Predictive Modeling of soil salinity integrating remote sensing and soil variables: An ensembled deep learning approach

Accurate predictions of soil salinity can significantly contribute to achieving the UN- Sustainable Development Goal (SDG-2) of ensuring ‘zero hunger.’ From this perspective, the current research aimed to predict soil electrical conductivity (EC) from remote sensing and soil data using advanced deep...

Full description

Saved in:
Bibliographic Details
Main Authors: Sana Arshad, Jamil Hasan Kazmi, Endre Harsányi, Farheen Nazli, Waseem Hassan, Saima Shaikh, Main Al-Dalahmeh, Safwan Mohammed
Format: Article
Language:English
Published: Elsevier 2025-03-01
Series:Energy Nexus
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2772427125000154
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850278390337110016
author Sana Arshad
Jamil Hasan Kazmi
Endre Harsányi
Farheen Nazli
Waseem Hassan
Saima Shaikh
Main Al-Dalahmeh
Safwan Mohammed
author_facet Sana Arshad
Jamil Hasan Kazmi
Endre Harsányi
Farheen Nazli
Waseem Hassan
Saima Shaikh
Main Al-Dalahmeh
Safwan Mohammed
author_sort Sana Arshad
collection DOAJ
description Accurate predictions of soil salinity can significantly contribute to achieving the UN- Sustainable Development Goal (SDG-2) of ensuring ‘zero hunger.’ From this perspective, the current research aimed to predict soil electrical conductivity (EC) from remote sensing and soil data using advanced deep learning (DL) architectures. A total of 109 soil samples were analyzed for agricultural land use in the Middle Indus Basin of Pakistan. Seven salinity indices (SI-1 to SI-7) were derived from the 10m to 20m wavelength bands of Sentinel-2, along with vegetation and topographic covariates. Initially, Recursive Feature Elimination was implemented as a feature-selection method to select the most effective predictors. Subsequently, deep learning architectures, including a Feedforward Neural Network (FFNN), Recurrent Neural Network (RNN), and Long Short-Term Memory (LSTM), were employed to predict soil salinity. Research findings showed that EC ranged between 0.57dS/m to 11.5 dS/m in the study area. The evaluation metrics of the DL models revealed that a simple FFNN with three fully connected dense layers achieved the highest R2 = 0.88 for model training. However, the ensemble of improved FFNN and LSTM outperformed with the highest R2 and NSE = 0.84, and the lowest RMSE and MAE = 1.38 and 1.01, respectively, on the testing dataset. Optimized deep learning architectures with adjustments to the learning rate, dropout rate, and activation functions achieved the highest prediction accuracy with the lowest validation loss. Finally, SHapely Additive exPlanations (SHAP) revealed that elevation, pH, NDVI, SI-1, and SI-7 had highly significant impacts on EC predictions. This research provides insight into implementing advanced and interpretable DL architectures, supporting informed decision-making by agricultural stakeholders.
format Article
id doaj-art-91dcf594aef949488f6237c53aa2737d
institution OA Journals
issn 2772-4271
language English
publishDate 2025-03-01
publisher Elsevier
record_format Article
series Energy Nexus
spelling doaj-art-91dcf594aef949488f6237c53aa2737d2025-08-20T01:49:31ZengElsevierEnergy Nexus2772-42712025-03-011710037410.1016/j.nexus.2025.100374Predictive Modeling of soil salinity integrating remote sensing and soil variables: An ensembled deep learning approachSana Arshad0Jamil Hasan Kazmi1Endre Harsányi2Farheen Nazli3Waseem Hassan4Saima Shaikh5Main Al-Dalahmeh6Safwan Mohammed7Department of Geography, The Islamia University of Bahawalpur, Bahawalpur 63100, PakistanDepartment of Geography, University of Karachi, Karachi 75270, PakistanInstitute of Land Use, Technical and Precision Technology, Faculty of Agricultural and Food Sciences and Environmental Management, University of Debrecen, 4032 Debrecen, Hungary; Institutes for Agricultural Research and Educational Farm, University of Debrecen, Böszörményi 138, 4032 Debrecen, HungaryInstitute of Agro industry and Environment, The Islamia University of Bahawalpur, Bahawalpur 63100, PakistanSoil and Water Testing Laboratory for Research Bahawalpur, Bahawalpur 63100, PakistanDepartment of Geography, University of Karachi, Karachi 75270, PakistanHourani Center for Applied Scientific Research, Al-Ahliyya Amman University, Amman, JordanInstitute of Land Use, Technical and Precision Technology, Faculty of Agricultural and Food Sciences and Environmental Management, University of Debrecen, 4032 Debrecen, Hungary; Institutes for Agricultural Research and Educational Farm, University of Debrecen, Böszörményi 138, 4032 Debrecen, Hungary; Corresponding author.Accurate predictions of soil salinity can significantly contribute to achieving the UN- Sustainable Development Goal (SDG-2) of ensuring ‘zero hunger.’ From this perspective, the current research aimed to predict soil electrical conductivity (EC) from remote sensing and soil data using advanced deep learning (DL) architectures. A total of 109 soil samples were analyzed for agricultural land use in the Middle Indus Basin of Pakistan. Seven salinity indices (SI-1 to SI-7) were derived from the 10m to 20m wavelength bands of Sentinel-2, along with vegetation and topographic covariates. Initially, Recursive Feature Elimination was implemented as a feature-selection method to select the most effective predictors. Subsequently, deep learning architectures, including a Feedforward Neural Network (FFNN), Recurrent Neural Network (RNN), and Long Short-Term Memory (LSTM), were employed to predict soil salinity. Research findings showed that EC ranged between 0.57dS/m to 11.5 dS/m in the study area. The evaluation metrics of the DL models revealed that a simple FFNN with three fully connected dense layers achieved the highest R2 = 0.88 for model training. However, the ensemble of improved FFNN and LSTM outperformed with the highest R2 and NSE = 0.84, and the lowest RMSE and MAE = 1.38 and 1.01, respectively, on the testing dataset. Optimized deep learning architectures with adjustments to the learning rate, dropout rate, and activation functions achieved the highest prediction accuracy with the lowest validation loss. Finally, SHapely Additive exPlanations (SHAP) revealed that elevation, pH, NDVI, SI-1, and SI-7 had highly significant impacts on EC predictions. This research provides insight into implementing advanced and interpretable DL architectures, supporting informed decision-making by agricultural stakeholders.http://www.sciencedirect.com/science/article/pii/S2772427125000154Electrical conductivityCanopy response salinity indexFeed forward neural networkPakistan
spellingShingle Sana Arshad
Jamil Hasan Kazmi
Endre Harsányi
Farheen Nazli
Waseem Hassan
Saima Shaikh
Main Al-Dalahmeh
Safwan Mohammed
Predictive Modeling of soil salinity integrating remote sensing and soil variables: An ensembled deep learning approach
Energy Nexus
Electrical conductivity
Canopy response salinity index
Feed forward neural network
Pakistan
title Predictive Modeling of soil salinity integrating remote sensing and soil variables: An ensembled deep learning approach
title_full Predictive Modeling of soil salinity integrating remote sensing and soil variables: An ensembled deep learning approach
title_fullStr Predictive Modeling of soil salinity integrating remote sensing and soil variables: An ensembled deep learning approach
title_full_unstemmed Predictive Modeling of soil salinity integrating remote sensing and soil variables: An ensembled deep learning approach
title_short Predictive Modeling of soil salinity integrating remote sensing and soil variables: An ensembled deep learning approach
title_sort predictive modeling of soil salinity integrating remote sensing and soil variables an ensembled deep learning approach
topic Electrical conductivity
Canopy response salinity index
Feed forward neural network
Pakistan
url http://www.sciencedirect.com/science/article/pii/S2772427125000154
work_keys_str_mv AT sanaarshad predictivemodelingofsoilsalinityintegratingremotesensingandsoilvariablesanensembleddeeplearningapproach
AT jamilhasankazmi predictivemodelingofsoilsalinityintegratingremotesensingandsoilvariablesanensembleddeeplearningapproach
AT endreharsanyi predictivemodelingofsoilsalinityintegratingremotesensingandsoilvariablesanensembleddeeplearningapproach
AT farheennazli predictivemodelingofsoilsalinityintegratingremotesensingandsoilvariablesanensembleddeeplearningapproach
AT waseemhassan predictivemodelingofsoilsalinityintegratingremotesensingandsoilvariablesanensembleddeeplearningapproach
AT saimashaikh predictivemodelingofsoilsalinityintegratingremotesensingandsoilvariablesanensembleddeeplearningapproach
AT mainaldalahmeh predictivemodelingofsoilsalinityintegratingremotesensingandsoilvariablesanensembleddeeplearningapproach
AT safwanmohammed predictivemodelingofsoilsalinityintegratingremotesensingandsoilvariablesanensembleddeeplearningapproach