Abstract We propose a methodology to assess tsunami risk in terms of human damage (HD: the quantification of people affected/killed by the tsunami) and economic loss (EL: the loss of the economic value of constructions damaged/destroyed by the tsunami). The methodology makes a combined use of detailed information on buildings, demography and reasonable tsunami inundation heights (IH). First, after defining the tsunami inundation area, we estimate the level of damage to each building within this area, which depends on the building vulnerability and on the flow depth (FD) affecting the building. To estimate the HD, by making use of census data, we infer the number of people present in each residential building of the inundated area. Then, based on FD and building damage, we evaluate the level of people exposure. Finally, we provide an estimate of fatalities. The EL is computed for two types of constructions (residential and trade/industrial) by using data on inundation and data from the real estate market. The methodology is applied to the town of Augusta (eastern Sicily coast, Italy), which, according to historical accounts, was affected by several tsunamis in the course of its history. The results are discussed in terms of the level of damage to buildings, the number of fatalities and the loss of economic value of the building stock for different tsunami sizes. This latter is prescribed through a simple bathtub approach: that is, by assuming uniform IHs with the maximum IH chosen to be consistent with historical data. Partial collapse occurs only above 6 m for residential buildings and above 8 m for trade/industrial buildings. Approximately 47% of the people living in Augusta are potentially affected by the tsunami inundation. For the highest selected IH (10 m), 70% of the people in the inundated area are effectively exposed and the fatalities are estimated at about 47%. The number of people involved increases linearly with the IH, while the number of fatalities increases following a quasi-quadratic law. Finally, the EL is also found to vary linearly with the IH, and the estimated loss in the case of the worst IH (10 m) is around 11% of the gross domestic product (GDP) of the entire province of Siracusa, to which Augusta belongs.

Preindustrial human societies living in ancient Central America apparently varied considerably in their vulnerabilities to the sudden massive stresses caused by explosive volcanic eruptions. The focus is on ancient complex societies and volcanic eruptions in the Zapotitán Valley during the past two millennia in what is now El Salvador, within a contextual and comparative framework of other eruptions and societies from other areas of Central America. Comparing eruptions and their societal effects indicates that there was a threshold of magnitude of these eruptions beyond which the society affected did not recover, and reoccupation was by a different ethnic group. An example is the TBJ (Tierra Blanca Joven) eruption of Ilopango where the original inhabitants never recovered to reoccupy the area. In contrast, the smaller Boquerón eruption devastated the eastern half of the Zapotitán Valley, but recovery was achieved by the indigenous society. The concept of “scalar vulnerabilities” is introduced, where physical factors such as the magnitude of eruptions are compared along with cultural factors such as variation in organizational complexity and institutionalized hostility. Simpler egalitarian societies in this sample were more resilient to sudden massive stresses than were the more complex societies. Simpler societies relied less on intensive agriculture, staple crops, fixed facilities, redistributive economies, and hierarchical institutions. Cultural factors such as chronic political hostilities can greatly increase the vulnerability of societies at any point on the simple-to-complex range. It is argued here that assessing a society's vulnerability must include the magnitude of stress, the complexity of the society (including the related factors of adaptation, demography, and the “built environment”), and the political landscape.