Abstract
The risk is evaluated as multiplication among socioeconomic value exposed to risk events,
a frequency of risk events and vulnerability (socioeconomic-technological responses), which depends
on time and place. We calculate 1-km grid square statistics of physical exposures in about 380,000
grids over Japan for four cases: population in 2010, firms and workers in 2012, and travelers in 2013.
The exposed values and the frequencies of risky events are estimated from government statistics data
and NOAA Tsunami catalog data by using a statistical inference method. We assumed a generalized
Pareto distribution to infer the frequency of tsunami run-up events and estimated its parameters in
each 1-km grid square from run-up events close to the grid square. Assuming that the vulnerability is
set as 1 (no responses to tsunami run-up events), we estimated physical exposures defined as multi-
plication between an exposed value and a frequency in each 1-km grid square. We further applied our
proposed method to evaluating tsunami risks of 3,916 airports used in international air transportation
in 2014.
