T he science of volcanology began with Pliny the Younger's observations of the devastating 79 A.D. Vesuvius eruption. The May 1980 Mount St. Helens eruption and its aftermath was the first major eruption observed and monitored systematically with a variety of modern instruments, and it revealed much about eruptive processes, leading in turn to the development of new monitoring instruments. These lessons were applied a decade later in responding to the much larger eruption of Mount Pinatubo in the Philippines. The science of hurricanes has improved from early observations in the same way. Our current ability to make useful estimates of hurricane course and strength derives from satellite and in situ observations. Data gathered from many prior storms, combined with computer modeling developed continually from a huge amount of weather data, have strengthened our prediction capacity. These are but two examples of how past observations, followed by detailed data obtained with modern instruments, have led to improved understanding of natural disasters and how to respond to them. The Sumatra-Andaman earthquake of 26 December 2004, with its consequent tsunami, was the first great earthquake to be observed with modern instruments. These include a global seismic network equipped with seismographs capable of recording over a broad frequency range (which was partly set up for observing and understanding such earthquakes), satellites that can record changes in ocean height, and a worldwide Global Positioning System (GPS) network capable of recording subtle movements of Earth's crust. Great earthquakes have moment magnitudes of about 9 or greater, release an enormous amount of energy, and are usually caused by the rupture of a convergent plate boundary (where one plate is subducting beneath another). As was the case here, they can produce devastating tsunamis. ![Figure][1] CREDIT: AMMON ET AL. Four papers and a Viewpoint in this issue, and one Report published online on Science Express, describe the seismic, satellite, and GPS observations of this earthquake and some of its effects. As shown in these papers, such large earthquakes have important differences from smaller earthquakes. One is their huge rupture length, which is many times the width across the rupture. The Sumatra-Andaman earthquake ruptured about 1300 km of the plate boundary. The rupture itself took about 1 hour, moving mostly from south to north, and the GPS data indicate some additional deformation after that. It rang the earth much like a bell (it is still ringing), and it produced measurable deformations in Earth's crust—as recorded in GPS data—over nearly an entire hemisphere. This protracted and uneven release of energy complicates the determination of magnitude. These data, as well as observations of the geometry of the generated tsunami, can be inverted to show that the rupture progressed rapidly at first, producing displacements of up to 20 m, then more slowly farther north. Seismic waves from the Sumatra-Andaman earthquake, like some others, also triggered other earthquakes in distant volcanic areas; this time, however, possibly revealing the mechanism of that triggering. The earthquake also produced numerous aftershocks and was followed by another large temblor. These and other data provide a new basis for understanding these largest, rarest, and most destructive seismic events. [1]: pending:yes