Search for a command to run...
The application and validation of complex atmosphere-soil water transport models demands knowledge of the parameters that describe hydraulic properties over extensive areas. Such information is rarely available, but Pedo Transfer Functions (PTFs) provide a means of predicting these parameters from soil survey data. However, most PTFs have been derived and validated using information from soils of temperate regions and have not been tested for the soils of tropical areas, for which chemical, physical, and pedogenetic processes are different. The equations of Rawls and colleagues, for example, overestimate water content when applied to the soils of Brazilian Amazonia. In this paper, we have developed a PTF to predict Brooks-Corey parameters from texture using data from soils of Amazonia. Multiple linear regressions were fitted to estimate, from soil texture (% sand, silt, and clay), the bulk density and porosity and the water content at a range of matric potentials. Brooks-Corey parameters were then derived and correlated independently with soil texture, providing a straightforward method for deriving soil retention parameters from the percentage of clay and silt. The method was validated using an independent data set for which textural and water release data were available. The agreement between the observed and measured values was very significant, but the results showed that the differences between predictions and measurements also depended on bulk density. However, the Amazonian soil survey data, which may be used to extrapolate these results spatially, do not generally include bulk density, and for this reason they were not included in the regressions.