time series analysis of water quality parameters

Water quality is a worldwide problem which affects human beings lives fundamentally. water scarcity is intensified in result of quality deterioration. different factors such as population increase, economic development and water pollution could be considered as the origins of the problem. the study and forecasting of water quality is necessary to prevent serious water quality deteriorations in future. different methodologies have been used to predict and estimate the quality of water. in present study using time series modeling, the quality of hor rood river is studied at kakareza station using time series analysis. 9 parameters of water quality are studied such as: tds, ec, hco^3-, so4^2-, mg^2+, ca^2+, na^+, ph and sar. investigation of observed time series show that there is an increasing trend for all parameters unless na+, ph and sar. the order of model for each parameter was determined using auto correlation function (acf) and partial auto correlation function (pacf) of time series. arima (autoregressive, integrated, moving average) model was found suitable to generate and forecast the quality of river water. aic, r2, rmse and ve % criteria were used for evaluating the generation and forecasting results. results show that time series modeling is quite capable of water quality forecasting. for all generated and forecasted parameters the value of r^2 was greater than 0.66 except for so4^2-. the value of r^2 for generated so4^2- was 0.48 and this value was 0.43 for forecasting this parameter. also the study show that the quality of water is deteriorating based on an increasing trend for the majority of parameters and needs serious managerial actions.