See the layers of sediments in the first of these 2 photos? They were deposited on a memorably bad day. These sediments come from a cave found on the edge of the island of Sumatra, and all of that sand and silt above the scouring layer was deposited on December 26, 2004; the date of the great Sumatran Earthquake and tsunami.
Those sediments were deposited by the tsunami, but they’re not the only tsunami remnants in this cave. This cave sits very near the shoreline and until recent uplift it was actually underwater, so it has taken multiple hits from tsunami waves over the last few thousand years. This cave was discovered a few years ago by scientists from Nanyang Technological University in Singapore and using results from this cave they have constrained the behavior of the Sumatran megathrust going back nearly 10,000 years.
The modern tsunami wave has an erosional scour at its base and the sediments just below it are 2900 years old – probably the date the cave was uplifted out of the water. The tsunami waves in 2004 entered the cave in pulses and dropped sand in layers on the surface. After trenching at this site, the scientists found 11 similar sand deposits in the sequence at this site. Each of them has thin layers of fine marine silt and clay in-between, showing that the sequence was a rapid deposition of sand followed by a long period of slow sedimentation.
Each of these 11 sand deposits represents a tsunami wave. The scientists carbon dated the layer at the bottom of their trench and found it to be 7400 years old, giving a recurrence interval of 450 years between 7400 and 2900 years ago.
That average would seem to be a statement of how often the fault breaks, but the scientists looking at the layers also found that average number to be almost meaningless – the fault doesn’t care about the average. They found that several thin layers of sand were packed close together, with as many as 3 smaller tsunamis within a 100 year period and as long as 2200 years of no sand deposits after one of the largest tsunami waves.
There are other geologic records around the Indian Ocean like this showing repeating tsunami waves, but none has as many waves recorded as this site and none of them are detailed enough to show the clustering. This site indicates that the Sumatran fault sometimes breaks in small earthquakes that trigger small tsunami waves and then occasionally breaks in a rupture like 2004 where it produces a major wave. Although there are only a couple examples in this cave, it is after big events like the 2004 quake where the fault is quiet for over a thousand years; so one possible interpretation is that the Sumatran fault may take centuries or millennia to produce major waves again.
This behavior is similar to the behavior observed at the Cascadia Megathrust off of Oregon, Washington, and British Columbia. Evidence from landslides and turbidites off the coast suggests that the fault there may break in a rupture that triggers a magnitude 8 earthquake and tsunami waves, but then occasionally the entire fault breaks, triggering a massive magnitude 9+ quake and much larger wave (https://tmblr.co/Zyv2Js29RKOmF).
There are often press reports saying a fault is “overdue” based on analyses of how often the fault moves on average, but these types of studies show that type of calculation just doesn’t express the way major faults move. The size of the quakes and size of the fault that breaks changes over time, and the exact future behavior is extremely difficult to predict.
The Sumatran fault is guaranteed to produce another large quake and tsunami. It could be thousands of years, or it could literally be less than 100 years. The best guess from this cave is that the huge 2004 rupture means it will be millennia before it goes again, but that's not the lesson of this cave. There are areas in the cave with 3 quakes in less than 100 years, and we don’t know for certain there won’t be another one soon.
Rather than measuring a recurrence interval, the lesson of this cave is that these megathrusts are very complicated and we’ve only seen the tiniest window of their history, so if the fault decided to release another quake and tsunami 20 years from now or 2000 years from now, humans in the area can’t be surprised. That lesson isn’t one for just this site either; it’s a lesson for faults around the world.
