A few months ago, I guaranteed the readership of the Lounge that none of them would be killed by a meteorite impact. In laying out the estimates that allowed me to do this, I took an equal area map and bombarded it with one hundred 400m projectiles. Objects of this size hit the Earth about once every 100,000 years, and are locally devastating but globally insignificant, so this seemed like a good way to look at where an impactor “big enough to wipe out LA” was actually likely to land.
Of these 100 impactors, 71 landed in the ocean. 19 of these were within 1000 km of the coastling of an inhabited continent (e.g. not Antarctica), while the others were far out in the ocean basins.
For impacts more than 1000 km offshore, the impact effect calculator of Marcus, Melosh, and Collins suggests that the main effect would be a tsunami. The tsunami details are not in their linked paper, and the amplitudes vary significantly, but the maximum amplitude at 1000 km from the impact area is about 4 meters or smaller. This is broadly similar to that of a magnitude 9 earthquake such as those that struck Japan this year and Sumatra (and the Bay of Bengal) four years ago. The tsunami takes about 1.8 hours to travel 1000 km, so warning times would depend greatly on detecting the impactor in space and seeing the fireball with antiproliferation satellites (this impactor is equivalent to a 3000 megaton bomb, so the fireball would be far larger than that of a nuclear weapon). The seismic signal of a hit to the deep ocean would actually be fairly minor, as most of the energy would be absorbed by the water.
Of course, the main difference is timing. Half of the impacts in this simulation were in the deep ocean, so with an impact repeat rate of 1 every 100,000 years, we would expect one deep ocean impact every 200,000 years. In contrast, a magnitude 9 earthquake strikes about once every 25 years. So over a million year period, we would expect 40,000 tsunamis from earthquakes, and five from deep ocean impacts.