A soil sampling program for the Netherlands

Soil data users in The Netherlands were inventoried for current and future data needs. Prioritized data needs were used to design the Netherlands Soil Sampling Program (NSSP) as a framework containing 3 groups of related projects: map upgrading, map updating and upgrading of pedotransfer functions. In each one group, the sampling design, performance criteria and optimal sample size were defined. This paper focuses on the upgrading of the existing soil map of The Netherlands at scale 1:50,000, and extensively treats the user inventory and the sampling strategy. The sampling design, performance criteria of the sampling and associated optimal sample size were obtained by statistical analysis of soil data available before the sampling. The Phosphate Sorption Capacity (PSC) was chosen as target variable to optimize sarnpling, because it dominated total cost per sample. A prior analysis of a performance criterion related to the sampling error of PSC resulted in a cost saving of 13% relative to total cost determined earlier by expert judgment. A posterior analysis showed that the set quality criterion was reached or better in 6 out of 7 cases. The NSSP resulted in a data base with soil data from 2524 sample points selected by stratified random sampling, and a collection of 5764 aliquots taken at these points. The NSSP has been showing its usage potential for various kinds of environmental studies and could be a sound future basis for a national scale monitoring program. (c) 2007 Elsevier B.V. All rights reserved.