Soil Carbon Fractions with 1 km Resolution Predictions for the Great Plains

From May 2007 to October 2010, this dataset provides estimates for total organic soil carbon and its pyrogenic, particulate, and other fractions. It includes 473 surface soil samples from Colorado, Kansas, New Mexico, and Wyoming, alongside terrain, climate, soil, fire, and land cover data used to generate 1 km resolution predictive maps via a random forest model. The dataset is produced by the National Aeronautics and Space Administration.

Use Cases

• Modeling spatial distribution of soil organic carbon fractions using predicted SOC, PyC, POC, and OOC values.
• Analyzing relationships between soil carbon stocks and environmental drivers like terrain, climate, and fire history.
• Validating regional-scale carbon cycle models with ground-sampled soil carbon data from stratified locations.
• Assessing land cover impacts on carbon sequestration potential in the Great Plains region.

Strengths

• Contains 473 physically collected surface soil samples from a stratified sampling design.
• Provides 1 km resolution predictive maps for four distinct soil carbon fractions across the study region.
• Integrates multiple predictor variable types including terrain, climate, soil, fire, and land cover data.

Limitations

• Specific column names and the total row count for the predictive maps are not provided by any source.
• Sources conflict on the last updated date: Data.gov lists March 2026, while NASA Earthdata lists October 2010.

Provenance

Source

National Aeronautics and Space Administration

Collection Method

Estimates derived from field-collected soil samples and a best random forest regression model applied to ancillary geospatial data.

Time Range

2007-05-01 to 2010-10-01

Freshness

2026-03-12 21:22:42.737403

Geography

Colorado, Kansas, New Mexico, and Wyoming, USA (Great Plains region)