Soil Salinity Estimates with Vegetation Cover from GF-1 Satellite, Inner Mongolia 2018

Junrui Zhang from Northwest A&F University created this dataset to systematically explore the effect of Fractional Vegetation Cover (FVC) on soil salinity monitoring accuracy. It contains field-sampled Soil Salt Content (SSC) data from May to August 2018 at different depths, paired with spectral covariates and FVC calculated from GF-1 satellite images in the Jiefangzha Irrigation District. The study constructed SSC inversion models using Partial Least Square Regression, Cubist, and Extreme Learning Machine.

Use Cases

• Training soil salinity inversion models based on spectral covariates from satellite imagery.
• Analyzing the relationship between soil salt content and fractional vegetation cover across different land classifications.
• Comparing the performance of different machine learning models (PLSR, Cubist, ELM) for predicting soil salinity.
• Studying seasonal variations in soil salinity based on monthly sampling from May to August.

Strengths

• Data collection includes field samples from three soil depth ranges (0-20cm, 0-40cm, 0-60cm).
• Analysis is based on satellite imagery from the GF-1 satellite and includes multiple modeling approaches.
• The study systematically tests the effect of FVC by dividing data into five experimental treatments (TA to TE).

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.

Provenance

Source

Northwest A&F University

Collection Method

Field sampling paired with satellite remote sensing (GF-1) data processing.

Time Range

May to August 2018

Geography

Jiefangzha Irrigation District, Inner Mongolia