Statistical Approach in Determining the Spatial Changes of Surface Water Quality at the Upper Course of Kano River, Nigeria

In this study, statistical techniques were used to evaluate the spatial variation of surface water quality and pollution source apportionment in Kano River, Nigeria. Triplicate water samples were collected from 30 sampling sites along the upper course of Kano River and analyzed for 23 parameters which include dissolved oxygen (DO), 5-day biochemical oxygen demand (BOD5), chemical oxygen demand, pH, temperature, salinity, conductivity, dissolved solids, suspended solids, total solids, turbidity, chloride (Cl), ammonia (NH3), nitrate (NO3), potassium (K), magnesium (Mg), sodium (Na), calcium (Ca), phosphate (PO4), iron (Fe), zinc (Zn), Escherichia coliform (E. coli), and total coliform (T. coli). Hierarchical agglomerative cluster analysis grouped the 30 sampling sites into three statistically significant clusters based on similarities of surface water quality characteristics. Principal component and factor analyses (PCA and FA) were used to investigate the source of water quality parameters and to identify three major water pollution sites: high pollution sites, moderate pollution site, and low pollution site which explained more than 65 % of the total variance in water quality. Discriminant analysis provided a better result with great discriminatory ability, pattern recognition, and important data reduction using only seven parameters (DO, BOD5, pH, NH3, Cl, E. coli, and T. coli) for spatial variation and affording more than 90 % correct cases assignation. Further, one-way analysis of variance (one-way ANOVA) was applied to the three factors obtained from PCA and FA to compare the variation within and between the factors, the result showed significant differences (p < 0.05) between the factors. The study provides a new insight into the environmental quality in Kano River for effective surface water quality management and protection.