Just how big was 2004's Sumatra earthquake?
One of the best-known solutions for measuring an earthquake's magnitude is the "Richter" scale. In 1935, Charles Richter was looking for a simple way to categorize earthquakes by the energy they release, rather than subjective descriptions of their impacts on human settlements. He decided on a scale that would emulate the stellar magnitude scale used by astronomers to measure the strength of stars. Like that scale, Richter's magnitude scale measures an earthquake by one criterion alone: the energy it sends out to recording instruments around the globe.
Like stars, earthquakes emit energy at many different frequencies, analogous to stars' emission of different kinds of electromagnetic energy. However, when seismologists started studying the largest earthquakes on the planet, they discovered an important failing of Richter's original magnitude scale. Because the energy at a given frequency is related to the size of the fault that produced the earthquake, the scale tends to 'saturate' when an earthquake reaches a certain size.
That's why a new magnitude scale was developed in the 1980's; the moment-magnitude scale relies on the lowest seismic frequencies, which are most sensitive to the scale of the earthquake-producing fault.
The different magnitude scales used by seismologists are why we often hear different values reported about the same earthquake. While the initial estimates of last year's Indonesian earthquake's magnitude reported 8.0, the moment-magnitude estimate, determined only after several hours of seismic recording, came to a 9.0. That means the earthquake was a whopping ten times larger than the original magnitude indicated.