Aquifer vulnerability assessment for fecal coliform bacteria using multi-threshold logistic regression.

Area of Science:

• Environmental Science
• Hydrogeology
• Public Health

Background:

• Groundwater pollution poses a significant risk to human health and ecosystems.
• Aquifer vulnerability assessment is essential for effective groundwater resource management and pollution prevention.
• Fecal coliform (FC) contamination is a primary indicator of groundwater's microbial safety.

Purpose of the Study:

• To assess aquifer vulnerability to fecal coliform (FC) pollution in the Pingtung Plain, Taiwan.
• To identify key environmental factors contributing to varying levels of groundwater FC contamination.
• To analyze human health risks associated with incidental ingestion of FC-polluted groundwater.

Main Methods:

• Utilized geographic information systems (GIS) to gather 23 explanatory variables in 2014.
• Employed multi-threshold logistic regression (LR) for aquifer vulnerability modeling and cross-validation.
• Conducted risk assessment to determine thresholds for incidental ingestion of FC-polluted groundwater.

Main Results:

• Identified sand and gravel soil textures and antecedent precipitation as key factors for low-to-medium FC pollution.
• Determined chicken farming, urban land use, and tap water usage ratio as key factors for high FC pollution.
• Risk assessment indicated unacceptable human health risks at medium- and high-level FC contamination thresholds.

Conclusions:

• The multi-threshold logistic regression model effectively characterizes environmental factors influencing groundwater FC pollution.
• Ranking of environmental factors is crucial for accurate aquifer vulnerability assessment.
• Findings provide critical insights for targeted interventions to protect groundwater quality in the Pingtung Plain.