A three-dimensional spatial analysis of fine-scale soil pH data
by
Kathy Haskard
University of Adelaide
Coauthors: Brian Cullis (NSW Agriculture), Mark Conyers (NSW Agriculture)

10 cm x 10 cm x depth 6 cm blocks of soil were divided into 1 cm cubes, and pH obtained for each cube. Spatial models were fitted, based on a 3-dimensional separable process, using as a starting point AR1 x AR1 for the horizontal dimensions x antedependence order 1 in the vertical dimension. Antedependence models are described, including a motivation as a generalisation of autoregressive models, and (with maximum order) their use as an alternative parametrisation of a general covariance matrix, as suggested by Pourahmadi (Biometrika 1999). Issues of non-stationarity, separability, model checking and diagnostics are also discussed.

This approach could also be applied to spatio-temporal data.

Date received: August 29, 2002

Copyright © 2002 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cajn-21.