This dataset reflects the distribution of soil silt content in the Songnen Plain of China, and can provide data support for soil physical properties and various agricultural research. The dataset is based on field sampling and laboratory analysis in 2020, covering the main agricultural areas of the Songnen Plain with a spatial resolution of 10 meters, providing high-precision information on soil silt content. The establishment of the dataset adopts advanced remote sensing technology and geographic information system (GIS) methods, combined with ground sampling data, to ensure the accuracy and reliability of the data.
This dataset is sourced from the Science Data Bank website（ https://www.scidb.cn/en/detail?dataSetId=DS_8c64142471e54b5a95b410976976793b&version=V1 ）.
This dataset analyzed the relationship between the band and spectral index of Sentinel-2 and soil particle content in the Songnen Plain, and combined it with the Random Forest algorithm (RF) to produce soil particle content at a resolution of 10m in the Songnen Plain.
The original data is Sentinel-2 product, which is widely used and has reliable accuracy. For powder particles, the coefficient of determination (R2) and root mean square error (RMSE) of Sentinel-2 imaging and RF algorithm are 0.75 and 9.38%, respectively.
This work is licensed under a
CC BY 4.0.
CSTR:11738.11.NCDC.SCIDB.DB6761.2025
10.57760/sciencedb.IGA.01002
