Backed by the latest research
- Perennial builds on decades of research in land remote sensing and statistical methods for the quantification of carbon density
- We use an in-house archive of in-situ soil samples across cropland, pastureland, and rangeland joined with remote sensing and environmental variables to train machine learning algorithms for the prediction of soil organic carbon content over time
- Every algorithm we develop undergoes rigorous cross-validation to ensure we meet performance thresholds, even when no soil samples are available