Merger of Network Graph Indicators to Estimate Resilience in Latin American Cities

Abstract

In Latin America, cities are characterized by uncontrolled urban growth and locations in critical geographic areas. This factor affects an individual's ability to react efficiently to an action known as resilience to disasters caused by natural phenomena or catastrophes that involve human participation. This study proposes a method for detecting vulnerabilities in urban road systems in large, intermediate, and small cities in Latin America, so that resilience can be better supported. Depending on the structural characteristics of a city, measurement techniques can be used to combine the topological, geographic, and spatiotemporal indicators. Several measurements, such as mesh and connectivity, betweenness centrality, closeness centrality, robustness of networks with interruption attacks, location of dead ends, measurement of travel times between origin and destination, elevations and fusion of indicators with clusters on the same map, were applied using network graphs from OpenStreetMap (OSM) to estimate the resilience of road networks. The results are exposed to clusters of fused geospatial maps, which show that intersections and streets tend to be classified as vulnerable, with morphological structures of growth, resistance to degradation of the urban network, marginal urban sectors with high conglomeration, and mobility. The results highlight the importance of applying resilient practices in the region and generating urban management options to strengthen the response capacity of cities in Latin America.