Water Quality Monitoring on Streaming Data DOI Creative Commons

Bhawnesh Kumar,

Tinku Singh, Anuj Kumar

et al.

Deleted Journal, Journal Year: 2024, Volume and Issue: 3(2), P. 98 - 113

Published: Feb. 12, 2024

The increasing contamination of natural water bodies due to diverse human activities necessitates a comprehensive approach monitoring quality, especially considering its widespread use in daily life. This study addresses the escalating bodies, emphasizing need for robust real-time quality system. Focused on evaluating Triveni Sangam, Prayagraj, where Ganga and Yamuna rivers converge, recognizes crucial role continuous safeguarding precious resources. To achieve this, sophisticated framework has been proposed, leveraging Spark server simulate streaming data. dynamic ensures uninterrupted assessment effective management system categorizes training data using Water Quality Index (WQI) employs Naive Bayes classification data, achieving an impressive accuracy 82.21%. results underscore effectiveness learning from utility real-time. contributes significantly ongoing resource initiatives but also highlights pivotal machine addressing pressing environmental challenges.

Language: Английский

A machine learning approach to mapping suitable areas for forest vegetation in the eThekwini municipality DOI Creative Commons
Mthokozisi Ndumiso Mzuzuwentokozo Buthelezi, Romano Lottering, Kabir Peerbhay

et al.

Remote Sensing Applications Society and Environment, Journal Year: 2024, Volume and Issue: 35, P. 101208 - 101208

Published: April 23, 2024

Driven by climate change, global forests are undergoing significant transformations in growth, ecology, and distribution, necessitating informed restoration conservation strategies, particularly the eThekwini Municipality where anthropogenic activities exacerbate these trends. Modelling current forest suitability (2023) utilized bioclimatic variables from WorldClim dataset, alongside elevation slope Shuttle Radar Topography Mission (SRTM) with remote sensing data acquired Landsat 9 Sentinel 2A. Future (2021 – 2040) was projected also using two Global Climate Models (GCMs) under four Shared Socioeconomic Pathway (SSP)-based Representative Concentration (RCP) scenarios. Employing Random Forests (RF), Light Gradient Boosting (LightGBM), Artificial Neural Networks (ANN), processing carried out Google Earth Engine (GEE), QGIS Python, model accuracy primarily assessed Receiver Operating Characteristic (ROC) curves Area Under ROC Curve (AUC). LightGBM demonstrated superior performance, achieving AUCs of 96.88% 93.75% for future mapping, respectively, annual precipitation vegetation changes identified as crucial variables. Currently, 30% municipality's land is deemed suitable, concentrated central region. projections highlight mountainous north-western region most notably SSP370 scenario a suitable area 63%. Strategic recommendations include prioritizing reforestation efforts, engaging private landowners, exploring urban opportunities, implementing continuous monitoring adaptive management, thereby enhancing carbon sequestration, biodiversity conservation, ecosystem resilience. This study provides valuable insights decision-making despite inherent uncertainties.

Language: Английский

Citations

5
Integration of Watershed eco-physical health through Algorithmic game theory and supervised machine learning DOI Creative Commons
Ali Nasiri Khiavi,

Mohammad Tavoosi,

Hamid Khodamoradi

et al.

Groundwater for Sustainable Development, Journal Year: 2024, Volume and Issue: 26, P. 101216 - 101216

Published: May 31, 2024

The eco-physical health assessment of watersheds is crucial for sustainable water resource management and ecosystem services. This study quantifies the Talar watershed in Iran using geometric mean method (GMM), game-theoretic algorithm (GTA), machine learning algorithms including Random Forest (RF), Support Vector Machine (SVM), Simple Linear Regression (SLR), K-Nearest Neighbor (KNN) distributed semi-distributed monitoring. results show that RF performed better than other models, as indicated by MAE, MSE, RMSE, AUC statistics with values 0.032, 0.003, 0.058, 0.940, respectively. index prioritization different approaches showed pattern changes positively from upstream to downstream. Based on GMM, it can be said sub-watersheds Int6 Int5 are healthiest studied watershed, 0.93 0.90, GTA approach, also Int6, Int5, Int01 ones. In case algorithm, average pixels each sub-watershed were recognized 0.91 0.88, consistently emerged across all methods, attributed high TWI NDVI low slope, DEM, erosion, CN values. general, catchment fully followed factors affecting catchment's spatial patterns change this consistent physiographic hydroclimatic conditions three approaches. study's implications underline importance multi-criteria multi-algorithm accurately assessing managing development.

Language: Английский

Citations

4
Enhancing BOD5 Forecasting Accuracy with the ANN-Enhanced Runge Kutta Model DOI

Rana Muhammad Adnan,

Ahmed A. Ewees, Mo Wang

et al.

Journal of environmental chemical engineering, Journal Year: 2025, Volume and Issue: unknown, P. 115430 - 115430

Published: Jan. 1, 2025

Language: Английский

Citations

0
Assessing the feasibility of using Machine learning algorithms to determine reservoir water quality based on a reduced set of predictors DOI
Natalia Walczak, Zbigniew Walczak

Ecological Indicators, Journal Year: 2025, Volume and Issue: 175, P. 113556 - 113556

Published: May 2, 2025

Language: Английский

Citations

0
Integration of Machine Learning Augmented With Biosensors for Enhanced Water Quality Monitoring DOI
Pokkuluri Kiran Sree, S. S. S. N. Usha Devi N., Alex Khang

et al.

Advances in environmental engineering and green technologies book series, Journal Year: 2024, Volume and Issue: unknown, P. 181 - 192

Published: Jan. 22, 2024

Monitoring water quality is essential to guaranteeing the sustainability and safety of supplies. Conventional techniques for evaluating might be laborious may not able provide results instantly. The suggested system makes use a wide range biosensors assess important aspects quality, including microbial activity, pH, dissolved oxygen, chemical pollutants. Following collection, data are analysed using recurrent neural networks (RNNs). An RNN trained identify patterns, correlate information from several sensors, forecast trends in quality. Early detection problems with prompt reaction possible contaminants, flexibility response changing environmental conditions some benefits this integrated approach. enhanced monitoring (BEWQM) useful tool long-term management because its learning characteristics, which allow it continuously improve accuracy performance over time.

Language: Английский

Citations

3
The Contribution of Open Source Software in Identifying Environmental Crimes Caused by Illicit Waste Management in Urban Areas DOI Creative Commons
Carmine Massarelli, Vito Felice Uricchio

Urban Science, Journal Year: 2024, Volume and Issue: 8(1), P. 21 - 21

Published: March 19, 2024

This study focuses on the analysis, implementation and integration of techniques methods, also based mathematical algorithms artificial intelligence (AI), to acquire knowledge some phenomena that produce pollution with an impact environmental health, which start from illicit practices occur in urban areas. In many areas (or agroecosystems), practice illegal waste disposing by commercial activities, abandoning it countryside rather than spending economic resources ensure correct disposal, is widespread. causes accumulation these (which can be protected natural areas), are then set fire reduce their volume. Obviously, repercussions such actions many. The burning releases contaminants into environment as dioxins, polychlorinated biphenyls furans, deposits other elements soil, heavy metals, which, leaching percolating, contaminate water rivers aquifers. main objective design monitoring programs against specific activities take account territorial peculiarities. advanced approach leverages AI GIS environments interpret states, providing understanding ongoing phenomena. methodology used algorithms, integrated a address even large-scale issues, improving spatial temporal precision analyses allowing customization peri-urban characteristics. results application show percentages different types found agroecosystems area degree concentration, identification similar greater criticality. Subsequently, through network nearest neighbour possible targeted checks.

Language: Английский

Citations

3
A Critical Review of the Modelling Tools for the Reactive Transport of Organic Contaminants DOI Creative Commons

Katarzyna Samborska-Goik,

Marta Pogrzeba

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(9), P. 3675 - 3675

Published: April 25, 2024

The pollution of groundwater and soil by hydrocarbons is a significant growing global problem. Efforts to mitigate minimise risks are often based on modelling. Modelling-based solutions for prediction control play critical role in preserving dwindling water resources facilitating remediation. objectives this article to: (i) provide concise overview the mechanisms that influence migration improve understanding processes affect contamination levels, (ii) compile most commonly used models simulate fate subsurface; (iii) evaluate these terms their functionality, limitations, requirements. aim enable potential users make an informed decision regarding modelling approaches (deterministic, stochastic, hybrid) match expectations with characteristics models. review 11 1D screening models, 18 deterministic 7 stochastic tools, machine learning experiments aimed at hydrocarbon subsurface should solid basis capabilities each method applications.

Language: Английский

Citations

3
A Novel Deep Learning Approach for Real-Time Critical Assessment in Smart Urban Infrastructure Systems DOI Open Access
Abdulaziz Almaleh

Electronics, Journal Year: 2024, Volume and Issue: 13(16), P. 3286 - 3286

Published: Aug. 19, 2024

The swift advancement of communication and information technologies has transformed urban infrastructures into smart cities. Traditional assessment methods face challenges in capturing the complex interdependencies temporal dynamics inherent these systems, risking resilience. This study aims to enhance criticality geographic zones within cities by introducing a novel deep learning architecture. Utilizing Convolutional Neural Networks (CNNs) for spatial feature extraction Long Short-Term Memory (LSTM) networks dependency modeling, proposed framework processes inputs such as total electricity use, flooding levels, population, poverty rates, energy consumption. CNN component constructs hierarchical maps through successive convolution pooling operations, while LSTM captures sequence-based patterns. Fully connected layers integrate features generate final predictions. Implemented Python using TensorFlow Keras on an Intel Core i7 system with 32 GB RAM NVIDIA GTX 1080 Ti GPU, model demonstrated superior performance. It achieved mean absolute error 0.042, root square 0.067, R-squared value 0.935, outperforming existing methodologies real-time adaptability resource efficiency.

Language: Английский

Citations

2
Financial Analytics with Artificial Neural Networks: Predicting Loan Repayment DOI

Siddharth Thakar,

Deep Patel, Vaibhav Gandhi

et al.

Published: Oct. 3, 2024

Language: Английский

Citations

1
Water Quality Monitoring on Streaming Data DOI Creative Commons

Bhawnesh Kumar,

Tinku Singh, Anuj Kumar

et al.

Deleted Journal, Journal Year: 2024, Volume and Issue: 3(2), P. 98 - 113

Published: Feb. 12, 2024

The increasing contamination of natural water bodies due to diverse human activities necessitates a comprehensive approach monitoring quality, especially considering its widespread use in daily life. This study addresses the escalating bodies, emphasizing need for robust real-time quality system. Focused on evaluating Triveni Sangam, Prayagraj, where Ganga and Yamuna rivers converge, recognizes crucial role continuous safeguarding precious resources. To achieve this, sophisticated framework has been proposed, leveraging Spark server simulate streaming data. dynamic ensures uninterrupted assessment effective management system categorizes training data using Water Quality Index (WQI) employs Naive Bayes classification data, achieving an impressive accuracy 82.21%. results underscore effectiveness learning from utility real-time. contributes significantly ongoing resource initiatives but also highlights pivotal machine addressing pressing environmental challenges.

Language: Английский

Citations

0