Portal de Periódicos da CAPES

Using the EM38DD Soil Sensor to Delineate Clay Lenses in a Sandy Forest Soil

2007; Wiley; Volume: 71; Issue: 4 Linguagem: Inglês

10.2136/sssaj2006.0323

1435-0661

Liesbet Cockx, Marc Van Meirvenne, Bruno De Vos,

Soil and Unsaturated Flow

The objective of this study was to locate clay lenses in a sandy forest soil. The study site was a Scots pine ( Pinus sylvestris L.) plantation in the Campine region of Belgium, which has been selected as a European Union‐funded Level II Research Site for soil monitoring. Typically, the soils in the area have homogeneous sandy soil profiles, but clay lenses occur locally within a depth of 2 m of the soil surface. Locating these clay lenses is necessary because they can have a substantial impact on soil processes. Therefore, we used the EM38DD soil sensor to measure the soil apparent electrical conductivity (ECa) simultaneously in two orientations. Apparent electrical conductivity maps were generated and it was found that the variation in ECa was mainly driven by the spatial variability of soil texture across the study site. The ratio of the two orientations (profile ratio or PR) clearly revealed a circular pattern with decreased PR values (<1), which was identified as a clay lens. To delineate the extent of this lens, two numerical methods were used: (i) a fuzzy‐ k ‐ means classification of the PR map focusing on the lowest centroid class, and (ii) a probability approach through indicator kriging. Using a validation image obtained from directed auguring, cell‐by‐cell comparisons were made for these two methods complemented with spatial accuracy measures. Although both methods tended to underestimate the spatial extent of the clay lens, the indicator kriging method was the most accurate, with an overall accuracy of 0.838, a proportion of error due to locational errors (κ‐loc) of 0.864, and an average similarity of 0.841.