Aggregated multivariate environmental risk (AMER) index and predictive models for heavy metal pollution trends in ecosystems

Abstract

Heavy metal (HM) contamination in estuarine sediments represents a critical environmental challenge, with profound implications for ecosystem health and human safety. To address this pressing issue, proactive monitoring and predictive modeling have emerged as essential tools. This study investigates the application of three predictive models—Grey Model (GM (1,1)), Autoregressive Integrated Moving Average (ARIMA), and Exponential Smoothing (ES)—to analyze historical and current data on eight key HMs (Cu, Zn, Cd, Pb, Ni, Cr, Mn, and Fe) in sediment samples from the Rosetta and Damietta estuaries in Egypt. Utilizing these models, five-year contamination trends were forecasted, incorporating both single-element pollution indices (SEPIs) and multiple-element pollution indices (MEPIs) to evaluate pollution levels and elucidate complex metal interactions. Furthermore, this study introduces a new MEPI, the Aggregated Multivariate Environmental Risk (AMER) index, derived from Principal Component Analysis (PCA). The AMER index demonstrated approximately 20 % higher responsiveness—measured by earlier detection and greater sensitivity to pollution fluctuations—compared to traditional multimetal indices such as PERI. Among the models evaluated, ES exhibited the highest predictive accuracy, achieving a mean absolute percentage error (MAPE) of 6.3 % for Cu predictions. Significant positive correlations were identified between Pb and Ni (r = 0.77, p < 0.05) in Damietta and Cu and Zn (r = 0.65, p < 0.05) in Rosetta, suggesting common pollution sources. The five-year forecasts indicate a declining trend for most metals, reflecting the efficacy of recent pollution control measures. This integrated modeling approach provides enhanced pollution assessment and forecasting capabilities, offering valuable insights for the development of sustainable estuarine management strategies.