Source apportionment and quality assessment of surface water using principal component analysis and multiple linear regression statistics

Abstract

Principal component analysis (PCA) and multiple linear regressions (MLR) analysis were applied on the data set of surface water quality for source identification of pollution and their contribution on the variation of water quality. Results revealed that, most of the water quality parameters were found to be toxic compare to the national standard of Malaysia. PCA identified the sources as, ionic groups of salts, soil erosion and agricultural runoff, organic and nutrient pollutions from domestic wastewater, industrial sewage and wastewater treatment plants. MLR investigated the contribution of every variable with R= 0.968 and R2=0.934 and it was highly significant (p<0.01).

Keywords

• Multiple linear regressions
• principal component analysis
• source apportionment
• varimax rotation
• water pollution
• water quality