Abstract
We presented a statistical approach employing the hierarchical bayesian approach to analyze environmental monitoring data set containing non-detected observations. The advantages of the presented method are as follows: (1) the method is applicable to data set containing non-detected observations. (2) Uncertainties in parameter estimation are quantitatively analyzed based on posterior distributions. (3) Temporal and spatial variations in the environmental concentrations are separately analyzed and quantified. (4) Distribution of the environmental concentrations in a particular monitoring site in which few monitoring data are available can be predicted based on information from all monitoring sites. (5) Analysis can be implemented by a commonly-used freeware and a simple programming code.
