Tsunami, USA
I've never lived anywhere that didn't have natural disasters: east coast hurricanes, pacific rim typhoons and earthquakes (e.g. '95 Kobe, which woke me from a sound sleep), plains tornados, not to mention the ever-present tectonic uncertainty -- the"BIG ONE" which was theoretically due in the next century, or overdue, everywhere I've lived for the last fifteen years.
Now, I live in Hilo, Hawai'i --"the tsunami capital of the United States" -- one of the few American cities to have been devastated by tsunami multiple times over the last century. So it's no wonder that one of our few local museums is the Pacific Tsunami Museum, dedicated to the proposition that tsunami deaths can be eliminated with proper preparation and warning. Obviously, they've got some work to do, but to be fair they've only taken on the Pacific as their project. The museum's website has been rather hard to access lately, presumably because it is tsunami.org and they've gotten a lot of traffic related to the current crisis. If you really want to root around their site, I highly recommend coming back to it in a few weeks. Their FAQ page is quite substantial, though, if you need a basic primer on the science and signs.
Hilo has been hit by several major tsunami in recent history -- 1946 and 1960 were the largest, but 1975 was actually more destructive locally than the 1960 -- and so doesn't have quite the wealth of early architecture or downtown vigor you would expect. We do have clearly marked evacuation zones and routes, and the monthly tests of tsunami early-warning sirens sound just like the monthly tests of the tornado warning sirens in Iowa. Our home is a couple of hundred feet up from sea level, in about as much danger from tsunami as from lava..... which is to say that if I'm in danger from either of these things, so are a lot of other people.
I am sure that Glenn Reynolds is wrong about the value of tsunami warning systems, even aside from the fact that the investment necessary with 21st century technology is much less than the return in saved lives. As human population densities increase (and there's little evidence that they will be decreasing soon, though the rate of increase may be slowing) and as coastline has always been desirable non-agricultural living space, the potential loss from tsunami is increasing. Therefore the relative cost of precautionary spending will not only decrease, but it will be cheaper to maintain a system that we build now than to build a new system when it seems more necessary in the future.