Powerline's Hindrocket raises this interesting question:
It's always struck me that casualties resulting from natural disasters inspire less horror than those caused by violence. More people have been killed today by tidal waves in Asia than have been killed in the last year and a half of violence in Iraq. Yet it is unlikely that today's earthquake will stay in the news for more than a day or two. I'm not sure why this is, but, frankly, I share the tendency to pay much greater attention to political violence than to natural disasters. But that shouldn't make us indifferent to the tragedy suffered by so many today in South Asia.
Another variant is raised by the carnage caused by Israel's notoriously dangerous drivers: traffic deaths far exceed deaths from the intifada, but the question of road safety is off the Israeli radar screen.
Although there's some social psychology work on risk assessment, we probably don't need much research to tell us that intentionally caused deaths raise more hackles than accidentally caused ones. And of course:
As Holmes observed, even a dog knows the difference between being tripped over and being kicked.' Id." American National Fire Ins. Co. v. Schuss, 221 Conn. 768, 775-76, 607 A.2d 418 (1992).
There's also the question of consequences. Although sea walls and a warning system might reduce tsunami deaths, in generally these deaths are perceived as "acts of God" that human action can't prevent. Deaths caused by terrorism or war, on the other hand, are caused by intentional human action, and in additon to the moral questions involved, defense, deterrence, and retaliation all might reduce such deaths, and because we can do something to prevent them or reduce their numbers, we are no doubt more attentive to them.
Frankly, I'm not certain I've fully answered the question, but it's a start.