March 8, 2024

Contribution of Sentinel-2 Seedbed Spectra to the Digital Mapping of Soil Organic Carbon

View Full Paper

Ask AI

Bookmark

Authors

FVFien VanongevalKU Leuven JOJos Van OrshovenResearch Institute for Nature and Forest AGAnne GobinFlemish Institute for Technological Research

Discussion

Member takes

Implication

Comparative modeling analysis reveals poor satellite spectral prediction of soil organic carbon in agricultural parcels, highlighting the dominant influence of topography and vegetation.

Key Points

Satellite seedbed reflectance yields poor accuracy for digital soil mapping of soil organic carbon, whereas spring spectra adequately estimate soil texture across agricultural fields.
Predictive performance for soil organic carbon peaked at R² of 0.26 with gradient boosting, while bare soil spectra alone reached an R² of only 0.19 across 175 field parcels.
Assessment using machine learning algorithms compared five regression frameworks linking multispectral bands, environmental covariates, and soil texture across agricultural parcels.
Terrain attributes and prior vegetation cover dominate regional soil organic carbon distribution, calling for higher-resolution modeling to capture fine-scale landscape heterogeneity.

Cite This Study

Vanongeval et al. (2024) studied this question.

synapsesocial.com/papers/68e752c8b6db6435876cb0a8 https://doi.org/10.5194/egusphere-egu24-11583

View Full Paper

Ask AI

Bookmark

Implication

Comparative modeling analysis reveals poor satellite spectral prediction of soil organic carbon in agricultural parcels, highlighting the dominant influence of topography and vegetation.

Contribution of Sentinel-2 Seedbed Spectra to the Digital Mapping of Soil Organic Carbon

Discussion

Member takes

Implication

Key Points

Cite This Study

Contribution of Sentinel-2 Seedbed Spectra to the Digital Mapping of Soil Organic Carbon

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider