Modelización de problemas ambientales en entornos urbanos utilizando sistemas de información geográfica y métodos multivariantes

In the present work a methodological proposal is presented for modeling urban environmental problems. It combines Geographic Information Systems and numerical techniques in order to generate probable impact scenarios. Additionally, a specific application is presented as an example oriented to creating susceptibility maps of landslides risk scenarios in an intermediate Colombian city. More specifically, the necessity of finding transversal methods to analyze the complex urban realities is presented. Among these realities, we pick up the possibility of landslides in Andes cities and a whole methodology based on a multifactorial cause analysis is introduced. This methodology takes into account past experiences to propose possible futures scenarios, relying on both Geographic Information Systems to store data and on Artificial Neural Networks to classify them.