The data comes from Landsate TM and ETM image data from the entire Xinjiang area around 2005 downloaded from the website of the University of Maryland, and the 1: 100,000 topographic map published in China in the 1970s collected by the Documentation Center of the Xinjiang Branch of the Chinese Academy of Sciences.

The TM data is first filtered to select those with cloud coverage of less than 4%, or less coverage of the main features. Combined with a 1: 100,000 topographic map, the selected images are geometrically corrected, then spliced, and evenly colored, and finally become a vectorized product. Professionals visually interpret and manually vectorize the TM data of each scene according to the interpretation mark and classification system to generate COVERAGE format data. Then Arcgis software is used to convert COVERAGE into SHP data, and the data is generated through projection conversion.

This data is processed by the geographical information and mapping professionals of the Xinjiang Institute of Ecology and Geography, China Academy of Sciences. From scanning of paper materials, screening of image data, precise geometric correction, splicing, color uniformity to digital vectorization and manual interpretation, the entire operation process complies with relevant national standards. The digital error is controlled within 2 pixels, ensuring that the vectorized boundary is completely consistent with the boundary of the topographic map. The vector data truly reflects the shape and attribute information of the feature. Data accuracy is as high as 99%.