Information on soil spatial variation is required for several soil science applications, such as land management, engineering and modelling the transport of nutrients and pollutants through the soil. Geostatistical methods are known to provide accurate maps of soil properties, provided that the sampling is appropriate. However, to implement these methods properly can be prohibitively expensive for many users. Traditional soil surveys used expert knowledge about how the soil was likely to vary based on parent material, and topography to map it as discontinuous classes. Can such knowledge be quantified and used to map the soil in a continuous way as embodied in the geostatistical approach?

McBratney & Pringle (1999) suggested the use of average variograms to guide sampling and avoid the need for a reconnaissance survey. Kerry & Oliver (2004) showed that this was possible if average variograms were related to parent material and topography. Furthermore, Kerry & Oliver (2002) showed that when no other variograms were available, average variograms could be used to krige sparse soil data at sites with similar parent material, topography or both. Kerry (2003) suggests that average variograms could be used more generally for kriging when topography and parent material specific databases of variograms for a variety of soil properties have been established. The validity of this concept depends on there being a plausible explanation, based on soil science theory, as to why the variogram of a soil property is likely to vary with parent material and topography. Analysis of soil and ancillary data for four sites on different parent materials with varied topography showed that variograms had longer ranges where the soil was more clayey and the landscape was fairly level. The variogram range describes the average size of patches of similar soil. Here we present an explanation and supporting evidence for differences in the variogram range with parent material and topography. The hydraulic conductivity and cohesiveness of soil materials could result in differences in variogram range or the patchiness in the distribution of top-soil properties associated with parent material. The amount of potential energy that influences the movement of water in given landscape positions could cause differences in variogram range with topographic unit.

References Kerry, R. 2003. Determining the effect of parent material and topography on the structure of spatial variation in soil properties for precision agriculture. Ph.D., University of Reading, Reading, England. Kerry, R. & Oliver, M.A. 2002. A comparison of kriged predictions using average variograms of soil properties and standardized average variograms of ancillary data. In: IAMG 2002, 8th Annual Conference of the International Association for Mathematical Geology (ed. U. Bayer, H. Burger and W. Skala), pp. 93-98. Kerry, R. & Oliver, M.A. 2004. Average variograms to guide soil sampling for land management. The International Journal of Applied Earth Observation and Geoinformation, 5, 307-325 McBratney, A.B. & Pringle, M.J. 1999. Estimating average and proportional variograms of soil properties and their potential use in precision agriculture. Precision Agriculture, 1, 125-152.