Using isometric log-ratio in compositional data analysis for developing a groundwater pollution index.

Area of Science:

• Environmental Science
• Hydrogeology
• Data Science

Background:

• Groundwater quality assessment is crucial for environmental protection and public health.
• Traditional statistical methods often fail to account for the relative nature of hydrochemical data.
• Assessing pollution from events like disease outbreaks requires robust analytical tools.

Purpose of the Study:

• To introduce a novel groundwater pollution index (GPI) for enhanced water quality assessment.
• To apply compositional data analysis (CoDa) and robust principal component analysis (RPCA) for developing the GPI.
• To address the limitations of traditional methods in analyzing hydrochemical data.

Main Methods:

• Utilized groundwater quality data from South Korea impacted by a 2010-2011 disease outbreak.
• Applied CoDa to identify hydrochemical differences between influenced and background groundwater.
• Developed the GPI using RPCA, isometric log-ratio (ILR) transformation, and normalization.

Main Results:

• The GPI effectively distinguishes between polluted and unpolluted groundwater based on key parameters (NH4+-N, Cl-, NO3--N).
• Validation against government criteria and receiver operating characteristic curve analysis confirmed the GPI's reliability.
• CoDa, particularly ILR transformation, proved essential in overcoming traditional statistical limitations.

Conclusions:

• The novel GPI offers a more precise and accurate method for groundwater pollution assessment.
• This approach significantly advances groundwater quality monitoring and management strategies.
• The study highlights the importance of advanced statistical techniques in hydrogeological research.