Interpolating Vancouver's daily ambient PM10 field

Abstract

In this article we develop a spatial predictive distribution for the ambient space – time response field of daily ambient PM10 in Vancouver, Canada. Observed responses have a consistent temporal pattern from one monitoring site to the next. We exploit this feature of the field by adopting a response model with two components, a common deterministic trend across all sites plus a stochastic residual. We are thereby able to whiten the temporal residuals without losing much of the spatial correlation in the original log-transformed series. This in turn enables us to develop an effective spatial predictive distribution for these residuals at unmonitored sites. By transforming the predicted residuals back to the original data scales, we can impute Vancouver's daily PM10 field for purposes such as human exposure and health impacts analysis. Copyright © 2000 John Wiley & Sons, Ltd.