Texture-dependent bias in a European vis–NIR map of particulate and mineral-associated soil carbon

Abstract Particulate (POM) and mineral-associated organic matter (MAOM) carbon (C) predicted from visible–near-infrared (vis–NIR) spectra now inform a C saturation risk index for European agricultural soils. Their uncertainty was treated as random. I tested whether the error depends on soil texture at 6,399 LUCAS 2009 points not used for calibration, by comparing the summed predicted POM C and MAOM C with measured soil organic carbon (SOC). Overall, the sum tracked SOC ( R 2 = 0.58; slope 0.92). At the same predicted value, however, measured SOC fell by 0.155 g kg − 1 for every percent of sand (95% interval 0.138 to 0.174). Sandy soils were over-predicted by up to 14.0 g kg − 1 and clayey soils under-predicted by up to 8.3 g kg − 1 . In sandy soils the sum explained none of the variation in SOC ( R 2 = − 0.31). A sand correction, validated by leaving out one 2° grid cell at a time, raised R 2 to 0.55 in sandy soils. Fractionation recovery did not explain the bias. The error is systematic and largest in coarse soils, where MAOM C capacity is smallest. It should be tested and corrected before saturation is computed from such maps.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-16
DOI
https://doi.org/10.21203/rs.3.rs-11562265/v1
Primary Topic
Soil Geostatistics and Mapping
Type
preprint

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Texture-dependent bias in a European vis–NIR map of particulate and mineral-associated soil carbon

Md. Kamruzzaman
2 links
Research Square
Soil Geostatistics and Mapping
preprint

Texture-dependent bias in a European vis–NIR map of particulate and mineral-associated soil carbon

Md. Kamruzzaman
preprint en

Abstract

Abstract Particulate (POM) and mineral-associated organic matter (MAOM) carbon (C) predicted from visible–near-infrared (vis–NIR) spectra now inform a C saturation risk index for European agricultural soils. Their uncertainty was treated as random. I tested whether the error depends on soil texture at 6,399 LUCAS 2009 points not used for calibration, by comparing the summed predicted POM C and MAOM C with measured soil organic carbon (SOC). Overall, the sum tracked SOC ( R 2 = 0.58; slope 0.92). At the same predicted value, however, measured SOC fell by 0.155 g kg − 1 for every percent of sand (95% interval 0.138 to 0.174). Sandy soils were over-predicted by up to 14.0 g kg − 1 and clayey soils under-predicted by up to 8.3 g kg − 1 . In sandy soils the sum explained none of the variation in SOC ( R 2 = − 0.31). A sand correction, validated by leaving out one 2° grid cell at a time, raised R 2 to 0.55 in sandy soils. Fractionation recovery did not explain the bias. The error is systematic and largest in coarse soils, where MAOM C capacity is smallest. It should be tested and corrected before saturation is computed from such maps.

Research Square
Magyar Agrár- és Élettudományi Egyetem (HU)
European Commission, Joint Research Centre
Life on land
Soil Geostatistics and Mapping
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon