No other natural disaster, apart from the impact of an asteroid or comet from space, can cause such far-reaching devastation as the tsunami resulting from a large subsea earthquake.
The earthquake off the Sumatran coast released more energy than 10 million one-megaton bombs, as the ocean floor shot upwards by as much as 10 metres and displaced hundreds of cubic kilometres of water.
The resulting waves travelled across the open Indian Ocean at the speed of a jumbo jet, carrying enough energy to cause death and destruction 7,000 km away on the East African coast.
Although the earthquake itself could not have been predicted, a tsunami alarm system certainly could have saved lives further afield.
The authorities in Sri Lanka, India and the Maldives would have had two or three hours to organise at least a partial evacuation of coastal areas.
Sadly, the Indian Ocean has no tsunami-warning system like the one that has operated in the Pacific since 1968. But this quake was so big that, even without the infrastructure of a formal system, some seismologists were able to give unofficial warnings of the tsunami, though they had no way to get the word out to the coasts of south Asia.
Some scientists have been pressing for an Indian Ocean tsunami-warning system since the early 1990s but there was little sense of urgency in the region. In contrast to the Pacific, where the vast majority of the world's tsunami occur, the Indian Ocean has not suffered a serious tsunami since the 19th century.
In the aftermath of last week's events, Indian Ocean countries seem certain to agree this year to set up a regional warning network.
The basic technology has been tried and tested in the Pacific, where alerts have saved hundreds of lives. The system starts with a network of seismographs to detect the initial earthquake. The seismic data is used to make an immediate prediction of whether the location and severity of the quake makes a tsunami likely. If so, an alert is issued. Then the progress of the tsunami is monitored through sensors that detect changes in sea level, along coastlines and (via seabed pressure sensors and buoys) in the open ocean.
The basic network is not expensive because the Indian Ocean has far fewer zones of seismic activity than the Pacific, so less equipment is needed. A functional system could be installed for between $5 million and $10 million, according to United Nations experts and the US National Oceanic and Atmospheric Administration.
But individual countries would then have to invest further in the infrastructure required to interpret warnings for local conditions and relay them immediately to relevant areas.
Reliable evacuation plans are needed too. The system will do its job only if local inhabitants trust it so it is essential to avoid false alarms, such as the one issued last Thursday in south India, which caused widespread panic and disrupted the disaster relief effort.
Although an Indian Ocean tsunami-monitoring system would not have been needed during the 20th century, seismologists say this was an unusually quiet period.
Ominously, the geological record shows that severe earthquakes off west Sumatra, such as the one on December 26th, tend to occur in couples separated by just a few decades. The last pair of giant quakes occurred there in 1797 and 1833.
If the pattern repeats itself, the Indian Ocean will face another tsunami catastrophe before the middle of the century.
The disaster may also lead to the establishment of an Atlantic tsunami-monitoring network, although it is a less-active seismic zone than the Indian Ocean.