MULTIVARIATE STATISTICAL ANALYSIS OF CONTINUOUS AUTOMATIC SURFACE WATER QUALITY DATA AT BA LANG AND CAI CON STATIONS IN CAN THO CITY

Abstract

The study utilized multivariate statistics to analyze water quality data continuously measured at two monitoring stations in Can Tho City, Vietnam. Surface water quality is assessed using national technical regulations on surface water (QCVN 08:2023/BTNMT). The water quality difference between two monitoring stations and between months in 2023 was analyzed using the Independent-Sample Test and One-way ANOVA. The relationship between water quality indicators is analyzed using the Pearson correlation. Cluster analysis (CA) is used to group water quality over time and suggest monitoring frequency, while principal component analysis (PCA) is used to identify the number of main sources affecting the water quality monitoring parameters. Surface water quality classification results show that pH and NH₄⁺-N are in level A while TSS is in level B, DO is in level D, COD is in level C - D. Water quality is site and time dependent. Cluster analysis results show that water quality fluctuates greatly over time and the months to be monitored are January, March, May, July, September, November. PCA identified 2 - 3 main sources that have a significant impact on the parameters of salinity, pH, DO, COD, TSS, NH₄⁺-N. The current study results contribute to providing important information for designing the frequency of monitoring surface water quality in Can Tho city. Subsequent studies need to identify monitoring indicators to continue perfecting the monitoring network in the study area.