Variabilidade espacial da condutividade hidráulica do solo saturado e implicações no escoamento superficial direto em escala de bacia hidrográfica

Abstract

Soil hydrological properties, specifically the saturated soil hydraulic conductivity (Ksat) and initial soil moisture (ϴi) are key factors to understand the different processes of the hydrological cycle, as well as, the critical parameters to the application in hydrological models. Therefore, these properties are of great importance to support the decision-making process about water resources management at the watersheds scale. The present study had the aim to evaluate the spatial variability influence of saturated soil hydraulic conductivity and initial soil moisture on direct surface runoff hydrographs (DSRH) originated from extreme rainfall events, based on hydrological simulation by the Limburg Soil Erosion Model (LISEM) in the sang Ellert basin (BHSE) located in the municipality of Canguçu – RS, Brazil. Used for Gaussian sequential simulation (GSS) to analyze Ksat spatial variability and its uncertainty for different changes in scenarios, as well as in Ordinary Kriging. Later, these scenarios were incorporated into LISEM, together with information about 5 precipitation events, vegetation and soil characteristics, Manning roughness coefficient, digital elevation model and ϴi, to simulate the ESD hydrographs at BHSE. Calibration of LISEM was performed using the Manning roughness coefficient and ϴi (multiplicative factor on saturation moisture). The GSS was of great importance to describe the behavior of spatial variability of Ksat, as well as extremely useful to evaluate the propagation of the uncertainties to the peak river flows. The results showed that K0 is heterogeneous, being that their uncertainty ranged 718,59 - 866,57 mm h-1 to the percentiles of 5% and 95% respectively, possibly due to the soil type, use and soil occupation, textures, climate, as well as, sampling grid used. ESD is more sensitive to ϴi than Ksat and has been influenced by the 5 precipitation events. Peak flows are sensitive to ϴi and also to Manning's coefficient and were influenced by the 5 precipitation events. It was noticed that there was no difference between the calibrated parameters when considering the Ksat obtained by GSS and OK. It was found that peak flows did not change strongly with the incorporation of Ksat uncertainties for events analyzed using GSS and OK. As the intensity and volume of precipitation decreased, a greater spread of Ksat uncertainties was noticed at peak flows. LISEM was unable to capture (in relation to the multiplicative factor on saturation moisture) the Ksat variations between GSS and OK. Even using Ksat as field truth, it was realized that LISEM was able to represent peak flows at BHSE well, which are also strongly dependent on ϴi.