Award details

Complex spatial variation of environmental variables: sampling, prediction and interpretation

Reference	BBS/E/C/00004943
Principal Investigator / Supervisor	Professor Walter Gilks
Co-Investigators / Co-Supervisors	Dr Ben Marchant
Institution	Rothamsted Research
Department	Rothamsted Research Department
Funding type	Research
Value (£)	959,082
Status	Completed
Type	Institute Project
Start date	01/04/2008
End date	31/03/2012
Duration	48 months

Abstract

The aim of this project is to develop theoretically sound and practically workable and robust methods to allow cost-effective prediction, estimation and scientific interpretation of complex spatial variation. These methods are to be applied to understanding the behaviour of soil systems, and to guide their management. The specific scientific objectives are as follows: 1. To extend and critically assess the linear mixed model framework for spatial analysis of environmental properties to cover complex and non-stationary variables. 2. To develop Bayesian approaches to spatial prediction, with particular emphasis on how this allows us to incorporate process information and use soft data or data of variable quality. 3. To invent methods for spatial analysis based on the wavelet transform, with particular interest in multivariate problems and irregularly sampled data. 4. To tackle the problem of modelling spatio-temporal covariance for complex processes. Objective 1 will deliver a suite of methods for spatial analysis and prediction, and for optimizing sample design. The methods will also allow us to partition a complex data set into realizations of a background process and of a contaminating process. The methods will be exemplified in case studies with collaborators such as the British Geological Survey and INRA Objective 2 will deliver an evaluation of an approach to spatial analysis that has attracted much interest but which is thought to have much unrealized potential (particularly with respect to the incorporation of process understanding). Objective 3 will develop substantial advances in wavelet methodology, allowing these powerful methods to be applied to new data and to new problems (particularly with respect to the characterization of multivariate spatial variation). Objective 4 will give insight into the complex spatio-temporal variation of some key processes, and how they may best be sampled and predicted.

Summary

unavailable

Committee	Closed Committee - Engineering & Biological Systems (EBS)
Research Topics	Soil Science, Technology and Methods Development
Research Priority	X – Research Priority information not available
Research Initiative	X - not in an Initiative
Funding Scheme	X – not Funded via a specific Funding Scheme

I accept the terms and conditions of use (opens in new window)

export PDF file

back to list new search