Abstract: | The aim of this study is to evaluate two different approaches [the layered two-dimensional (2-D) method and a full 3-D approach] to describe the 3-D spatial variability of soil NO3--N in the top 1 m of an agricultural field of about 1 ha. NO3--N concentrations were determined in layers of 5 cm to a depth of 1 m at 26 locations. These samples were complemented with less intensive sampling at only three depths (0-5, 30-35 and 60-65 cm) at 26 other locations and with topsoil (0-5 cm) samples at another 75 locations. Variogram analysis showed a strong anisotropy when the horizontal dimension was compared to the vertical. A nested model of four structures was used to represent a 3-D variogram. This allowed us to predict the average NO3--N concentration for blocks of 5ǹ m (horizontally) by 0.05 m (vertically). A cross-validation showed a clear improvement in using the full 3-D interpolation instead of the layered 2-D approach. The full 3-D interpolation seems to be especially useful in situations where knowledge of the 3-D distribution of soil properties is essential to evaluate soil management practices in respect to environmental considerations, like N fertilisation and the subsequent leaching of N to the ground water, or N2O emissions. |