Pixel-wise spectral correction framework for soil total carbon mapping: integrating hyperspectral image and laboratory Vis–NIR spectra

Soil total carbon (STC) is pivotal to evaluate soil fertility and enable sustainable agricultural management. Laboratory visible and near-infrared (Vis-NIR) spectroscopy holds potentials to quantify STC rapidly and efficiently at the field or farmland scale, but the approach cannot be transferred to large-scale mapping. Satellite hyperspectral image offers a promising solution, yet faces challenges such as atmospheric interference, soil moisture and vegetation residuals. Here we proposed a novel STC mapping framework that integrates the high-fidelity laboratory spectra with the satellite hyperspectral image for pixel-level spectral correction across the 400~2500 nm range. Specifically, laboratory-measured soil Vis–NIR spectra were resampled using the Gaofen-5 (GF5) spectral response function to simulate sensor-level reflectance. A band-wise mapping between the simulated spectra and the original GF5 reflectance was established, thereby enabling pixel-wise spectral correction and generating corrected hyperspectral images with enhanced spatial detail and spectral fidelity. Results demonstrated that the proposed spectral correction framework not only enhanced the correlation between STC and GF5 images but also improved the precision of STC mapping. For corrected images, the validation R2, RMSE, and CCC were 0.59, 2.57 g/kg, and 0.75, respectively; compared with the uncorrected images, the validation R2 increased by 180.9%, the RMSE decreased by 0.97 g/kg, and the CCC increased by 127.3%. Furthermore, the corrected image yielded smoother STC maps with reduced uncertainty and diminished spatial residual autocorrelation. This integrative approach bridges laboratory and satellite hyperspectral through pixel-wise spectral correction, and offers a valuable tool for agricultural and environmental assessments.

Authors

Institutions

Publication Details

Journal
Journal of Remote Sensing
Published
2026-11-17
DOI
https://doi.org/10.34133/remotesensing.1093
Primary Topic
Soil Geostatistics and Mapping
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Pixel-wise spectral correction framework for soil total carbon mapping: integrating hyperspectral image and laboratory Vis–NIR spectra

Yiyun Chen, Yongshen Hong, Jiaxue Wang, Bo Wang et al.
2 links
Journal of Remote Sensing
Soil Geostatistics and Mapping
article

Pixel-wise spectral correction framework for soil total carbon mapping: integrating hyperspectral image and laboratory Vis–NIR spectra

Yiyun Chen, Yongshen Hong, Jiaxue Wang, Bo Wang, Long Guo, Yujiao Wei, Shuangyin Zhang, Abdul Mounem Mouazen, Huanfeng Shen, Jianping Qian, Chen Wu
article en

Abstract

Soil total carbon (STC) is pivotal to evaluate soil fertility and enable sustainable agricultural management. Laboratory visible and near-infrared (Vis-NIR) spectroscopy holds potentials to quantify STC rapidly and efficiently at the field or farmland scale, but the approach cannot be transferred to large-scale mapping. Satellite hyperspectral image offers a promising solution, yet faces challenges such as atmospheric interference, soil moisture and vegetation residuals. Here we proposed a novel STC mapping framework that integrates the high-fidelity laboratory spectra with the satellite hyperspectral image for pixel-level spectral correction across the 400~2500 nm range. Specifically, laboratory-measured soil Vis–NIR spectra were resampled using the Gaofen-5 (GF5) spectral response function to simulate sensor-level reflectance. A band-wise mapping between the simulated spectra and the original GF5 reflectance was established, thereby enabling pixel-wise spectral correction and generating corrected hyperspectral images with enhanced spatial detail and spectral fidelity. Results demonstrated that the proposed spectral correction framework not only enhanced the correlation between STC and GF5 images but also improved the precision of STC mapping. For corrected images, the validation R2, RMSE, and CCC were 0.59, 2.57 g/kg, and 0.75, respectively; compared with the uncorrected images, the validation R2 increased by 180.9%, the RMSE decreased by 0.97 g/kg, and the CCC increased by 127.3%. Furthermore, the corrected image yielded smoother STC maps with reduced uncertainty and diminished spatial residual autocorrelation. This integrative approach bridges laboratory and satellite hyperspectral through pixel-wise spectral correction, and offers a valuable tool for agricultural and environmental assessments.

Journal of Remote Sensing
Vytautas Magnus University (LT), Huazhong Agricultural University (CN), Changjiang Water Resources Commission (CN), Ghent University (BE), Wuhan University (CN), Institute of Agricultural Resources and Regional Planning (CN), State Key Laboratory of Information Engineering in Surveying Mapping and Remote Sensing (CN), Chinese Academy of Agricultural Sciences (CN), Changjiang River Scientific Research Institute (CN)
Life on land, Climate action
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