Principal Components Analysis (PCA) of Water Catchments in BUI Division North West Region of Cameroon: Implications on the Origin of Ions and Correlation

Abstract

Principal Component Analysis (PCA) helps in interpreting data and simplifies the complexity in high-dimensional or multivariate data while retaining trends and patterns. This is done by transforming the data into fewer dimensions that act as summaries of characteristics. The multivariate or high-dimensional data was prevalent in the physico- chemical analysis of samples from some water catchments in Bui division, North West Region of Cameroon. The water catchments were in Elak (ELA) Oku subdision, Mbiame (MBI) Mbven subdivision, Belem (BEL) Nkum subdivision, Nkar (NKA) Jakiri subdivision and YEH in Kumbo subdivision. This type of data presents several challenges that PCA mitigates. With the application of SPSS (Software Package for Social Sciences) it transformed the multivariables obtained in the physico-chemical analysis from the water catchments into a correlation matrix either Pearson’s or Spearman’s matrix. In this work I Pearson’s correlation matrix was used. It reduced the data by geometrically projecting onto lower dimensions in correlation circles. Pearson’s correlation matrix for the water catchments between the variables had a medium or moderate correlation with most of the coefficient values between -/+ 0.39 and +/- 0.49. Nonetheless, some strong positive correlations existed between Mg2+ / pH, TDS / EC, HCO3 - / T°C, and SiO2 / Mg2+, while a strong negative correlation existed between, Mg2+/ T°C, and Mg2+/ NO3 - . Generally, a weak correlation was observed between most of the variables. Eighty-seven variables were correlated from the catchments with 51 having a positive correlation, whereas 36 were negative. Positive correlation values indicated a common source with the ions evolving concomitantly by similar weathering processes or inputs from a common source. This concomitant correlation was expressed between Ca2+ and HCO3 - within the catchments. The negative values revealed that the ions evolve in an antagonistic manner. Consequently, in the water catchments, 58.62 % of the variables had a positive correlation while. 42.48 % had a negative correlation with antagonistic behaviour