This is a tough one (Watching a ridge slide in slow motion ...):
The fissure was first spotted in October on Rattlesnake Ridge in south central Washington State, overlooking Interstate 82 and the Yakima River. Since then, a 20-acre chunk of mountainside — roughly four million cubic yards of rock, enough to fill 25 football stadiums to the top of the bleachers, eight stories up — has been sliding downhill. Geologists can measure its current speed — about two and a half inches a day — but they cannot say for certain when, or if, it might accelerate into a catastrophe. And they are powerless to stop it. ...
The usual course of events when nature comes unhinged is for researchers and responders to look back in time, trying to understand what happened, as they assess and address the damage done. Here it is precisely the opposite: They are looking forward in time at a disaster in waiting, in which all is still potential and nothing is certain.
The worst-case scenarios — considered unlikely but possible — have the slide breaking loose suddenly and roaring down toward the Yakima River, blocking the channel and flooding the valley, or burying the interstate that runs along the river, carrying 30,000 vehicles a day. Either would mean big trouble for a rich agricultural district where apples and hops are king, about two and a half hours’ drive from Seattle. ...
By giving scientists, officials and the people who live nearby so much time to stew over what could be happening, the slow-motion nature of the slide — at least so far — has already created tensions. About 75 people have been evacuated from nearby houses that may or may not ever be damaged. The local road at the foot of the hillside has been closed. Irrigation canal operators stand ready to open dams and take water from the river if it is blocked by a rock slide. Alternate highway routes have been mapped out, in case the interstate becomes impassable.
Current projections, based on dozens of motion sensors that have been installed on the ridge, suggest that the collapse is most likely to happen sometime in late February. ...
Mr. Emmons, the emergency management director, says he gets updates every day from state geologists and other experts about the odds and percentage predictions of the slide’s likely path. And then he tosses those odds out the window.
“I can’t rely on percentages, because in my world, 1 percent is still a chance,” he said. “I don’t feel comfortable banking on a percent number for public safety, or life safety. In my view, everything is possible.”
My guess is that if the analysis was done using the expected* benefits and costs then the net benefits of evacuation are less than zero. The economic analysis would suggest no evacuation. But the worst case scenario is likely very large and a risk averse emergency manager (see the last two paragraphs) will likely base the decision on that number.
*For example, the expected benefits would be estimated as E(B) = p1*B1 + p2*B2 + ... + pn*Bn; where p is the probability, B is the benefit of evacuation and slide scenarios go from 1 to n.
