A Kriging-based Estimate of the Survey Quality in Biomonitoring of Trace Element Air Pollution

Abstract

The present paper addresses Kriging-based approaches to estimate survey quality, thereby using data on biomonitoring surveys of trace element air pollution. Data were used of a 1995 moss monitoring study in The Netherlands, thereby taking sodium (Na), arsenic (As) and bromium (Br) to illustrate the points made. Survey quality was expressed as the survey signal-to-noise ratio (SSNR), with the survey variance as the signal and the local variance as the noise. The Kriging approaches were regarded in parallel to direct approaches where local variances are determined from multi-fold local sampling.

The paper addresses the uncertainties in SSNR associated to the uncertainties in local values and variances if based on pooled samples and a limited 5-fold sub-sampling in a restricted number of sampling sites. It further introduces a Kriging approach to SSNR _K ( _K for Kriging), by Kriging-assisted recalculation of local values and variances, in turn based on fitted experimental variograms. The outcomes suggest that the Kriging approach may yield reliable SSNR _K values; the approach implies that SSNR _K depends on survey set-up, site density and sampling grid characteristics. The discussion of the paper addresses the points made and further suggests the use of large sample analytical facilities to permit a change in survey approach from mixing and milling of pooled sub-samples to bulk analyses of all sampled material in higher site-density surveys.