Sampling strategies for toxic air contaminants

Area of Science:

• Environmental Science
• Toxicology
• Statistics

Background:

• Assessing chronic human exposure to toxic air contaminants is crucial for risk analysis.
• Limited ambient concentration measurements due to resource constraints hinder accurate environmental risk assessments.
• Understanding data representativeness and uncertainty is vital before quantitative risk analysis.

Purpose of the Study:

• To discuss factors for designing effective field-sampling programs for toxic air contaminants.
• To examine SO2, TSP, and CO data as surrogates for various pollutant types in sampling design.
• To present statistical methods for estimating annual mean concentrations, confidence bounds, and determining optimal sample sizes for regulatory compliance.

Main Methods:

• Utilizing bootstrap resampling and normal theory to estimate annual mean concentrations and 95% confidence intervals from limited data.
• Applying operating characteristic (OC) curves to determine optimal sample sizes and sampling strategies.
• Implementing a one-sided t-test procedure for evaluating compliance with ambient guideline concentrations.

Main Results:

• Demonstrated methods for estimating reliable annual mean concentrations and their uncertainties from restricted datasets.
• Illustrated how OC curves can optimize sampling strategies and sample sizes for toxic air contaminants.
• Provided a statistical framework for assessing regulatory compliance using sampled concentration data.

Conclusions:

• Statistical methods like bootstrapping and OC curves enhance the reliability of risk assessments with limited air quality data.
• Optimized sampling designs are essential for cost-effective monitoring and regulatory compliance of toxic air contaminants.
• The proposed statistical procedures aid in making informed decisions regarding environmental regulations and human exposure assessment.