Application of Regression and Correlation Analysis for the Prediction of Groundwater Quality Variables in Benin City, Edo State, Nigeria

Abstract

Purpose: Groundwater is the major source of drinking water in Benin City. The focus of this research is to encourage regular monitoring of groundwater parameters for the assessment of the level of water quality for health benefits.

Methodology: This was carried out by studying interrelationship between parameters measured on spot and those measured in the laboratory. This research attempts to establish regression equations using pH, TDS and DO that are measured on site for prediction of cations and anions prior to their measurement in the laboratory within the study area. Water samples were analyzed for the following parameters pH, Total Dissolve Solids (TDS), Dissolved Oxygen (DO), Bicarbonate (HCO3), Sodium (Na), Potassium (K), Magnesium (Mg), Chloride (Cl­-), Nitrate (NO3), Sulphate (SO4). Correlation analysis with ±0.25value was performed first to investigate the relationship between independent variables (pH, TDS, DO) and dependent variables (cations and anions). Multiple regression models was used to determine significant predictors (with p-value < 0.05) for the prediction of each ion.

Findings: TDS is a significant predictor with more than 95% confidence level for predicting all the ions in both seasons. DO and pH contributed in predicting Cl- and NO32- respectively in wet season. The independent variables (predictors) can easily be done using meter on site. The use of prediction equations will give an over view of groundwater quality, save time, money and resources.

Unique contribution to theory, practice and policy: Awareness programs and enlightenment should be continuously done to educate the people. Government and stake holders should make funds available for more research and enact laws that will improve groundwater quality for human health.