Abstract

GOMEZ, Laura Fabiana. Simulation of future scenarios of land use changes using Artificial Neural Networks. Application in Curuzú Cuatiá city, Corrientes, Argentina. Estudios Socioterritoriales [online]. 2020, vol.28, pp.54-54. ISSN 1853-4392.  http://dx.doi.org/10.37838/unicen/est.28-054.

Simulations of future land use scenarios allow obtaining results that can become valuable information for land planners while decreasing the degree of uncertainty. The aim of this contribution is to simulate a linear path land use scenario with a linear trajectory for the year 2030 in the city of Curuzú Cuatiá, Corrientes. Topographic factors, distances and flood areas defined by current regulations were chosen as explanatory variables of the changes, a model based on artificial neural networks, included in the Land Change Modeler -LCM- in the Idrisi Selva software was used, in addition to two maps of land uses of five categories elaborated for two moments of the past. The results obtained show the maps of transition potentials towards the Consolidated Urban Area with accuracies greater than 72%; simulated scenarios for the year 2030 (hard and soft prediction), and the surfaces of each class at each moment were determined, finding the increase only of the Urban Consolidated class in all of them. The study was completed with the validation of the model by class, using the global reliability and the Kappa index.

Keywords : Simulation; Future Scenarios; Land use change; Artificial Neural Network.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )