Application of Principal Component Analysis in Determining the Hydrogeochemistry Characteristics of Karst Water in Bac Kan and Cao Bang

Abstract

Research on the origin and processes affecting the hydrogeochemical characteristics of karst water provides useful information for the management and exploitation of karst water resources. Principal Component Analysis (PCA) was applied to the karst water physicochemical data, including temperature, pH, Electronic Conductivity (EC), Ca²⁺, Mg²⁺, Na⁺, K⁺, Cl^-, HCO₃^-, SO₄^2-, and NO₃^- in Ba Be lake (Bac Kan province) and Ha Quang, Trung Khanh (Cao Bang province). The Piper diagram described the main type of hydrochemical facies in karst water at 24/25 survey points as Ca²⁺-HCO₃^-. The combination of eleven physicochemical properties was reduced into four main components (PC1, PC2, PC3, PC4) based on eigenvalues of 4,0; 2,5; 1,3 and 1,2, respectively. A total of 81,53% of the variance was represented by four main components, in which PC1 (36,36%) represented the rock-water interaction, PC2 (22,88%) represented pollution sources related to human activities, PC3 and PC4 (22,29%) represented the dissolution of magnesium minerals. The moderate-strong correlation between parameters of the Mia Le Formation’s sample in Ba Be Lake area and PC1 indicated that the major chemical composition of karst water is strongly influenced by the interaction between carbonate rocks and water. This process also occurs intensely at some survey points in the Bac Son, Song Hien, and Cao Bang formations. Meanwhile, some samples NK.19, NK.23, NK.25 of the Bac Son and Song Hien formations had a moderate positive correlation with PC2 and are affected by human activities. Research results have shown that the combination of principal component analysis PCA and correlation analysis is an effective tool in determining karst water characteristics in the area.