Evaluation of two methods of interpolating quarterly trihalomethane levels between sampling dates

Area of Science:

• Environmental Health
• Epidemiology
• Water Quality Monitoring

Background:

• Epidemiological studies often rely on single quarterly samples for disinfection byproduct (DBP) exposure assessment.
• Trihalomethanes (THMs), a common DBP class, exhibit significant temporal fluctuations, potentially misrepresenting exposure over pregnancy.
• Existing methods may not adequately capture the dynamic nature of THM concentrations in drinking water.

Purpose of the Study:

• To evaluate interpolation methods for improving exposure estimates of trihalomethanes (THMs) using available water quality data.
• To compare the accuracy of cubic splines and linear interpolation against actual quarterly measurements.

Main Methods:

• Utilized monthly THM sampling data from the New York City water supply system.
• Employed cubic spline and linear interpolation techniques to estimate THM levels between quarterly compliance samples.
• Validated interpolated values against actual data from the remaining months within each quarter.

Main Results:

• Both interpolation methods demonstrated generally acceptable fits, with 90% of discrepancies under 14 µg/l for total THMs (THM4).
• Linear interpolation outperformed cubic splines, showing lower average discrepancies and fewer large deviations.
• Interpolated values were significantly more precise than typical quarterly variations and the range of values used in many existing studies.

Conclusions:

• Linear interpolation offers a practical and effective approach to enhance exposure estimates from sparse quarterly water quality monitoring data.
• This method can improve the accuracy of epidemiological studies linking DBPs to adverse birth outcomes by better characterizing exposure.
• The findings support the use of linear interpolation for more robust environmental exposure assessments in public health research.