Neural Network-Based Modeling of Water Quality in Jodhpur, India

A case study of using artificial neural networks to predict heavy metal pollution in Lake Iznik

Artificial neural networks offer a viable route in assessing and understanding the presence and concentration of heavy metals that can cause dangerous complications in the wider context of water quality prediction for the sustainability of the ecosystem. In order to estimate the heavy metal concentrations in Iznik Lake, which is an important water source for the surrounding communities, characterization data were taken from five different water sources flowing into the lake between 2015 and 2021. These characterization results were evaluated with IBM SPSS Statistics 23 software, with the addition of the lake water quality system. For this purpose, seven distinct physicochemical parameters were measured and monitored in Karasu, Kırandere, Olukdere and Sölöz water sources flowing into the lake, to serve as input data. Concentration levels of 15 distinct heavy metals in Karsak Stream originating from the lake were as the output. Specifically, Sn for Karasu (0.999), Sb for Kırandere (1.000), Cr for Olukdere (1.000) and Pb and Se for Sölöz (0.995) indicate parameter estimation R2 coefficients close to 1.000. Sn stands out as the common heavy metal parameter with best estimation prospects. Given the importance of the independent variable in estimating heavy metal pollution, conductivity, COD, COD and temperature stood out as the most effective parameters for Karasu, Olukdere, Kırandere and Sölöz, respectively. The ANN model emerges as a good prediction tool that can be used effectively in determining the heavy metal pollution in the lake as part of the efforts to protect the water budget of Lake Iznik and to eliminate the existing pollution.

Water Quality Prediction Based on Machine Learning and Comprehensive Weighting Methods

In the context of escalating global environmental concerns, the importance of preserving water resources and upholding ecological equilibrium has become increasingly apparent. As a result, the monitoring and prediction of water quality have emerged as vital tasks in achieving these objectives. However, ensuring the accuracy and dependability of water quality prediction has proven to be a challenging endeavor. To address this issue, this study proposes a comprehensive weight-based approach that combines entropy weighting with the Pearson correlation coefficient to select crucial features in water quality prediction. This approach effectively considers both feature correlation and information content, avoiding excessive reliance on a single criterion for feature selection. Through the utilization of this comprehensive approach, a comprehensive evaluation of the contribution and importance of the features was achieved, thereby minimizing subjective bias and uncertainty. By striking a balance among various factors, features with stronger correlation and greater information content can be selected, leading to improved accuracy and robustness in the feature-selection process. Furthermore, this study explored several machine learning models for water quality prediction, including Support Vector Machines (SVMs), Multilayer Perceptron (MLP), Random Forest (RF), XGBoost, and Long Short-Term Memory (LSTM). SVM exhibited commendable performance in predicting Dissolved Oxygen (DO), showcasing excellent generalization capabilities and high prediction accuracy. MLP demonstrated its strength in nonlinear modeling and performed well in predicting multiple water quality parameters. Conversely, the RF and XGBoost models exhibited relatively inferior performance in water quality prediction. In contrast, the LSTM model, a recurrent neural network specialized in processing time series data, demonstrated exceptional abilities in water quality prediction. It effectively captured the dynamic patterns present in time series data, offering stable and accurate predictions for various water quality parameters.