Regression Model for Predicting Water Quality Parameters: Study of Igando, Lagos-Nigeria

Abstract

Fifteen groundwater samples were collected from hand dug well. The depths of wells were measured and the water quality analyzed for 7 heavy metals (Fe, Cu, Mn, Pb, Zn, Cr and Cd) after standard method. The study aimed at predicting groundwater quality parameters with depth with a view to ascertain its quality using both correlation and regression model. The results show that depths of hand dug wells ranged between 15.5 m-22.60 m with a mean of 19.47.Iron,Copper, Cadmium, Manganese, Lead, Zinc and Chromium ranged between (0.00-10.16, 0.02-8.71, 0.02-0.32, 0.04-30.00, 0.00-3.14, 1.40-55.18 and 0.00-0.04 mg/L) respectively. The Coefficient of variation revealed that all the examined groundwater parameters with the exception of Chromium are highly variable. All the parameters examined with the exception of Chromium were found to be above the maximum permissible limit for drinking water standard in the area. The regression coefficient of determination accounts for 75.17%.The regression model significantly explains how the depth of well affects groundwater parameters representing about 25% of the information not accounted for in the model. The functional relationship and the degree of correlation between depth of well and parameter indicates that among all the parameter examined only Iron, Copper and Zinc have a significant linear association with depth at 0.05 level of significance (0.024, 0.020 and 0.001 respectively).The study demonstrates the effectiveness of regression model in predicting water quality parameters for the present and future purposes.