TESKE, Rodrigo e GIASSON, Elvio. Delimitação de áreas para plantio de eucalipto utilizando regressões logísticas. Cienc. Rural [online]. 2012, vol.42, n.7, pp. 1219-1222. ISSN 0103-8478. http://dx.doi.org/10.1590/S0103-84782012000700013.
Research uses logistic regressions to determine usable areas for planting eucalyptus trees in Rio Grande do Sul State, Brazil
Researchers from Universidade Federal do Rio Grande do Sul (UFRGS), Rio Grande do Sul, Brazil, used logistic regressions to select areas in a farm suitable for the planting of eucalyptus trees. The research was published in Ciência Rural journal, v.42, n.7, of july 2012.
The study generated predictive maps of areas suitable for planting eucalyptus using binary logistic regressions and geomorphometric variables. The researchers related land characteristics with river distances and the occurrence of permanent preservation areas. Thus, predictive maps of areas were obtained of planting of eucalyptus, which simulated areas planted after field evaluation.
In practice, the results can be used in initial analyzes of land acquisition and quantification of the area to be planted and in the exact location of plots for forestation with eucalyptus. Beyond, they assist in the evaluations of the real value of the land.
The research innovates by using knowledge about training algorithms in Agricultural Sciences, which, unlike other areas, are still poorly explored. To develop the model by the algorithms, it is necessary to supply information of relevant characteristics such as altitude, slope and area hydrography. Once established the mathematical relationships, the models can be implemented in GIS (Geographic Information Systems) for the generation of maps of areas potentially suitable for the planting of eucalyptus.