Contextualization: Incorporation of new techniques and technologies to know the characteristics and composition of soils used in agricultural production is booming. The free license products such as Sentinel-2 satellite images, allow correlating the edaphic properties with spectral indices.
Knowledge gap: In the agricultural sector, the indices calculated from satellite images have been centered on characterization of vegetation. To a lesser proportion, studies have been carried out with indexes used in mineral prospecting to correlate with the chemical characteristics of the soil.
Purpose: This study was developed to evaluate the correlation of ferric oxide, ferrous silicate, ferrous iron, and alteration indices with the content of chemical elements of importance in crop production.
Methodology: A crop field of 282 hectares in the municipality of Puerto López in the department of Meta was selected. The evaluated indices from a set of Sentinel-2 imagery were calculated. After indices obtention, 40 sites were selected to do a soil composite sampling and determine the following characteristics: pH in water (1:1 potentiometric), Organic Matter (Walkey and Black), P (Bray II), Al (KCl 1N), Ca, Mg, K y Na (Ammonium acetate), Fe (DTPA). With the obtained information, a multiple regression analysis using the stepwise method was performed; as independent variables, the raster layers of the indexes calculated were used, and the variable to predict were the properties determined in the laboratory.
Results and conclusions: The evaluated indices showed a higher correlation to 0.60 with Ca and Mg content. A correlation between 0.40 and 0.60 was found between the indices and organic carbon y Fe. The regression models calculated give adjustments of 0.54, 0.47, and 0.33 for Mg, Ca, and Fe, respectively. The findings of this research allow us to understand from Sentinel-2 images that it is possible to obtain relevant indexes for agricultural production in conditions of the Colombian Plateau.