Quantitative determinations of soil attributes, via spectral sample data

Another application is to obtain quantity of certain attributes or quantitative analysis. The user who has spectral sample curves of his interest can enter the BraSpecS Spectral Processing system to upload his data and receive the results of clay, sand, organic carbon, cation exchange capacity (CTC), pH, base saturation (V%), calcium, magnesium, potassium, sum of bases, aluminum, saturation by aluminum, phosphorus among others. To do so, the user must initially fill out the model spreadsheet and upload the data. After a few minutes, the user will receive by e-mail the results of the soil analysis generated from the spectral data. These data are generated through attribute prediction models generated from the BSSL database (detailed analysis and explanation of the modeling are found in the article by Demattê et al., 2019). The models proposed by Rizzo et al. 2016 were used for soil color estimation. IMPORTANT: the literature demonstrates that the use of spectral soil data for attribute estimation is assertive and robust when dealing with particle size, organic carbon and CTC attributes. For the other attributes there is a great variability of results leading to even greater errors when used in practice.