A markov-monte carlo simulation model to support urban planning decisions: a case study for medellín, colombia

Abstract

The identification of properties and land destinations are key factors in urban planning decisions, especially in rapid-growing urbanized cities. This information is vital for cadaster matters, property taxes calculations, and therefore for the financial sustainability of a city. In this work we present a Markov-Monte Carlo simulation model to predict changes in land destinations. First, a Markov chain is established to identify the transition finite-state matrix of property destinations, and then a Monte Carlo simulation model is used to predict the changes. We present a case study for the city of Medellín, Colombia, using historical information from the cadaster office from 2004 to 2016. Results obtained allow identifying the urban areas with the larger number of changes. Moreover, these results provide support for urban planning decisions related to workforce sizing and visits sequences to the identified areas.