Soil Organic Carbon Prediction and Mapping in Morocco Using PRISMA Hyperspectral Imagery and Meta-Learner Model

Accurate mapping of soil organic carbon (SOC) supports sustainable land management practices and carbon accounting initiatives for mitigating climate change impacts. This study presents a novel meta-learner framework that combines multiple machine learning algorithms and spectra processing algorithm...

Full description

Saved in:
Bibliographic Details
Main Authors: Yassine Bouslihim, Abdelkrim Bouasria, Budiman Minasny, Fabio Castaldi, Andree Mentho Nenkam, Ali El Battay, Abdelghani Chehbouni
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Remote Sensing
Subjects:
Online Access:https://www.mdpi.com/2072-4292/17/8/1363
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850180664145477632
author Yassine Bouslihim
Abdelkrim Bouasria
Budiman Minasny
Fabio Castaldi
Andree Mentho Nenkam
Ali El Battay
Abdelghani Chehbouni
author_facet Yassine Bouslihim
Abdelkrim Bouasria
Budiman Minasny
Fabio Castaldi
Andree Mentho Nenkam
Ali El Battay
Abdelghani Chehbouni
author_sort Yassine Bouslihim
collection DOAJ
description Accurate mapping of soil organic carbon (SOC) supports sustainable land management practices and carbon accounting initiatives for mitigating climate change impacts. This study presents a novel meta-learner framework that combines multiple machine learning algorithms and spectra processing algorithms to optimize SOC prediction using the PRISMA hyperspectral satellite imagery in the Doukkala plain of Morocco. The framework employs a two-layer structure of prediction models. The first layer consists of Random Forest (RF), Support Vector Regression (SVR), and Partial Least Squares Regression (PLSR). These base models were configured using data smoothing, transformation, and spectral feature selection techniques, based on a 70/30% data split. The second layer utilizes a ridge regression model as a meta-learner to integrate predictions from the base models. Results indicated that RF and SVR performance improved primarily with feature selection, while PLSR was most influenced by data smoothing. The meta-learner approach outperformed individual base models, achieving an average relative improvement of 48.8% over single models, with an R<sup>2</sup> of 0.65, an RMSE of 0.194%, and an RPIQ of 2.247. This study contributes to the development of methodologies for predicting and mapping soil properties using PRISMA hyperspectral data.
format Article
id doaj-art-bcb1b6c3952b450587a8a79840145ebf
institution OA Journals
issn 2072-4292
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Remote Sensing
spelling doaj-art-bcb1b6c3952b450587a8a79840145ebf2025-08-20T02:18:04ZengMDPI AGRemote Sensing2072-42922025-04-01178136310.3390/rs17081363Soil Organic Carbon Prediction and Mapping in Morocco Using PRISMA Hyperspectral Imagery and Meta-Learner ModelYassine Bouslihim0Abdelkrim Bouasria1Budiman Minasny2Fabio Castaldi3Andree Mentho Nenkam4Ali El Battay5Abdelghani Chehbouni6National Institute for Agricultural Research, Rabat 10000, MoroccoFaculty of Science, Chouaib Doukkali University, El Jadida 24000, MoroccoSchool of Life & Environmental Sciences, Sydney Institute of Agriculture, The University of Sydney, Sydney, NSW 2006, AustraliaInstitute of BioEconomy, National Research Council of Italy (CNR), Via Giovanni Caproni 8, 50145 Firenze, ItalySchool of Life & Environmental Sciences, Sydney Institute of Agriculture, The University of Sydney, Sydney, NSW 2006, AustraliaCenter for Remote Sensing Applications (CRSA), Mohammed VI Polytechnic University (UM6P), Lot 660, Hay Moulay Rachid, Ben Guerir 43150, MoroccoCenter for Remote Sensing Applications (CRSA), Mohammed VI Polytechnic University (UM6P), Lot 660, Hay Moulay Rachid, Ben Guerir 43150, MoroccoAccurate mapping of soil organic carbon (SOC) supports sustainable land management practices and carbon accounting initiatives for mitigating climate change impacts. This study presents a novel meta-learner framework that combines multiple machine learning algorithms and spectra processing algorithms to optimize SOC prediction using the PRISMA hyperspectral satellite imagery in the Doukkala plain of Morocco. The framework employs a two-layer structure of prediction models. The first layer consists of Random Forest (RF), Support Vector Regression (SVR), and Partial Least Squares Regression (PLSR). These base models were configured using data smoothing, transformation, and spectral feature selection techniques, based on a 70/30% data split. The second layer utilizes a ridge regression model as a meta-learner to integrate predictions from the base models. Results indicated that RF and SVR performance improved primarily with feature selection, while PLSR was most influenced by data smoothing. The meta-learner approach outperformed individual base models, achieving an average relative improvement of 48.8% over single models, with an R<sup>2</sup> of 0.65, an RMSE of 0.194%, and an RPIQ of 2.247. This study contributes to the development of methodologies for predicting and mapping soil properties using PRISMA hyperspectral data.https://www.mdpi.com/2072-4292/17/8/1363hyperspectral satellite imageryPRISMAmeta-learner modeldigital soil mappingsoil organic carbon
spellingShingle Yassine Bouslihim
Abdelkrim Bouasria
Budiman Minasny
Fabio Castaldi
Andree Mentho Nenkam
Ali El Battay
Abdelghani Chehbouni
Soil Organic Carbon Prediction and Mapping in Morocco Using PRISMA Hyperspectral Imagery and Meta-Learner Model
Remote Sensing
hyperspectral satellite imagery
PRISMA
meta-learner model
digital soil mapping
soil organic carbon
title Soil Organic Carbon Prediction and Mapping in Morocco Using PRISMA Hyperspectral Imagery and Meta-Learner Model
title_full Soil Organic Carbon Prediction and Mapping in Morocco Using PRISMA Hyperspectral Imagery and Meta-Learner Model
title_fullStr Soil Organic Carbon Prediction and Mapping in Morocco Using PRISMA Hyperspectral Imagery and Meta-Learner Model
title_full_unstemmed Soil Organic Carbon Prediction and Mapping in Morocco Using PRISMA Hyperspectral Imagery and Meta-Learner Model
title_short Soil Organic Carbon Prediction and Mapping in Morocco Using PRISMA Hyperspectral Imagery and Meta-Learner Model
title_sort soil organic carbon prediction and mapping in morocco using prisma hyperspectral imagery and meta learner model
topic hyperspectral satellite imagery
PRISMA
meta-learner model
digital soil mapping
soil organic carbon
url https://www.mdpi.com/2072-4292/17/8/1363
work_keys_str_mv AT yassinebouslihim soilorganiccarbonpredictionandmappinginmoroccousingprismahyperspectralimageryandmetalearnermodel
AT abdelkrimbouasria soilorganiccarbonpredictionandmappinginmoroccousingprismahyperspectralimageryandmetalearnermodel
AT budimanminasny soilorganiccarbonpredictionandmappinginmoroccousingprismahyperspectralimageryandmetalearnermodel
AT fabiocastaldi soilorganiccarbonpredictionandmappinginmoroccousingprismahyperspectralimageryandmetalearnermodel
AT andreementhonenkam soilorganiccarbonpredictionandmappinginmoroccousingprismahyperspectralimageryandmetalearnermodel
AT alielbattay soilorganiccarbonpredictionandmappinginmoroccousingprismahyperspectralimageryandmetalearnermodel
AT abdelghanichehbouni soilorganiccarbonpredictionandmappinginmoroccousingprismahyperspectralimageryandmetalearnermodel