Predição de solos moles no Brasil: elaboração de uma chave de classificação

Abstract

The knowledge of the soil characteristics is vital for land planning. Although soils have several characteristics that can lead to their classification as soft soils, so far in practice, this concept is limited to Geotechnics. Not with standing, critical infrastructure projects, especially transport, such as roads and railways, are great demanders of information regarding the occurrence of soft soils because of the roads' geographic scope or the need to identify locations with suitable soils to receive the road infrastructure and the vehicles. The objective of this study was to prepare a predictive map of the potential for the occurrence of soft soils in Brazil. The investigation started with an in-depth literature review seeking to identify the main characteristics described as soft soils, followed by their correlation with the physical, chemical, and mineralogical characteristics of the Brazilian Soil Classification System (SiBCS) soil classes. Then, a classification lookup table was elaborated to translate the potential for soft soils from the soil database. Thus, using the pedological maps of Brazil and the State of Goias,at scales of 1: 5.0000.000 and 1:250.000, respectively, the potential for soft soils for the areas was spatialized. Data validation included 1,470 samples along the North-South railroad within the state of Goias between Sao Simao and Quirinopolis. As a result, a map was generated containing two classes of Occurrence Potential of Soft Soils in Brazilian territory: Low and High Occurrence Potential. The low potential class encompassing soils with great physical stability, well-drained, well developed, and low primary mineral content, and absence of expansive minerals are represented by Oxisols, Argissolos, Nitosols, Planosols, and Plintosols. The high potential class comprises hydromorphic soils, with a high degreeof organic matter content, primary minerals of the 2:1 and expansive type, and sandy soils; Neosols, Organosols, Vertisols. Findings show that bridging Soil Sciences and Geotechnics allowed the construction of a model capable of providing predictive information on soft soils. The model is limited to the map scale; however, it covers a significant gap that existed until then in transport planning.