One pertinent question in the wake of the earthquake near Aceh (亚齐省) and the tsunami (海啸) it generated is how much notice of an approaching wave can be given to vulnerable people without the risk of crying "wolf" too often. Earthquakes themselves are unpredictable, and likely to remain so. But detecting them when they happen is a routine technology. That was not the problem in this case, which was observed by monitoring stations all over the world. Unfortunately for the forecasters, although any powerful submarine earthquake brings the risk of a erous tsunami, not all such earthquakes actually result in a big wave, and false alarms cost money and breed cynicism. On top of that, most "tsunamigenic" earthquakes, which are caused when the processes of plate tectonics force heavy, oceanic crystal rock below lighter, continental rock to create a deep trench at the bottom of the sea, occur in the Pacific, which is almost surrounded by such trenches. In the Indian Ocean, deep trenches are confined to the southern coast of Indonesia, and tsunamis are rare. Since most of the countries affected by this tsunami are poor, or middle-income at best, and monitoring costs money, this might suggest that a fatalistic approach to the question is reasonable. But American and Japanese experience suggests that effective monitoring need not be that expensive. These two countries have networks of seabed pressure-detectors that can monitor tsunamis and indicate whether and where evacuation is necessary data they share with their Pacific neighbors. A system of seven detectors, run from Hawaii, cost about $18m to develop, and the experience gained doing so means a similar system might now be had for as little as $ 2m. So, to the sound of stable doors being bolted firmly shut, politicians in South-East Asia and Australia are proposing one for the Indian Ocean. Even if you have an effective detection system, though, it is useless if you cannot evacuate a threatened area. Here, speed is of the essence. Computer modelling can help show which areas are likely to be safest, but common sense is often the best guide-run like the wind, away from the sea. Evacuation warnings, too, should be easy to give as long as people are awake. Radios are ubiquitous, even in most poor places. It is just a matter of having systems in place to tell the radio stations to tell people to run. The problem was that no one did. To the countries in South-East Asia, building a tsunami monitoring system _________.