Modeling and Prediction of Water Quality: Evidence of Benarpota Station of Sundarbans Region

Abstract

The Sundarbans contain the world's largest coastal mangrove forest, with an area of about 10,000 km2 (3,900 sq mi), of which about 6,000 km2 (2,300 sq mi) are located in Bangladesh and about 4,000 km2 (1,500 sq mi) in India. Salinity will be the major issue in the future which could destroy the natural ecological balance. The major salinity components are electric conductivity (EC) at low tide and high tide with the corresponding chloride in low tide and high tide. This study develops a SARMA (2, 0, 2) (2, 0, 0) [4] model as it provides the minimum value of model selection criterion (AIC, BIC, SIC, RMSE) compare to all other models. It was found that the bestfitted model for EC in a high tide of study area follows a seasonal autoregressive moving average model of order (2, 0, 2)(2, 0, 0) [4] and chloride in high tide follows SARMA (1, 0, 1) (2, 0, 2) [4]. After fitting the time series model and forecasting the fitted model, the study showed that in both electric conductivity (EC) and salinity (Chloride ppm) increasing in the last three decades. EC crosses the poor water quality level which is more than 4 ppt or 2560 ppm. Salinity also crosses the poor water quality level which is more than 5 ppt or 3200 ppm. The findings of this study would be a potential contribution to make a comprehensive management plan for the long-term conservation and protection of the cultural landscape and mangrove wetlands ecosystem in the Sundarbans region.