During the last 20 years the climate-related disasters have dominated the picture accounting for 91% of all 7255 recorded events, being the floods the most frequent type of disaster. Several decisions, such as the allocation of shelters and relief distribution, are made to minimize the aftermath impact on the population. In this chapter, we present a model to develop a reliable network based on a hierarchical preferences multi-criteria framework. This work aims to minimize the distance between the affected populations and the available shelters as well as their exposition to the risk due to damaged routes integrating the several stakeholders' preferences. The proposed solution is tested with the hurricane Stan case, which impacted the Mexican Republic southeast in 2005, affecting Quintana Roo, Yucatan, Oaxaca, and Veracruz. The case study is based solely on the state of Veracruz's situation, which considers 27 available distribution centers, 109 affected populations, and 1,379 available shelters. The problem is solved in GAMS commercial software, and the results showed that a reduction of non-used capacity of the opened temporary shelters up to 90.33% could be obtained when the integration of stakeholders' preferences and adequate decision-making tool.
