Grassland resources are the material basis of animal husbandry production and the gene pool of biodiversity, which have the ecological functions of preventing wind, fixing sand and conserving water. Grassland classification based on satellite remote sensing can not only provide data support for the investigation of grassland resources, but also provide a basis for the study of grassland species diversity. Based on the field survey data, the grassland classification information of real sample plots was obtained. Based on field survey data, the grassland type information of real sample plots is obtained. Landsat8 remote sensing images that have been preprocessed by radiation calibration, atmospheric correction, image fusion, seamless mosaic, etc. are used. After the operation process of dataset sample labeling, sample data clipping, result inspection, data enhancement, etc., combined with visual interpretation of remote sensing interpretation, ArcGIS software is used to construct 8 types of grassland in Tarbagatay Prefecture, a total of 3360 multispectral remote sensing image classification training datasets, and the number of sample images in each category is between 300-580.This data set can not only provide data basis for remote sensing image classification of grassland in Tarbagatay Prefecture, but also provide data reference for researchers in related fields in other areas.

Data summary: