MODELLING SOIL WATER REPELLENCY IN AN ABANDONED AGRICULTURAL FIELD

Soil Water Repellency (SWR) is a natural property of soils with impacts on soil erosion, water infiltration, superficial and sub-surface hydrology, nutrients leaching and plant growth. Purpose: Study the spatial distribution and identify the most accurate interpolation method to estimate SWR in an abandonedagricultural field. Methodology: A plot with 21 m2 (07x03 m) was designed. Inside this plot SWR was measured in the field every 50 cm. In order to identify the most reliable map, we tested several interpolation methods, as Ordinary Kriging (KRG), Inverse Distance to a Weight (IDW) with the power of 1, 2, 3, 4 and 5, Radial Basis Function (RBF) (Inverse, Multiquadratic, Multilog, Multiquadratic, Natural Cubic Spline and Thin Plate, Spline) and, Local Polynomial, with the power of 1 and 2. Findings: The results show that SWR was very heterogeneous, even in small distances, showing that soil hydrological prop- erties can change very quickly in space. The spherical model was the best predictor of SWR and the most accurate interpolation method was the Multilog and the more biased the Natural Cubic Spline. Originality: The test of several interpolation methods in SWR spatial distribution were not explored in detail, and this study represents an advance in this field. Practical value: A better interpolation of SWR and other variables will help to have a better understanding of small scale processes in larger areas. Mapping with a better accuracy will improve models and contribute to a better prediction.