A Dataset of Desertification Distributions along the China–Pakistan Economic Corridor 2000 – 2017
>>
： 2018 - 09 - 11
： 2018 - 09 - 23
： 2019 - 08 - 02
171 1 0
Abstract & Keywords
Abstract: Desertification has caused severe ecological environment problems worldwide. It is very important to monitor the dynamic change of land desertification and grasp its change rule. The China-Pakistan Economic Corridor is under execution smoothly for the last couple of years and most of the early harvest projects have been completed or are at an advanced stage of completion. Also, it is a critical part of the One Belt And One Road (OBOR) initiative. The Corridor is about 3,000 kilometers long, covering more than 1,300 kilometers of permafrost, arid and desert regions. Especially its southern section is characterized by drought and large-scale desert, which constitutes the Corridor’s major ecological constraints. In this study, desertification difference index (DDI) was used to evaluate the degree of desertification in the Corridor. Based on Normalized Difference Vegetation Index (NDVI) and Albedo, we established the Albedo-NDVI space and a DDI formula through negative correlation between Albedo and NDVI. Based on DDI, we built the dataset of desert distributions along the China-Pakistan Economic Corridor during 2000–2017. The quality and accuracy of the dataset were verified by vegetation coverage inversed by Landsat ETM+. Taking 2010 data as an example, the overall evaluation accuracy reached 81.67%, and the Kappa coefficient was 75.42%. The dataset reflects the degree of desertification along the Corridor, providing information for desertification prevention and control in the countries around.
Keywords: China-Pakistan Economic Corridor; desertification; Albedo-NDVI space; desertification difference index (DDI); Albedo
Dataset Profile
 Title A dataset of desert distributions along the China-Pakistan Economic Corridor 2000 – 2017 Data corresponding author Min Yufang (myf@lzb.ac.cn) Data authors Min Yufang, Feng Keting, Kang Jianfang, Ai Minghao Time range From 2000 to 2017 Geographical scope 23°47′N–40°55′N, 60°20′E–80°16′E Spatial resolution 1 km Data volume 569 MB Data format TIF Data service system http://www.crensed.ac.cn/portal/metadata/0464e947-96a0-452d-903b-a4040d6debba Sources of funding Data sharing fundamental program for Construction of the National Science and Technology Infrastructure Platform (Y719H71006); The 13th Five-Year Informatization Plan of Chinese Academy of Sciences (XXH13506, XXH13505). Dataset composition This dataset contains eighteen Albedo, NDVI and DDI raster data of the China-Pakistan Economic Corridor from the year 2000 to 2017. Each folder corresponds to a specific year, and names after the year. Each folder contains three files: “Albedo_ year (min).tif” stores the smallest Albedo raster data; “NDVI_ year (Max).tif” stores the maximum NDVI raster data; “DDI_ year.tif” stores DDI raster data.
1.   Introduction
Desertification, including land degradation in drought, semi-drought and drought-sub humid regions due to various factors including climate change and human activities, is a destructive biogeographic process that typically results in reduced biodiversity, reduced soil fertility and even loss of ecological capacity[1][2] . Land desertification is one of the most serious ecological and environmental problems in the world. According to statistics, more than 100 countries are being affected by land desertification. Desertification not only destroys the ecological environment, but also weakens social and economic development. Monitoring the dynamic changes of land desertification and mastering the law of change are of great significance for combating desertification.
The starting point of the China-Pakistan Economic Corridor is in Kashgar, Xinjiang, China. The end point is in Gwadar Port, Pakistan. The “Silk Road Economic Belt” is in the north and the “21st Century Maritime Silk Road” is in the south, the Corridor is the key hub for the North-South Silk Road. At the same time, it is a trade corridor covering roads, railway, oil-natural gas and fiber optic channels, and is the important component of “Belt and Road” initiative. The China-Pakistan Economic Corridor is about 3,000 kilometers long, of which more than 1,300 kilometers are severely cold plateau, arid and desert sections. Especially in the southern section, drought and large-scale desert are the main ecological constraints for this region.
Remote sensing technology provides people with a new means of desertification monitoring. It has the characteristics of wide observation range, large amount of information, fast data updates and high precision. It can accurately and timely acquire the changed information of the desertified land of Xingjiang through remote sensing image interpretation or quantitative inversion. Desertification is represented by remote sensing images as the enhancement of bare ground information and the attenuation of vegetation information. It can be characterized by surface albedo, surface temperature, surface moisture, vegetation index and vegetation coverage as indicators. The Normalized Vegetation Index (NDVI) is an important biophysical parameter reflecting the state of surface vegetation, while the surface albedo (Albedo) is a physical parameter reflecting the surface reflection characteristics of solar short-wave irradiation. With the aggravation of desertification, the surface vegetation suffered severe damage, the coverage of surface vegetation decreased, the biomass decreased, and the surface roughness increased. On the remote sensing image, the NDVI value decreased correspondingly, and the surface albedo increased correspondingly. Li[3]et al’s research results by positional observation have shown that when the albedo of the local surface reaches a certain value, grassland desertification will occur, and the surface albedo threshold of desertification onset is 0.3. The two indicators of Albedo and NDVI are related to the desertification process by characterization of bare surface information and vegetation coverage information, respectively, and exists a good correlation. Zeng Yongnian et al.[4]proposed not to rely on a single parameter, but to construct the "NDVI-Albedo" feature space through the negative correlation between NDVI and Albedo, and calculate the desertification difference index (DDI), and use DDI to carry out desertification categorization. The method is simple to use and easy for obtaining the indicators. In the classification and categorization of desertification, the method of classification using only remote sensing spectral information is more accurate, so it has been applied in the quantitative evaluation of desertification in different regions in recent years[4][5][6][7] . This paper uses DDI to evaluate the degree of desertification in the China-Pakistan Economic Corridor. The study zone is shown in Figure 1. Using NDVI and Albedo as monitoring indicators, the Albedo-NDVI feature space was constructed, and the negative correlation between Albedo and NDVI was used to construct the DDI formula of the China-Pakistan Economic Corridor, and 18 sessions of dedicated data sets for China-Pakistan Economic Corridor desertification classification was completed for 2000–2017, which offer technical support for governance and decision-making by relevant authorities. By using the remote sensing technology, we can predict the change trends for desertified lands in the drought, semi-drought and drought-sub humid Land at the beginning of the twenty-first century. Based on the data obtained through the research, dynamic monitoring of desertified Land of Xinjiang can be realized.
2.   Data Collection and Processing Method
2.1   Data Source and Pre-treatment
The preparation of desertification distribution data in China-Pakistan Economic Corridor is based on two kinds of data: MODIS vegetation index product MOD13A3 data and surface albedo product MCD43A3 data. The data verification uses Landsat 7 data. The detailed list is shown in Table 1.
Table 1   List of data use by this study
EntryTitleTimeSourceType
1MOD13A32000–2017LPDAACGrid
2MCD43A32000–2017LPDAACGrid
3Landsat 72010USGSGrid
Albedo is an important parameter in remote sensing inversion, which refers to the ratio of the total reflected irradiation flux to the total incidence solar irradiation flux per unit time and unit area. With the increase of desertification degree, the surface morphology changed significantly, that is demonstrated in the decline of surface vegetation coverage, soil organic matter content decreases, soil moisture decreases, surface roughness increases, and albedo increases. The MCD43A3 is a data product that combines the Terra and Aqua satellites. It is a day-scale L3 product with a space-time resolution of day/500 m. The MCD43A3 product data includes white space albedo and black air albedo of the local noon sun angle, and both white albedo and black air albedo contain 7 narrow bands as bands 1–7 and 3 wide-band visible lights (0.3 ~0.7 μm), near-infrared (0.7 to 5.0 μm) and short-wave (0.3 to 5.0 μm). The data comes from the National Aeronautics and Space Administration (NASA) Land Process Distributed Data Archives Center (LPDAAC). The data is stored in HDF-EOS format with sinusoidal projection. Covering the entire China-Pakistan corridor requires 6 scenes and 6 bands of data. This study obtains a total of 39,090 scenes of MCD43A3 data for the period of 18 years from 2000–2017. Firstly using the MRT tool to do data mosaic and resampling, and the 500 m resolution data is resampled to 1 km, and all data is converted to WGS84 by sinusoidal projections. Then use the China-Pakistan Corridor vector border and Gdal's warp tool for data cutting.

Fig. 1   Sketch of geographical location of the study zone
NDVI is an important indicator for the detection of plant biomass and vegetation. It is mainly used to monitor vegetation growth status, vegetation coverage and eliminate some radiation errors. In this paper, the 1 km MODIS monthly synthetic vegetation index product MOD13A3 is used. MOD13A3 data includes 1 km monthly synthetic NDVI, enhanced vegetation index EVI, NDVI quality file, EVI quality file, visible red light band reflectance, near infrared reflectance, visible blue light band reflectance, mid-infrared reflectance, average observation zenith angle, average solar zenith angle and average azimuth. The data comes from LPDAAC. The data is stored in HDF-EOS format with Sinusoidal projection. This study obtained a total of 1296 scenes of MOD13A3L3 data for 18 years from 2000 to 2017. Firstly using the MRT tool to do data mosaic and resampling, and the 500 m resolution data is resampled to 1 km, and all data is converted to WGS84 from sinusoidal projection coordinates. Then use the China-Pakistan Corridor vector border and Gdal's warp tool for data cutting, and finally get the NDVI data of the 216 scenes of the research zone. Landsat7 ETM+ data is also used in this paper for data quality assessment. The Landsat 7 ETM+ image consists of 8 bands, the spatial resolution of band 1–band 5 and band 7 is 30 m, the spatial resolution of band 6 is 60 m, the spatial resolution of band 8 is 15 m, and the time resolution is 16 days. In this study, 8 scenes and 7 images of Landsat 7 with a cloud cover of less than 5% were obtained from July to August 2010. Using ENVI software to perform geometric correction and atmospheric correction on the data as well as NDVI calculation and statistics, and finally use the pixel binary model to estimate the vegetation coverage by remote sensing to obtain the vegetation coverage data of the study zone. The NDVI and Albedo raw data have outliers and missing pixels, and the quality is not very reliable. Invalid value culling and spatial interpolation are required. The valid range of NDVI data is −2000 to 10000, and the invalid value is filled with −3000. The 1–6 white space albedo band used in the MCD43A3 product has a valid value range of 0 to 1000, and the invalid value is filled with 32766. In this paper, the inverse distance weighting (IDW) method is used to realize the spatial difference of data. The IDW method parameters are set as follows:
(1) Distance index: Usually ranges from 0.5 to 3, and the value chosen in this paper is 2;
(2) Search radius: Define the input point for interpolating the missing pixel value, select the variable search radius mode, and the number of nearest neighbor input sampling points for interpolation is 12.
2.2   Data Processing Procedure
2.2.1 Reconstruction of NDVI and Albedo Data Annual Scale
The change of vegetation coverage is the most intuitive manifestation of desertification. It is necessary to evaluate the degree of desertification under the most lush vegetation time in a year. Therefore, based on the annual desertification index constructed by the annual NDVI and the annual Albedo, it is necessary to use the annual maximum of NDVI and the annual minimum of Albedo as the basic data. The calculation process includes the following aspects:
（1）Albedo Data Calculation
The conversion of MODIS narrow-band albedo to wide-band albedo is based on the algorithm of Liang[8][9] . The short-wave albedo calculation method is used. The formula is as follows:
$${\alpha }_{short}=0.16{\alpha }_{1}+0.291{\alpha }_{2}+0.243{\alpha }_{3}+0.116{\alpha }_{4}+0.112{\alpha }_{5}+0.081{\alpha }_{6}-0.0015$$ (1)
Among them, αshort is short-wave albedo, α1–α6 represents bands 1–6 in MCD43A3, respectively. In this study, white space albedo is used to calculate Albedo, because white space albedo is the integral of each incident angle, which is closer to the surface albedo in general meaning.
（2）Calculation of Annual NDVI Data
Maximum Value Compositing (MVC) is used. This method is mainly used for remote sensing data processing, and is mainly used for data analysis and reconstruction of pixels in a certain period of time. Calculation formula:
$${NDVI}_{i}=\underset{1\le \mathrm{j}\le \mathrm{n}}{\mathrm{max}}{NDVI}_{ij}$$ (2)
Where i is the pixel name, j is the time point of the [1, n] time interval, and NDVIij refers to the NDVI value of the pixel i at the time j.
（3）Calculation of Albedo Annual Data
The annual reconstruction of Albedo data uses the minimum synthesis method. The method selects the minimum value of the pixel as the new pixel value within a specified period of time. The calculation formula of the annual Albedo is:
$${Albedo}_{i}=\underset{1\le \mathrm{j}\le \mathrm{n}}{\mathrm{min}}{Albedo}_{ij}$$ (3)
Where i is the pixel name, j is the time point of the time interval [1, n], and Albedoij refers to the Albedo value of the pixel i at the time j.
2.2.2 Constructing Desertification Index based on Albedo and NDVI
NDVI values are directly proportional to vegetation coverage, while Albedo is inversely proportional to vegetation coverage. According to the positive correlation between NDVI and vegetation coverage, and the negative correlation between NDVI and Albedo, the two-dimensional （2D）spatial feature map of Albedo-NDVI can be obtained. In this paper, taking 2010 data as an example, the NDVI and Albedo data are normalized firstly. The normalization processing formula is as follows:
$$\mathrm{N}\mathrm{D}\mathrm{V}\mathrm{I}=\mathrm{N}\mathrm{D}\mathrm{V}\mathrm{I}/10000$$ (4)
$$\mathrm{A}\mathrm{l}\mathrm{b}\mathrm{e}\mathrm{d}\mathrm{o}=\mathrm{A}\mathrm{l}\mathrm{b}\mathrm{e}\mathrm{d}\mathrm{o}/1000$$ (5)
Then, 1500 sample points are evenly selected in the study zone, and NDVI is used as the X axis, and Albedo is used as the Y axis to construct the NDVI-Albedo feature space. The feature space scatter diagram is shown in Figure 2. According to the results of linear statistical regression analysis, the negative correlation between Albedo and NDVI can be expressed by the following linear relationship:
$$Albedo=-0.2303\mathrm{N}\mathrm{D}\mathrm{V}\mathrm{I}+0.3258$$ (6)

Figure 2   Two-dimensional (2D) feature space map of Albedo-NDVI in China-Pakistan Economic Corridor 2010
As shown in Fig. 2, in the feature space constructed by Albedo and NDVI, the different positions of the negative correlation between Albedo and NDVI represent the state and degree of different stages of desertification, and the degree of desertification increases with the decrease of NDVI value and is up with the rise of Albedo value, that is, the negative correlation linear expression of Albedo and NDVI can reflect the trend of desertification. According to the conclusions of Verstraete and Pinty[10], if the Albedo-NDVI feature space is divided in the vertical direction representing the trend of desertification, different desertification land can be effectively distinguished as shown in Figure 3, and the desertification difference index model DDI is used to represent:
$$DDI=\alpha ×NDVI-Albedo$$ (6)

Fig.3   Graphical expression of desertification difference index (DDI) in NDVI-Albedo space
In a specific application, the constant α can be determined from the slope in the Albedo-NDVI feature space, where k is the slope of the characteristic equation.
$$\alpha ×k=-1$$ (7)
According to the above calculation results, the slope of the negative correlation of Albedo-NDVI is k=-0.2303, then α=4.3422, and the expression of DDI is as follows:
$$DDI=4.3422×NDVI-Albedo$$ (8)
2.2.3 Desertification Categorization
In order to meet the needs of desertification evaluation and mapping, in 1984, the United Nations Food and Agriculture Organization (FAO) and the United Nations Environmental Programme (UNEP) in the “Desertification Assessment and Mapping Program” came up with the specific quantitative vegetation of desertification status, development rate and intrinsic risk assessment from vegetation degradation, wind erosion, water erosion and salinization as four aspects, and desertification was divided into mild, moderate, severe and extremely severe (level) according to the degree of development.
Through the DDI model constructed in the previous section, the dedicated data on the desertification index of the China-Pakistan Economic Corridor from 2000 to 2017 was obtained. When making the desertification categorization based on DDI, most researchers use the natural fracture method[7][11][12] in ArcGIS reclassification. The natural fracture method is a categorization point based on the Jank optimal method of statistics[13]. This classification minimizes the difference within the category and maximize the difference among the categorized. In this paper, the natural fracture method and the desertification degree quartering method are also used. The DDI value is divided into five intervals, and the five-level categorization index is determined (Table 2). The user can also combine with the field survey data to further fine-tune the DDI categorization table.
Table 2   DDI categorization table
 Type DDI Values Snowy and icy waters DDI≤−0.26 Severe desertification −0.26＜DDI≤0.12 Moderate desertification 0.12＜DDI≤0.55 Mild desertification 0.55＜DDI≤1.6 Non desertification 1.6＜DDI≤4.2
2.2.4 Simulation Process of Dedicated Data for Desertification Index
Firstly, the NDVI and Albedo data are reconstructed, the annual NDVI value is calculated by the maximum synthesis method, and the annual Albedo value is calculated by the minimum synthesis method. Then, the Albedo-NDVI feature space is constructed by fitting the Albedo and NDVI data to obtain the fitting slope k. Finally, the desertification index is constructed, the desertification grade is divided, and the desertification classification map based on the desertification index is completed. The data processing flow is shown in Figure 4.

Fig. 4   Simulation process for desertification index’s dedicated data
3.   Data Specimen Description
The China-Pakistan Economic Corridor annualized desertification data-set has 18 files from 2000–2017, each of which includes annual NDVI, Albedo, DDI grid data and 2010 desertification grading map examples. The spatial resolution of all data is 1 km. The annual NDVI, annual Albedo, and DDI grid data storage formats are tif, and the desertification classification map is saved in jpg format. Users can create desertification grading data based on DDI data. All data geographic coordinate systems are WGS1984. The data results are shown in Figure 5.

Fig. 5   China-Pakistan Economic Corridor desertification grading diagram 2010
4.   Data Quality Control and Assessment
This paper uses high-resolution data to evaluate the quality of desertification grading data. Taking the 2010 data as an example, the 8 scenes Landsat 7 image data is selected for verification on a small scale. The spatial resolution of the 8 scenes image is 30 m, the average cloud volume is less than 5%, and the data quality is good. DDI is calculated from the annual maximum of NDVI and the annual minimum of Albedo. Therefore, the Landsat data of August in the peak season of vegetation growth is used as the verification data to invert the annual maximum vegetation coverage. Data is preprocessed using ENVI software, including registration, geometry correction, image enhancement, and FLAASH atmospheric correction. Firstly the NDVI value of the 8 scene image is calculated, and then the vegetation coverage is estimated. The vegetation coverage calculation formula is as follows:
$$vfc=\frac{NDVI-NDV{I}_{soil}}{NDV{I}_{veg}-NDV{I}_{soil}}$$ （9）
Where: vfc is the vegetation coverage of the study zone, NDVIsoil is the NDVI value of the pure desert in the study zone, and NDVIveg is the NDVI value of the complete vegetation coverage area. Generally, the upper and lower thresholds of NDVI intercepted by 5% confidence are approximated to represent NDVIveg and NDVIsoil respectively.
This paper refers to other research works, and divides the degree of desertification in the China-Pakistan Economic Corridor by vegetation coverage into follows: non-desertification (high vegetation cover, vfc>60%), mild desertification (medium vegetation coverage, 30%-60%), moderate desertification (low to medium vegetation cover, 15% to 30%), severe desertification (low vegetation cover, <15%). During the verification, 30 desertification verification points of each level are selected on each image. Since the resolution of Landsat and Modis data is inconsistent, when selecting the verification point, selecting a small area with little difference in vegetation coverage on Landsat as a verification point, which is compared with the DDI grading results (Table 3) and Kappa coefficients are calculated. Table 4 is a data evaluation accuracy table after verification based on this method. The overall evaluation accuracy is 81.67%, and the Kappa coefficient is 75.42%. The data evaluation results of the two resolutions are basically the same, reflecting that it is feasible to use the data of this paper to evaluate the desertification of China-Pakistan Economic Corridor on a regional scale.
Table 3   DDI classification confusion matrix
DDI Classification Landsat ClassificationNon-desertificationMild desertificationModerate desertificationSevere desertificationTotal
Non desertification6155273
Mild desertification2502357
Moderate desertification5238449
Severe desertification3744761
Total71644956240
Table 4   Analysis of single type accuracy and overall accuracy
DDI classification Landsat classificationUser accuracyProducer accuracyOverall classification precisionKappa coefficient
Non desertification0.83560.85920.81670.7542
Mild desertification0.87720.7813
Moderate desertification0.77550.7755
Severe desertification0.77050.8393
5.   Data Value
Although the influencing factors of desertification are complex and the purpose of use is diversified, resulting in a wide range of indicators, all indicators can ultimately be attributed to the most basic impact factor - vegetation. Vegetation is not only a comprehensive demonstration of land cover quality, but also a main factor affecting soil and water conservation in the ecological environment. Based on the readily available remote sensing image data, the NDVI reflecting vegetation information and Albedo reflecting soil and water information are used to construct the DDI model of the China-Pakistan Economic Corridor, which directly reflects the degree of desertification in the China-Pakistan Economic Corridor. It is to provide reference for quantitative assessment of the severity of desertification, and provide basic information for desertification control and macro decision-making in countries along the China-Pakistan Economic Corridor.
6.   Data Use Method and Recommendations
The data on the distribution of desertification in the China-Pakistan Economic Corridor in 2000–2017 is saved in the grid tif format. Commonly used GIS and remote sensing software such as ArcGIS, QGIS, ENVI, ERDAS all support the reading and operation of this data.
Acknowledgments
Thanks to the MODIS data provided by LPDAAC, thanks to the Landsat 7 data provided by USGS. Thanks to the observation and data guidance and support given by the National Special Environment and Special Function Observation and Research Station’s shared service platform.
[1] Wang Xinyuan, Yang Xiaopeng, Chen Xiangshun, et al. Construction of Desertification Monitoring Indicator System——Taking Gansu Province as an Example [J]. ECOLOGICAL ECONOMY, 2016, 32(7): 174-177, 182.
[2] OH K, JEONG Y, LEE D, et al. Determining Development Density Using the Urban Carrying Capacity Assessment System[J]. Landscape & Urban Planning, 2005, 73(1): 1-15.
[3] SHENG GL,HARAZONO Y,OIKAWA T, et al. Grassland Desertification By Grazing and the Resulting Micrometeorological Changes in Inner Mongolia[J]. Agricultural & Forest Meteorology, 2000, 102(2): 125-137.
[4] Zeng Yongnian, Xiang Nanping, Feng Zhaodong, et al. Study on Albedo-NDVI Feature Space and Desertification Remote Sensing Monitoring Index[J]. GEOGRAPHICAL SCIENCE, 2006, 26(1): 75-81.
[5] Guan Yuwei. Construction and Trend Analysis of Global Desertification Index based on Remote Sensing Image [D]: Chengdu: University of Electronic Science and Technology of China, 2015.
[6] Li Yan, Zhou Youyou, Hu Baoqing, et al. Comparative Study on Evolution Characteristics of Desertification in Typical North and South Regions Based on 3S Technology[J]. Journal of Guangxi Teachers University (Natural Science Version), 2017, 34(1): 82-90 .
[7] MA Z,XIE Y,JIAO J, et al. The Construction and Application of an Aledo-ndvi Based Desertification Monitoring Model[J]. Procedia Environmental Sciences, 2011, 10: 2029-2035.
[8] LIANG S,SHUEY CJ,RUSS AL, et al. Narrowband to Broadband Conversions of Land Surface Albedo: Ii. Validation[J]. Remote Sensing of Environment, 2003, 84(1): 25-41.
[9] LIANG S. Narrowband to Broadband Conversions of Land Surface Albedo I : Algorithms[J]. Remote Sensing of Environment, 2001, 76(2): 213-238.
[10] VERSTRAETE MM,PINTY B. Designing Optimal Spectral Indexes for Remote Sensing Applications[J]. Ieee Transactions on Geoscience & Remote Sensing, 1996, 34(5): 1254-1265.
[11] Guan Yuwei. Construction and Trend Analysis of Global Desertification Index based on Remote Sensing Image [D]: Chengdu: University of Electronic Science and Technology of China, 2015.
[12] Wu Zhaopeng, Wang Mingxia, Zhao Xiao. Study of Desertification of Jinghe River Basin Based on Desertification Difference Index (DDI)[J]. Bulletin of Soil and Water Conservation, 2014, 34(4): 188-192.
[13] Jenks GF. The Data Model Concept in Statistical Mapping[J]. International Yearbook of Cartography, 1967, 7: 186-190.
Data citation
Min Y, Feng K, Kang J, Ai M. Data Set for Annualized Distribution of Desertification in China-Pakistan Economic Corridor from 2000 to 2017 [DB/OL]. National Special Environment, Special Function Observation and Research Station’s Sharing Platform, 2018. DOI: 10.12072/ Casnw.046.2018.db.

Min Y, Feng K, Kang J, Ai M. Data Set for Annualized Distribution of Desertification in China-Pakistan Economic Corridor from 2000 to 2017 [J/OL]. Chinese Scientific Data, 4(2019). DOI: 10.11922/csdata.2018.0056.en.
Min Yufang
Mainly responsible for the work: MODIS and Landsat data download, desertification data preparation process design and implementation.
myf@lzb.ac.cn
(1983-), female, resident of Lintan County, Gansu Province, Ph.D. candidate, engineer, whose research direction is the big data application in the cold and dry regions.
Feng Keting
Mainly responsible for the work: MODIS data pre-processing scheme design.
(1980—), male, resident of Zhongwei City, Ningxia Hui Autonomous Region, Ph.D. candidate, engineer, whose research direction is the application of big data in the cold and arid regions.
Kang Jianfang
Mainly responsible for the work: MODIS data pre-processing.
(1981-), female, resident of Qinan, Gansu, master’s degree, engineer, whose research direction is the big data application in the cold and dry regions.
Ai Minghao
Mainly responsible for the work: MODIS data batch processing.
(1986-), male, resident of Jining City, Shandong Province, master’s degree, engineer, whose research direction is the big data application in the cold and dry regions.
Data sharing fundamental program for Construction of the National Science and Technology Infrastructure Platform (Y719H71006); The 13th Five-Year Informatization Plan of Chinese Academy of Sciences (XXH13506, XXH13505).

II区出版时间：2019年8月2日 （ 版本EN1

csdata