2026-10-12T05:15:19Z https://www.scidb.cn/oai

10.57760/sciencedb.IGA.001er 2026-09-15T13:23:47Z

2026-09-15 Black Soil Texture Distribution Map (2021) doi:10.57760/sciencedb.iga.001er en This dataset presents the results of systematic sampling and high-resolution spatial prediction of topsoil texture (particle size fractions) for cultivated land in Northeast China. It includes 1,413 topsoil (0–20 cm) samples collected in April and October 2021 across the region, covering diverse soil types such as black soil, chernozem, meadow soil, and others, ensuring broad representativeness. Each sample was analyzed using a Malvern MS‑2000 laser particle size analyzer to determine the mass fractions of sand (0.05–2 mm), silt (0.002–0.05 mm), and clay (<0.002 mm), with the three fractions summing to 100%. Based on the sample data and remote sensing variables, a Random Forest (RF) model was employed to spatially predict the particle size fractions, generating raster maps of sand, silt, and clay contents at 10 m resolution covering the entire study area, with a total data volume of approximately 48.17 GB. For the sand fraction, the validation R² was 0.77 and RMSE was 10.94%; for the silt fraction, R² was 0.75 and RMSE was 10.05%; and for the clay fraction, R² was 0.65 and RMSE was 2.27%. Soil Texture; Northeast China; Machine Learning 郑淼 PUBLIC https://creativecommons.org/licenses/by/4.0/ dataset Science Data Bank