Integrated processing methodology to identify the spatial variability of accumulated infiltration for agricultural soil

Abstract

The infiltration of water in the soil, and its variation in space, is essential to establish the irrigation schedule for crops and to evaluate the possible degrading effects on the soil. The objective was to develop an integrated processing methodology in Rstudio to identify the spatial variability of the accumulated infiltration, in two phases related to pea crops. Field sampling was carried out on a rectangular mesh with 48 points per moment, using double infiltrometer rings. The data were evaluated by means of geostatistical tools adjusted with programming code in Rstudio, defining the relationships between the magnitudes of the accumulated infiltration, for different test instants, without the need to make statistical adjustments to the normality of variables, discriminated over a period between 1 and 80 minutes. The results suggest the existence of spatial variability of the accumulated infiltration in the two evaluated phases, considering that most of the analyzed data were adjusted to multiple variance models, maintaining a degree of spatial dependence, and validating the effectiveness of the adjusted methodology developed and implemented. The spatial relationships were corroborated by means of contour maps, where the spatial variation of the accumulated infiltration between the two identified cultivation moments was observed. The reliability of the interpolation by the Ordinary Kriging method was verified by generating variance maps, establishing the degree of homogeneity of the interpolation. The variability of infiltration confirms the validity of the adjusted methodology implemented.