Assessing the Presence of Metals in Surface Waters: A Case Study Conducted in Algeria Using a Combination of Artificial Neural Networks and Multiple Indices

Abstract

Elevated concentrations of heavy metals in wetlands can contaminate surface water, posing hazards to human health and ecological balance. Given increasing urbanization and activities in places like Algeria, it is crucial to closely monitor and effectively control heavy metal pollution in surface water. This study proposes the use of artificial neural networks (ANN) and various indicators to comprehensively assess metal contamination in Algerian surface waters and its implications for public health. Sixteen water samples were collected for the composition analysis and source identification. Measurements indicated that several areas exceed the World Health Organization (WHO) limits for four metals. Methods such as the heavy metal evaluation index (HEI) and heavy metal pollution index (HPI) were employed to assess pollution levels. Results showed that over 99% of samples exhibited significant pollution according to HPI, with 60% showing elevated pollution levels by HEI, highlighting substantial contamination risks. Principal component analysis (PCA) revealed that the first two components accounted for 93.540% of total variation, with subsequent components contributing 6.459% or less. PCA 1 and PCA 2, representing 49.084 and 44.456% of variability, respectively, were identified as primary components, while PCA 3 and PCA 4 each contributed less than 5.015 and 1.444% to total variance. The study demonstrated minimal error values and R² values exceeding 0.5 during the testing of heavy metal models, indicating robust performance. Overall, this study underscores the prevalence of elevated metal levels in water bodies, providing comprehensive insights into heavy metal contamination in Algerian basins to assist environmental management decisions and protect public health.