Quantifying the humanitarian and economic impact of earthquakes on South American capital cities (Invited)

By 2000, an estimated 80% of South America’s population lived in urban areas (Veblen et al., The Physical Geography of South America, Oxford University Press, 2007). A significant fraction of those urban dwellers resides in the capital cities which are major economic centers and act as magnets for rural poor and refugees. This population concentration includes many residents living in extreme poverty in substandard and informal housing, often on the margins of these capital cities and sometimes on steep slopes, greatly compounding the vulnerability to natural hazards. We are analyzing the humanitarian and economic risk for six of the seismically most-at-risk South American capitals along the northern and western plate boundaries of South America: Caracas, Venezuela; Bogotá, Colombia; Quito, Ecuador; Lima, Perú; La Paz, Bolivia; and Santiago, Chile. Impacts are provided in the form of expected losses for a specific “likely” scenario earthquake and in a probabilistic format using exceedance probability curves (probability of exceeding a given loss in different return periods). Impacts to be quantified include: total economic losses, potential fatalities, potential serious injuries, and the number of displaced households. Probabilistic seismic hazard was developed in collaboration with numerous South American experts and includes subduction interface, intraslab, background crustal and, where available, active fault sources. A significant challenge for this study is to accurately account for the exposure and vulnerability of populations living in the informal, shanty areas. Combining analysis of aerial imagery and on-the-ground reconnaissance, we define between 20-30 “inventory districts” of relatively uniform construction styles within each capital. Statistical distributions of the different construction types and their characteristics (height, occupancy, year built, average value) are estimated for each district. In addition, working with local graduate students, we are also developing demographic information including the population, occupancy (day and night) of the various structures, as well as the total number of buildings for each inventory district. Estimating damage and losses due to triggered landslides presents another major research challenge for the project. Analysis of global earthquake fatalities in the last 40 years indicates that earthquake-triggered landslides are responsible, on average, for about 5% of the reported fatalities (Marano et al., 2009, Natural Hazards Research); however in South America in the last 50 years our own analysis showed that 29% of the approximately 79,000 earthquake fatalities have been caused by landslides. Our analyses are designed to provide objective and comparative data on the humanitarian impacts for the various cities. The project’s goal is to focus attention on this increasing problem throughout South America and much of the developing world. In the second phase of the project we plan to develop partnerships with government agencies, NGOs, and other local stakeholders to help implement mitigation solutions.