A sampling strategy for quantitative soil survey includes two different samplings : (i) a reconnaissance sampling to characterize the variogram required to optimize the sampling scheme for (ii) the actual survey. For the actual survey a sampling
geostatistics. The variability was found to be the largest for the sand and silt contents, but occurred at shorter distances for the clay content. In order to apply puntual-kriging optimally, observations are required at least every 14,5 m for the depth data