Systematic sample design for the estimation of spatial means

Publisher Website
Google Scholar
Abstract
This article develops a practical approach to undertaking systematic sampling for the estimation of the spatial mean of an attribute in a selected area. A design‐based approach is used to estimate population parameters, but it is combined with elements of a model‐based approach in order to identify the spatial correlation structure, to evaluate the relative efficiency of the sample mean under simple random and systematic sampling, to estimate sampling error and to assess the sample size needed in order to achieve a desired level of precision. Using two case studies (land use estimation and weed seedbank in soil) it is demonstrated how the practical basis for the design of systematic samples provided in this work should be applied and it is shown that if the spatial correlation is ignored the sampling error of the sample mean and the sample size needed in order to achieve a desired level of precision with systematic sampling are overestimated. Copyright © 2003 John Wiley & Sons, Ltd.
Funding Information
  • Ministerio de Ciencia y Tecnología, Spain (RNE2001-379/CLI)

This publication has 25 references indexed in Scilit: