| Popis: |
Surface electrical resistivity tomography (ERT) is recognized as a powerful non-invasive soil survey and monitoring method. Relationships between electrical resistivity (ER) and soil water contents that are needed to infer the spatial distribution of soil moisture from the ERT results, are known to reflect soil properties. However, pedotransfer relationships that relate ER, water content, and basic soil properties other than clay content have not been developed. The objective of this work was to evaluate the improvement in the estimates of ER from soil water contents if soil basic properties are included in the pedotransfer equations. Soil — coarse-loamy, siliceous, mesic Typic (or Aquic) Hapludult — was sampled in 12 locations at ten depths across a 30 × 30-m experimental site. The resistivity was measured in disturbed samples in four-electrode cells. Water content, bulk density, texture, organic carbon, and pH were measured in these samples. The coefficient and the exponent in the power law-dependence of ER on volumetric water content were approximated with linear regressions that included any number of measured soil basic properties as inputs. A total of 1024 regressions were compared using the Akaike information criterion, and the best model was found that provided the best approximation with minimum number of regression terms. Six soil textural classes were encountered in the samples. When no pedotransfer relationships were included, the R2 of the log-log relationships between ER and water content increased from 0.45 at 10 cm to 0.91 at 190 cm. Splitting the dataset into topsoil and subsoil subsets resulted in favorable pedotransfer relationships with R2 about 0.85 both in topsoil and subsoil. Bulk density, clay content, sand content and pH appeared to be influential soil properties. Overall, accuracy of relationships between soil electrical resistivity and volumetric water content at ten depths in highly heterogeneous sandy soil could be substantially improved if pedotransfer relationships were established. Future work is needed on the applicability of data obtained on repacked samples to undisturbed soils and on the selection of input variables and evaluation methods for pedotransfer electrical resistivity pedotransfer functions. |
Full Text from ScienceDirect