GIS approaches for the estimation of residential-level ambient PM concentrations
2006

Response to Critique on National Kriging Exposure Estimation

Sample size: 366 publication Evidence: moderate

Author Information

Author(s): Liao Duanping, Peuquet Donna J., Lin Hung-Mo, Duan Yinkang, Whitsel Eric A., Smith Richard L., Heiss Gerardo

Primary Institution: Pennsylvania State University

Hypothesis

Can national-scale kriging effectively estimate daily mean concentrations of PM10 at unmonitored locations in the contiguous United States?

Conclusion

National kriging performs comparably to regional kriging even when restricted to optimal days.

Supporting Evidence

  • The average RMS and SE from 366 daily PM10 cross-validations were 19.48 and 16.19 μg/m3 from log-normal kriging.
  • National kriging was found to be competitive when assessed by overall RMS error.
  • The average daily SD of PM10 measured at monitor locations was 27.20 μg/m3.

Takeaway

The study shows that a method for estimating air pollution can work well across the whole country, even if some areas have fewer data points.

Methodology

The study used log-normal kriging based on a spherical model to interpolate daily PM10 data.

Potential Biases

Manual adjustments to models can lead to inconsistencies in validation measures.

Limitations

Cross-validation may not represent performance at participant address locations.

Participant Demographics

The study involved large cohorts over long periods.

Digital Object Identifier (DOI)

10.1289/ehp.9169

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication