A soil organic carbon mapping method based on transfer learning without the use of exogenous data

Accurate and cost-effective mapping of soil organic carbon (SOC) is critical for understanding carbon dynamics and informing sustainable land management. Although deep learning-based methods have demonstrated strong potential in digital soil mapping, they typically require large amounts of data. How...

Full description

Saved in:
Bibliographic Details
Main Authors: Jingfeng Han, Mujie Wu, Yanlong Qi, Xiaoning Li, Xiao Chen, Jing Wang, Jinlong Zhu, Qingliang Li
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-05-01
Series:Frontiers in Environmental Science
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fenvs.2025.1580085/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850136729459097600
author Jingfeng Han
Mujie Wu
Yanlong Qi
Xiaoning Li
Xiao Chen
Jing Wang
Jinlong Zhu
Qingliang Li
author_facet Jingfeng Han
Mujie Wu
Yanlong Qi
Xiaoning Li
Xiao Chen
Jing Wang
Jinlong Zhu
Qingliang Li
author_sort Jingfeng Han
collection DOAJ
description Accurate and cost-effective mapping of soil organic carbon (SOC) is critical for understanding carbon dynamics and informing sustainable land management. Although deep learning-based methods have demonstrated strong potential in digital soil mapping, they typically require large amounts of data. However, the availability of site-level SOC observations is often limited, which poses a challenge for model performance. To address this, we propose a novel transfer learning approach based on a Convolutional Neural Network (CNN) model that does not rely on exogenous data. Specifically, when predicting SOC for a given soil layer, the model is first pre-trained on data from all layers and then fine-tuned using data from the target layer. This design enables more efficient use of limited site data. Experimental results show that the proposed transfer model consistently outperforms other machine learning models, including the Random Forest (RF), standard CNN, and Multi-Task CNN (MTCNN) models. The transfer model achieves a coefficient of determination (R2) of 0.374 and a root mean square error (RMSE) of 2.937%, indicating superior performance. These findings highlight the effectiveness of the proposed approach for digital soil mapping under data-scarce conditions and underscore its potential as a robust tool for accurate SOC estimation.
format Article
id doaj-art-56181cda2b464bde9bbe2ffab073afd7
institution OA Journals
issn 2296-665X
language English
publishDate 2025-05-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Environmental Science
spelling doaj-art-56181cda2b464bde9bbe2ffab073afd72025-08-20T02:31:03ZengFrontiers Media S.A.Frontiers in Environmental Science2296-665X2025-05-011310.3389/fenvs.2025.15800851580085A soil organic carbon mapping method based on transfer learning without the use of exogenous dataJingfeng HanMujie WuYanlong QiXiaoning LiXiao ChenJing WangJinlong ZhuQingliang LiAccurate and cost-effective mapping of soil organic carbon (SOC) is critical for understanding carbon dynamics and informing sustainable land management. Although deep learning-based methods have demonstrated strong potential in digital soil mapping, they typically require large amounts of data. However, the availability of site-level SOC observations is often limited, which poses a challenge for model performance. To address this, we propose a novel transfer learning approach based on a Convolutional Neural Network (CNN) model that does not rely on exogenous data. Specifically, when predicting SOC for a given soil layer, the model is first pre-trained on data from all layers and then fine-tuned using data from the target layer. This design enables more efficient use of limited site data. Experimental results show that the proposed transfer model consistently outperforms other machine learning models, including the Random Forest (RF), standard CNN, and Multi-Task CNN (MTCNN) models. The transfer model achieves a coefficient of determination (R2) of 0.374 and a root mean square error (RMSE) of 2.937%, indicating superior performance. These findings highlight the effectiveness of the proposed approach for digital soil mapping under data-scarce conditions and underscore its potential as a robust tool for accurate SOC estimation.https://www.frontiersin.org/articles/10.3389/fenvs.2025.1580085/fulltransfer learningsoil organic carbondigital soil mappingdeep learningsoil depth correlation
spellingShingle Jingfeng Han
Mujie Wu
Yanlong Qi
Xiaoning Li
Xiao Chen
Jing Wang
Jinlong Zhu
Qingliang Li
A soil organic carbon mapping method based on transfer learning without the use of exogenous data
Frontiers in Environmental Science
transfer learning
soil organic carbon
digital soil mapping
deep learning
soil depth correlation
title A soil organic carbon mapping method based on transfer learning without the use of exogenous data
title_full A soil organic carbon mapping method based on transfer learning without the use of exogenous data
title_fullStr A soil organic carbon mapping method based on transfer learning without the use of exogenous data
title_full_unstemmed A soil organic carbon mapping method based on transfer learning without the use of exogenous data
title_short A soil organic carbon mapping method based on transfer learning without the use of exogenous data
title_sort soil organic carbon mapping method based on transfer learning without the use of exogenous data
topic transfer learning
soil organic carbon
digital soil mapping
deep learning
soil depth correlation
url https://www.frontiersin.org/articles/10.3389/fenvs.2025.1580085/full
work_keys_str_mv AT jingfenghan asoilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT mujiewu asoilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT yanlongqi asoilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT xiaoningli asoilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT xiaochen asoilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT jingwang asoilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT jinlongzhu asoilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT qingliangli asoilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT jingfenghan soilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT mujiewu soilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT yanlongqi soilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT xiaoningli soilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT xiaochen soilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT jingwang soilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT jinlongzhu soilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata
AT qingliangli soilorganiccarbonmappingmethodbasedontransferlearningwithouttheuseofexogenousdata