Prediction of soil classes in a complex landscape in Southern Brazil

Abstract

The objective of this work was to evaluate the use of covariate selection by expert knowledge on the performance of soil class predictive models in a complex landscape, in order to identify the best predictive model for digital soil mapping in the Southern region of Brazil. A total of 164 points were sampled in the field using the conditioned Latin hypercube, considering the covariates elevation, slope, and aspect. From the digital elevation model, environmental covariates were extracted, composing three sets, made up of: 21 covariates, covariates after the exclusion of the multicollinear ones, and covariates chosen by expert knowledge. Prediction was performed with the following models: decision tree, random forest, multiple logistic regression, and support vector machine. The accuracy of the models was evaluated by the kappa index (K), general accuracy (GA), and class accuracy. The prediction models were sensitive to the disproportionate sampling of soil classes. The best predicted map achieved a GA of 71% and K of 0.59. The use of the covariate set chosen by expert knowledge improves model performance in predicting soil classes in a complex landscape, and random forest is the best model for the spatial prediction of soil classes.

Fine-scale soil mapping with Earth Observation data: a multiple geographic level comparison

“…Digital soil mapping (DSM) has been explored in Brazil to produce updated and higher resolution maps of soil properties and soil taxonomic classes (Chagas et al, 2010;ten Caten et al, 2012;Samuel-Rosa et al, 2013;Giasson et al, 2015;Arruda et al, 2016;Vasques et al, 2016;Moura-Bueno et al, 2019). Different methods were also adapted for predicting soil taxonomical classes (Nolasco de Carvalho et al, 2015;Coelho et al, 2021), such as the probability mapping (Giasson et al, 2006), categorical mapping (Bazaglia Filho et al, 2013;Wolski et al, 2017;Moura-Bueno et al, 2019), Boolean logic (Rizzo et al, 2020), spatial disaggregation (Sarmento et al, 2017), among others. The DSM relies on the relationship of soil observations with environmental data employed in a model, which delivers a predicted value for a specific location usually coupled with an uncertainty estimate (McBratney et al, 2003;Lagacherie et al, 2006;Minasny and McBratney, 2016).…”