Harnessing Explainable AI for Sustainable Agriculture: SHAP-Based Feature Selection in Multi-Model Evaluation of Irrigation Water Quality Indices DOI Open Access
Enas E. Hussein, Bilel Zerouali, Nadjem Bailek

et al.

Water, Journal Year: 2024, Volume and Issue: 17(1), P. 59 - 59

Published: Dec. 29, 2024

Irrigation water quality is crucial for sustainable agriculture and environmental health, influencing crop productivity ecosystem balance globally. This study evaluates the performance of multiple deep learning models in classifying Water Quality Index (IWQI), addressing challenge accurate prediction by examining impact increasing input complexity, particularly through chemical ions derived indices. The tested include convolutional neural networks (CNN), CNN-Long Short-Term Memory (CNN-LSTM), CNN-bidirectional Long (CNN-BiLSTM), Gated Recurrent Unit (CNN-BiGRUs). Feature selection via SHapley Additive exPlanations (SHAP) provided insights into individual feature contributions to model predictions. objectives were compare 16 identify most effective approach IWQI classification. utilized data from 166 wells Algeria’s Naama region, with 70% training 30% testing. Results indicate that CNN-BiLSTM outperformed others, achieving an accuracy 0.94 area under curve (AUC) 0.994. While CNN effectively capture spatial features, they struggle temporal dependencies—a limitation addressed LSTM BiGRU layers, which further enhanced bidirectional processing model. importance analysis revealed index (qi) qi-Na was significant predictor both Model 15 (0.68) (0.67). qi-EC showed a slight decrease importance, 0.19 0.18 between models, while qi-SAR qi-Cl maintained similar levels. Notably, included qi-HCO3 minor score 0.02. Overall, these findings underscore critical role sodium levels predictions suggest areas enhancing performance. Despite computational demands model, results contribute development robust management, thereby promoting agricultural sustainability.

Language: Английский

Using machine learning for the assessment of ecological status of unmonitored waters in Poland DOI Creative Commons
Andrzej Martyszunis, Małgorzata Loga, Karol Przeździecki

et al.

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Oct. 18, 2024

Advancements in Artificial Intelligence (AI) technology allow for development of new tools analytics and management which present opportunities field environmental protection. The following study showcases usage Machine Learning (ML) techniques as a complementary method water status assessment bodies. Since the main goal Water Framework Directive (WFD) is to improve quality reach good all bodies across Europe intensive monitoring program was launched together with procedure. Based on requirements European Union's WFD concerning ecological it presented how ML can be used Polish unmonitored river Due absence data, foremost challenge lay securing relevant alternative data set anthropogenic pressures. pivotal solution implementation enable processing seemingly unrelated information pressures catchment. Decision Tree, Random Forest, KNN, Support Vector Machine, Multinomial Naive Bayes, XGBoost models have been tested results indicated most suitable techniques. Study shows highest efficiency Forest algorithms classification were compared by their overall accuracy (OA) reaching approximately 93% binary 72% comprehensive well partial class accuracies Probability Misclassification (PoM) parameter. analyses demonstrates practical application AI case reporting objectives possible full planning operational uses. OA PoM are postulated best measures goodness classification.

Language: Английский

Citations

3
Advancing Deltaic Aquifer Vulnerability Mapping to Seawater Intrusion and Human Impacts in Eastern Nile Delta: Insights from Machine Learning and Hydrochemical Perspective DOI

Nesma A. Arafa,

Zenhom E. Salem, Abdelaziz Abdeldayem

et al.

Earth Systems and Environment, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 16, 2024

Language: Английский

Citations

3
Application of Geospatial and Machine Learning Algorithms for Groundwater Quality Prediction Used for Irrigation Purposes DOI Creative Commons
Hemant Raheja, Arun Goel, Mahesh Pal

et al.

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: Feb. 12, 2024

Abstract The main objective of the present study is to evaluate groundwater quality for irrigation purposes in central-western part Haryana state (India). For this, 272 samples were collected during Pre- and Post-monsoon periods 2022. Several indices, including Sodium Absorption Ratio (SAR), Permeability Index (PI), Percentage (Na %), Kelly (KR), Magnesium Adsorption (MAR), Irrigating water index (IWQI) derived. results terms SAR, Na%, KR values indicate that generally suitable irrigation. On other hand, PI MAR exceeded established limits, primarily showing issues related salinity magnesium content groundwater. Furthermore, according assessment based on IWQI classification, 47.06% 25% total fell under "Severe Restriction irrigation" category Pre-monsoon periods, respectively. Spatial variation maps western portion area unsuitable both periods. Three Machine learning (ML) algorithms, namely Random forest (RF), Support vector machine (SVM), Extreme Gradient Boosting (XGBoost) integrated validated predict IWQI. revealed XGBoost with searchachieves best prediction performances. approaches this have been confirmed be cost-effective feasible quality, using hydrochemical parameters as input variables, highly beneficial resource planning management.

Language: Английский

Citations

1
WQI Improvement Based on XG-BOOST Algorithm and Exploration of Optimal Indicator Set DOI Open Access
Jing Liu, Qi Chu,

Wenchao Yuan

et al.

Sustainability, Journal Year: 2024, Volume and Issue: 16(24), P. 10991 - 10991

Published: Dec. 14, 2024

This paper takes a portion of the Manas River Basin in Xinjiang Province, China, as an example and proposes improved traditional comprehensive water quality index (WQI) method using Extreme Gradient Boosting (XG-BOOST) to analyze groundwater levels region. Additionally, XG-BOOST is used screen existing dataset ten indicators, including fluoride (F), chlorine (Cl), nitrate (NO), sulfate (SO), silver (Ag), aluminum (Al), iron (Fe), lead (Pb), selenium (Se), zinc (Zn), from 246 monitoring points, order find that optimizes model training performance. The results show that, selected study area, categorized “GOOD” “POOR” accounts for majority, with covering 48.7% area 31.6%. Regions classified “UNFIT” are mainly distributed central–eastern parts located Changji Hui Autonomous Prefecture. Comparatively, western part better than eastern part, while areas “EXCELLENT” primarily southern area. optimal indicator consists five indicators: Cl, NO, Pb, Se, Zn, achieving accuracy 98%, RMSE = 0.1414, R2 0.9081.

Language: Английский

Citations

1
Analysis of Water Quality for Taal Lake Using Machine Learning Classification Algorithm DOI

Michelle C. Tanega,

Arnel C. Fajardo, Jomel S. Limbago

et al.

Published: June 28, 2023

The rapid population growth in the Philippines has increased demand for food and aquatic commodities, making fishing a crucial source of income coastal households [1]. Ensuring water quality Philippine lakes is essential to maintaining environment's integrity protecting communities that rely on resources. In this work, we applied machine learning classification algorithms such as Random Forest, Decision Tree, Support Vector calculate Taal Lake, Philippines's Water Quality Index (WQI) Classification (WQC). Weighted Arithmetic (WAWQI) approach was employed classify Lake. Our results showed lake's unsuitable between 2018 2022 at five selected stations. Moreover, evaluated model against two other demonstrated it outperformed Precision, Recall, F-1 score. Forest achieved highest overall accuracy rate 95.0% compared models tested. This study emphasizes importance utilizing monitor Philippines.

Language: Английский

Citations

3
AI-Based Smart Water Quality Monitoring and Wastewater Management DOI
Dipankar Ghosh,

Sayan Adhikary,

Srijaa Sau

et al.

Advances in civil and industrial engineering book series, Journal Year: 2023, Volume and Issue: unknown, P. 127 - 151

Published: Nov. 27, 2023

Water is unambiguously susceptible to contamination, as it able dissolve a broader spectrum of substances than any other liquid on Earth. Increasing population and urbanization have been imposed monitor water quality wastewater management in the current global scenario. Conventional monitoring involves sampling, testing, investigation, which are usually performed manually not dependable. Rapid economic prosperity generates larger quantity enriched with broad range pollutants that pose serious threats environment human health. Advancements artificial intelligence machine learning approaches shown breakthrough potential toward large dataset capture analysis datasets attain complex large-scale systems. The chapter summarizes prospects potentials AI technologies for amelioration establish an integrated sustainable biocomputation platform near future.

Language: Английский

Citations

1
Application of geospatial and machine learning algorithms to predict (under certain limitations) the quality of groundwater used for irrigation purposes DOI Creative Commons
Hemant Raheja, Arun Goel, Mahesh Pal

et al.

Water Science & Technology Water Supply, Journal Year: 2024, Volume and Issue: 24(11), P. 3724 - 3743

Published: Nov. 1, 2024

ABSTRACT The main objective of the present study is to evaluate groundwater quality for irrigation purposes in central-western part Haryana state (India). For this, 272 samples were collected during pre- and post-monsoon periods 2022. Several indices, including SAR, PI, Na%, KR, magnesium adsorption ratio (MAR), IWQI derived. results KR values indicate that generally suitable irrigation. On other hand, PI MAR exceeded established limits, primarily showing issues related salinity content groundwater. Furthermore, according classification, 47.06 25% total fell under ‘severe restriction irrigation’ category pre-monsoon periods, respectively. Spatial variation maps water western portion area unsuitable both periods. Three ML algorithms, namely RF, SVM, XGBoost integrated validated predict IWQI. revealed with random search achieves best prediction performances. approaches this have been confirmed be cost-effective feasible quality, using hydrochemical parameters as input variables, highly beneficial resource planning management.

Language: Английский

Citations

0
Harnessing Explainable AI for Sustainable Agriculture: SHAP-Based Feature Selection in Multi-Model Evaluation of Irrigation Water Quality Indices DOI Open Access
Enas E. Hussein, Bilel Zerouali, Nadjem Bailek

et al.

Water, Journal Year: 2024, Volume and Issue: 17(1), P. 59 - 59

Published: Dec. 29, 2024

Irrigation water quality is crucial for sustainable agriculture and environmental health, influencing crop productivity ecosystem balance globally. This study evaluates the performance of multiple deep learning models in classifying Water Quality Index (IWQI), addressing challenge accurate prediction by examining impact increasing input complexity, particularly through chemical ions derived indices. The tested include convolutional neural networks (CNN), CNN-Long Short-Term Memory (CNN-LSTM), CNN-bidirectional Long (CNN-BiLSTM), Gated Recurrent Unit (CNN-BiGRUs). Feature selection via SHapley Additive exPlanations (SHAP) provided insights into individual feature contributions to model predictions. objectives were compare 16 identify most effective approach IWQI classification. utilized data from 166 wells Algeria’s Naama region, with 70% training 30% testing. Results indicate that CNN-BiLSTM outperformed others, achieving an accuracy 0.94 area under curve (AUC) 0.994. While CNN effectively capture spatial features, they struggle temporal dependencies—a limitation addressed LSTM BiGRU layers, which further enhanced bidirectional processing model. importance analysis revealed index (qi) qi-Na was significant predictor both Model 15 (0.68) (0.67). qi-EC showed a slight decrease importance, 0.19 0.18 between models, while qi-SAR qi-Cl maintained similar levels. Notably, included qi-HCO3 minor score 0.02. Overall, these findings underscore critical role sodium levels predictions suggest areas enhancing performance. Despite computational demands model, results contribute development robust management, thereby promoting agricultural sustainability.

Language: Английский

Citations

0