers it is determined by the number of tributaries. Maximum elevation (Max_Elev) and minimum elevation (Min_Elev) record the elevation of the highest and lowest points of a glacier. Table 2 is a description of the data fields and their attributes.
Table 2   Data fields and field attribute description
1IDDouble6Identification code
2GLIMS_IDText14Glacier code
3CGI_IDText10Watershed code
4AreaDouble10Area of the glacier
5Mean_LenDouble10Mean length of the tributaries
6Max_LenDouble10Length of the longest tributary
7CountShort Integer4Number of the tributaries
8Max_ElevShort Integer4Elevation of the highest point
9Min_ElevShort Integer4Elevation of the lowest point
10DataSourceText30Data source
4.   Quality control and assessment
The glacier vector data were extracted through a combined use of band ratio and manual visual interpretation, and the interpretation accuracy was controlled within one pixel. The glacier length data were extracted by the glacier centerline method as per Yao et al.6 While automatic extraction alone was sufficient to achieve satisfactory results for single-basin and single-exit glaciers, the auxiliary reference line method6,17-18 was used to supplement automatic data extraction for composite-basin and single-exit glaciers which often have a vast glacier accumulation area or multiple tributaries, and to rectify errors caused by partial deviation of the centerline from the glacial material flow.
5.   Value and significance
As mentioned earlier, this dataset drew sources from FCGI, SCGI, 2016 glacier vector data and SRTM DEM data. Compared with other glacier inventories, this dataset aims to characterize the glacier length changes in the study region from 1959 (395 entries included) to 2016 (273 entries included). Its reliability is ensured through a combined use of automatic data extraction and manual data revision. This dataset can be used to analyze the spatiotemporal characteristics of the glaciers in the study area and the glacial responses to climate change. It can also be used to reveal a quantitative correlation between glacial area and length changes in the region, which would help unpack the response mechanism of regional glacial changes to climate change.
6.   Data usage and recommendations
The vector data in this dataset are stored in ESRI Shapefile, under WGS84 geodetic coordinate system and Albers equal area projection coordinate system. The data can be opened, viewed and edited in GIS software (e.g., ArcGIS) or remote sensing software (e.g., ENVI).
Thanks go to the U.S. Geological Survey and Geospatial Data Cloud for providing the Landsat OLI and DEM data, and the “Survey of Glacier Resources and Changes in China” project team for providing the CGI data.
Li X, Cheng G, Jin H et al. Cryospheric change in China. Global and Planetary Change 62(2008): 210–218.
Oerlemans J. Quantifying global warming from the retreat of glaciers. Science 264(1994): 243–244.
Xie ZC & Liu CH. Introduction to Glaciology. Shanghai: Shanghai Science Popularization Press, 2010.
Liu SY, Yao XJ, Guo WQ et al. The contemporary glaciers in China based on the Second Chinese Glacier Inventory. Acta Geographica Sinica 70(2015): 3–16.
Machguth H & Huss M. The length of the world’s glaciers: a new approach for the global calculation of center lines. The Cryosphere 8(2014): 1741–1755.
Yao XJ, Liu SY, Zhu Y et al. Design and implementation of an automatic method for deriving glacier centerlines based on GIS. Journal of Glaciology and Geocryology 37(2015): 1563–1570.
Lebris R & Paul F. An automatic method to create flow lines for determination of glacier length: A pilot study with Alaskan glaciers. Computers & Geosciences 52(2013): 234–245.
Leclercq PW, Oerlemans J, Basagic HJ et al. A data set of worldwide glacier length fluctuations. TheCryosphere 8(2014): 659–672.
Paul F, Barry RC, Cogley JG et al. Recommendations for the compilation of glacier inventory data from digital sources. Annals of Glaciology 50(2010): 119–126.
Kienholz C, Rich JL, Arendt AA et al. A new method for deriving glacier centerlines applied to glacier in Alaska and northwest Canada. The Cryosphere 8(2014): 503–519.
Yang BY, Zhang LX, Gao Y et al. An integrated method of glacier length extraction based on Gaofen satellite data. Journal of Glaciology and Geocryology 38(2016): 1615–1623.
Wang LL, Liu CH & Wang P. Modern Glaciers in Altay Mountains of China. Acta Geographica Sinica 40(1985): 142–154.
Liu CH, You GX & Pu JC. Glacier Inventory of China (II): Altay Mountains. Lanzhou: Lanzhou Institute of Glaciology and Cryopedology. Academia Sinica Press, 1982.
Yao XJ, Liu SY, Guo WQ et al. Glacier Change of Altay Mountain in China from 1960 to 2009—Based on Second Glacier Inventory of China. Acta Geographica Sinica 27(2012): 1734–1745.
Wang SH, Xie ZC, Dai YN et al. Structure, change and its tendency of glacier systems in Altay Mountains. Arid Land Geography 34(2011): 115–123.
Bai JZ, Li ZQ, Zhang MJ et al. Glacier changes in Youyi Area in the Altay Mountains of Xinjiang during 1959 – 2008. Arid Land Geography 35(2012): 117–124.
Guo WQ, Liu SY, Xu JL et al. The second Chinese glacier inventory: data, methods and results. Journal of Glaciology 61(2015): 357–372.
Guo WQ, Liu SY, Yu PC et al. Automatic extraction of ridgelines using on drainage boundaries and aspect difference. Science of Surveying and Mapping 36(2011): 191, 210–212.
Wu KP, Liu SY & Guo WQ. A dataset of glacier distribution and glacier changes in the Kangri Karpo Mountains during 1980 – 2015. China Science Data 4 (2019). DOI: 10.11922/csdata.2018.0013.zh.
Data citation
1. Jin H, Yao X & Zhang D. A dataset of glacier length changes in the Chinese Altay Mountains (1959 – 2016). Science Data Bank. DOI: 10.11922/sciencedb.753.
Article and author information
How to cite this article
Jin H, Yao X & Zhang D. A dataset of glacier length changes in the Chinese Altay Mountains (1959 – 2016). China Scientific Data 4(2019). DOI: 10.11922/csdata.2019.0013.zh.
Jin Huian
raw data collection and processing, manuscript writing.
MSc; research area: environmental remote sensing.
Yao Xiaojun
design of the overall scheme.
PhD, Professor; research area: GIS and cryospheric change.
Zhang Dahong
data processing and coding.
MSc; research area: GIS design and development.
National Natural Science Foundation of China (Grant No. 41561016, 41861013 & 41801052); Earlier Career Research Promotion Program of the Northwest Normal University (Grant No. NWNU-LKQN-14-4).
Publication records
Published: Oct. 22, 2019 ( VersionsEN2
Released: April 28, 2019 ( VersionsZH1
Published: Oct. 22, 2019 ( VersionsZH2