Sequential Gaussian simulation for mapping the spatial variability of saturated soil hydraulic conductivity at watershed scale

Abstract

The saturated soil hydraulic conductivity (K_sat) exhibits high spatial variability due to the various physical, chemical, and biological processes that act simultaneously with different intensities in soil formation. The use of geostatistics as a tool to study soil heterogeneity facilitates the understanding of the spatial variability of K_sat. This study aimed to simulate the spatial variability of K_sat and evaluate its uncertainties using sequential Gaussian simulation (SSG) in a tropical watershed located in southern Brazil. Soil sampling was conducted in an experimental watershed-scale grid with a sample spacing of 300 m, and K_sat was analyzed. Descriptive statistics were applied to assess the behavior of K_sat spatial variability, followed by geostatistical analysis, specifically SSG. Variogram parameters were defined, and SSG was used to generate 100 equiprobable random fields. The results showed that K_sat in the Santa Rita watershed (SRW) is heterogeneous, and uncertainties among the hundred fields ranged from 58.70 to 81.10 mm h-1 for the 5th and 95th percentiles, respectively, possibly influenced by soil type, land use, density, and texture. The criteria for validating SSG simulation were met and successfully described the spatial continuity of K_sat in the SRW. Thus, SSG proved to be an effective tool for understanding the magnitude and structure of K_sat spatial variability at the watershed scale, contributing to effective soil and water management in the SRW.