Dataset for 30 m Soil Organic Matter Mapping Using Long-Term Soil Erosion Dynamics in Northeast China

Published: 14 October 2026| Version 1 | DOI: 10.17632/rrs4prjm9r.1
Contributor:
Chong Luo

Description

This dataset contains the processed 30 m soil organic matter (SOM) inversion results for the black soil region of Northeast China. The results were generated using the ResCNN-ECA-MLP deep-learning framework and represent the spatially predicted SOM content derived from multi-source remote-sensing and environmental variables. The dataset is provided to support the manuscript entitled “A Hybrid ResCNN-ECA-MLP Framework for 30 m Soil Organic Matter Mapping Using Long-Term Soil Erosion Dynamics in Northeast China”. Only the processed SOM inversion results are shared in this dataset. The original soil sampling data, remote-sensing data, environmental covariates, and model training data are not included. SOM values are reported in g/kg. Detailed information on the file format, spatial resolution, coordinate reference system, spatial extent, and NoData values is provided in the accompanying README file.

Files

Steps to reproduce

The dataset contains processed 30 m soil organic matter inversion results for the black soil region of Northeast China. The raster results were generated using the ResCNN-ECA-MLP deep-learning framework based on multi-source remote-sensing and environmental variables, including long-term soil erosion information. The uploaded ZIP file contains the processed raster data. The raster values represent soil organic matter content in g/kg. The spatial resolution is 30 m × 30 m, and the coordinate reference system is WGS 1984 UTM Zone 51N (EPSG:32651). The NoData value is -3.40282346639e+38. The original soil sampling data, remote-sensing data, environmental covariates, model training data, and source code are not included in this dataset.

Categories

Soil Science, Remote Sensing, Machine Learning

Licence