Funções de pedotransferência na caracterização da distribuição espacial da condutividade hidráulica do solo saturado

Abstract

Saturated soil hydraulic conductivity (Ksat) is an important soil attribute for water resources management and soil management and conservation, because it governs soil water flow, infiltration rate, runoff, solute transport, among others environmental processes. However, it has been reported in the literature that Ksat has a high spatial variability for watershed scale measurements or even on small experimental plots areas. Thus, the identification and quantification of Ksat spatial variability has been the target of studies related to soil water dynamics. However, determining Ksat's spatial variability requires a large dataset. Due this, the application of mathematical models to estimate Ksat using low-cost independents variables as topographic attributes or another soils hydro physics attributes (SHPA) has become increasingly common. The objective of this study was to develop and evaluate pedotransfer functions (PTFs) to estimate the saturated soil hydraulic conductivity in an experimental sampling grid, containing 80 points with irregular spatial distribution, established in the Santa Rita watershed (Pelotas, Rio Grande do Sul), as well their potential to characterize spatial variability structure of hydraulic conductivity using geostatistics tools. At each point of the experimental sample grid were determined the SHPA and topographic, as well as information on the land use. The PTFs were developed using multiple linear regression technique and validated through the application of statistical performance parameters and by a ranking proposed in the literature. Applying geostatistical tools (variographic modeling and interpolation by ordinary kriging) where performed the characterization and mapping of the spatial variability distribution of Ksat values observed in the field and estimated by PTFs developed for the soils of the Santa Rita watershed (SRW). The PTFs found in another studies and tested in this work were unable to estimate Ksat for SRW soils. The PTFs that best described the spatial distribution of Ksat were developed from soil bulk density, clay content, slope and macroporosity data and from soil bulk density, clay content, sand content and elevation data. Topographic attributes (slope, aspect and elevation) were present as predictive attributes among the seven best performing PTFs. Land use data also proved to be an adequate predictor of Ksat estimation. So, it was possible to conclude that the PTFs developed to estimate Ksat showed potential to capture Ksat spatial dependence structure in the watershed area. The Ordinary Kriging indicated that the spatialization of the absolute errors of the estimated Ksat values by the soil bulk density, clay content, macroporosity and slope presented the smallest errors and the most similar spatial distribution pattern in relation to the observed Ksat data.