Performance of CA_Markov and DINAMICA EGO models to evaluate urban risk in Antofagasta and Mejillones, Chile

The rapid growth of cities worldwide is a phenomenon that has generated numerous debates about the effects it could have on the environment and society. An issue that has not been well addressed in the literature is the relationship between expected urban growth and risks to natural hazards. The relationship is an important aspect for planning because urban expansion is uncertain. In this sense, land change simulation models can be useful tools to address this uncertainty, because simulation models can produce growth scenarios, which allow anticipation of exposure to natural threats. This article compares two land change simulation models: the CA_Markov model in the Selva version of IDRISI and the DINAMICA EGO model. We apply both models to extrapolate areas of future urban gain in the coastal cities of Antofagasta and Mejillones, where there is high exposure to tsunamis, mudslides, and steep slopes. The models can extrapolate differently in terms of both quantity and spatial allocation of urban gain depending on the parameter settings. CA_Markov projected urban growth adjacent to existing urban patches, while best fits to Markov equation, and the urban form was achieved with DINAMICA EGO, according to our parameter settings. We conclude that if modelers understand the models’ behaviors, then applying these spatially explicit models to natural risks opens great prospects for urban planning and risk management, especially in countries highly exposed to dangerous natural events, such as in the case of Chile.