Renato Paiva de Lima

soilphysics: An R package to determine soil preconsolidation pressure

Preconsolidation pressure is a parameter obtained from the soil compression curve and has been used as an indicator of load-bearing capacity of soil, as well as to characterize the impacts suffered by the use of machines. Despite its importance in soil physics, there is a few software or computational routines to support its determination. In this paper we present a computational package in R language, the package soilphysics, which contains implementations of the main methods for determining preconsolidation pressure, such as the method of Casagrande, Pacheco Silva, regression methods and the method of the virgin compression line intercept. There is still a consensus that Casagrande is the standard method, although the method of Pacheco Silva has shown similar values. The method of the virgin compression line intercept can be used when trying to be more conservative on the value (smaller) of preconsolidation pressure. Furthermore, Casagrande could be replaced by a regression method when the compression curve is obtained from saturated soils. The theory behind each method is presented and the algorithms are thoroughly described. We also give some support on how to use the R functions. Examples are used to illustrate the capabilities of the package, and the results are briefly discussed. The latter were validated using a recently published VBA. With soilphysics, the user has all the graphical and statistical power of R to determine preconsolidation pressure using different methods. The package is distribution free (under the GPL-2|3) and is currently available from the Comprehensive R Archive Network (http://CRAN.R-project.org/package=soilphysics). The R platform and all the package dependencies are similarly available from CRAN.

Spatial variability and estimate of critical sites for bulk density in an oxisol

The soil physical condition is complex and must be analyzed in terms of a set of specific indicators and their inter-relations. This work aimed to evaluate the spatial variability of bulk density and related effects from other physical attributes in an experimental area of Oxisol in order to detect and estimate sites with critical values for root development. Linear models were fitted to the spatial bulk density in layers 0-0.2 and 0.2-0.4 m, considering the Gaussian process. Models were adjusted with and without a principal component, and models containing the covariate density in the previous layer to depth of 0.2-0.4 m. Performed kriging, probability maps to obtain bulk density above 1.4 kg dm-3 were constructed based on 1000 simulations from predictive distribution. There were sites with critical values of soil density. There were no related effects of moisture, sand, silt and clay in the bulk density levels, however, it was no influence from upper layer in the layer density levels on 0.2-0.4 m. There was evidence of sites over 80% probability of occurrence of critical levels of bulk density, in the two layers.