2026-10-14T18:56:36Z https://www.scidb.cn/oai

10.57760/sciencedb.j00100.00064 2026-06-24T19:25:52Z

2026-06-24 Dataset for Rocky Desertification Degree Extraction from Multi-source Gaofen Imagery in Puding County, Guizhou Province doi:10.57760/sciencedb.j00100.00064 en This dataset is a multi-source Gaofen imagery dataset for rocky desertification degree extraction in Puding County, Guizhou Province, China . It supports the reproduction and verification of Figures 1–10 and Tables 1–3 in the paper “Extracting Rocky Desertification Degree Using Gaofen Satellite Data” (Manuscript No. dlxb2024-0549). The dataset includes study area overview data, multi-source Gaofen satellite imagery and field samples, sample-plot land-cover classification data, field-measured endmember spectra, sample dataset construction data, bare rock fraction inversion and validation data, local comparison data of inversion results, local comparison data of classification results, FCN training performance data, model accuracy comparison data, and tabular data for GF imagery band parameters, rocky desertification classification criteria, and model accuracy assessment.The study area is mainly located in Puding County, Anshun City, Guizhou Province, China. The GF-2, GF-6, and GF-7 images were acquired on January 22, 2020, November 11, 2020, and April 3, 2022, respectively. The field photographs of the study area were taken on September 5, 2024, and the field validation samples were collected on February 28, 2022. The data types include GeoTIFF raster files, Shapefile vector data, Excel/CSV tables, PNG thumbnails, Word data documents, and ZIP archives. Most spatial data use the WGS_1984_UTM_Zone_48N coordinate system, while part of the study area overview data uses the GCS_WGS_1984 coordinate system. The spatial resolutions include 1 m for GF-2, 16 m for GF-6, 0.65 m for GF-7, and 2.6 m for some inversion or resampled products.The main processing steps include radiometric calibration, atmospheric correction, geometric correction, study area clipping, sample annotation, UAV sample-plot data organization, field endmember spectra compilation, model-based classification, linear spectral mixture model inversion, road-mask post-processing, accuracy assessment, and figure/table data organization. The dataset can be used for rocky desertification degree extraction, bare rock fraction inversion, comparison of classification accuracies from multi-source Gaofen imagery, sample dataset construction, and reproduction of the associated paper’s figures and tables. Some GF imagery and model result files are large and are recommended to be used with ArcGIS, QGIS, ENVI, Excel, Python, or R. rocky desertification; Gaofen satellite imagery; Puding County; remote sensing inversion; bare rock fraction; FCN; LSMM; classification accuracy liu gang Ji Cuicui RESTRICTED dataset Science Data Bank