The pollution of the soil by heavy metals in the cities is a big problem of public health, for this reason the concentrations of heavy metals in the soils has been studied in many cities. However, there are few studies on the evaluation of soil capacity for the sorption of heavy metals with fast and low-cost techniques, different from the empirical techniques of chemical sorption isotherms.

The objectives of this study were: a) to develop two new indices of heavy metal sorption in urban soils of low cost and fast analysis; b) evaluate three heavy metal sorption indices by decision trees for classification; and c) prepare the map of the best heavy metal sorption index in the city of Morelia Mexico.

100 samples of surface urban soil were systematically taken. Subsequently, the content of organic matter, pH, cation exchange capacity, percentage of clays, structure, stoniness and bulk density were analyzed.

The Lehmann index was evaluated, and two new indexes were developed: the modified Lehmann index and the Bautista-Campuzano index. The validation of the three indices was carried out through the elaboration of decision trees for classification with the WEKA software.

The three indices showed congruence between the soil properties and the heavy metal sorption classes with a Kappa index> 0.7. However, the Bautista-Campuzano index showed the greatest theoretical congruence according to the decision tree for classification. This study reveals the importance of the mathematical and theoretical evaluation of the indexes by means of decision trees for classification.

It is considered that the Bautista-Campuzano index is the most efficient methodology for decision making related to the improvement of soils contaminated with heavy metals. The soils in the center of the city of Morelia showed the lowest values of the heavy metal sorption indexes, so these soils should be improved urgently.