Modeling and Forecasting of Total Dissolved Solids for Irrigation Water Quality Assessment

Abstract

Industrial Effluents when discharged into water bodies contain untreated or partially treated substances with an enhanced
concentration of nutrient and sediments which poses serious negative impact on the quality and life forms of the receiving water
body .Ele River is surrounded by clusters of industries popularly among them which has a direct discharge outlet into the River is
Chicason industries Limited. Effluents discharged from these industries contain physico-chemical and heavy metal properties
predominantly among them is total dissolved solids (TDS) which comprise inorganic compounds such as salts, heavy metals and
some traces of organic compounds that are dissolved in water. The aim of this study is to model and forecast Ele river TDS using
artificial neural network for irrigation purposes. Monthly Water samples were collected at ten sampling points of 10m interval for
a period of forty-eight months. The water samples were analyzed for TDS concentrations. The results show that during rainy
season, The TDS exceeded the FAO permissible standard from point 1 to point 3 and later decreased as the river flows down the
river course. Also, the results show that during the dry season, the river TDS concentration values was found to have exceeded
the FAO permissible range from point 1 to Point 5 Generally, the river quality from point 1 to point 5 is not safe for agricultural
production as the TDS concentration was above FAO guideline for irrigation while from point 6 to 10, the river is safe for
agricultural production. The artificial neural network was able to model the River TDS very well as the R2 value for the model
training, testing and forecasting are 0.981 to 0.988, 0.953 to 0.970 and 0.946 to 0.968, respectively. It was recommended that the
river water from points 1 to point 5 needs to be treated before use for agricultural production.