How a volcano can be the trigger
Nobody had any clue. There was certainly no warning. It's part of the picture that now points to a large underwater landslide being the cause of Saturday's devastating tsunami in the Sunda Strait.
Of course everyone in the region will have been aware of Anak Krakatau, the volcano that emerged in the sea channel just less than 100 years ago. But its rumblings and eruptions have been described by local experts as relatively low-scale and semi-continuous.
In other words, it's been part of the background.
And yet it is well known that volcanoes have the capacity to generate big waves. The mechanism as ever is the displacement of a large volume of water.
Except, unlike in a classic earthquake-driven tsunami in which the seafloor will thrust up or down, it seems an eruption event set in motion some kind of slide.
It is not clear at this stage whether part of the flank of the volcano has collapsed with material entering the sea and pushing water ahead of it, or if movement on the flank has triggered a rapid slump in sediment under the water surface.
The latter at this stage appears to be the emerging consensus, but the effect is the same - the water column is disturbed and waves propagate outwards. Tide gauges in the Sunda Strait indicate high water around half an hour after Anak Krakatau's most recent eruptive activity at roughly 21:00 local time (14:00 GMT) on Saturday evening.
Prof Dan Parsons from Hull University, UK, told BBC News: "The sides of volcanoes, the flanks, are notoriously unstable and it looks like a landslip movement into, or below, the sea has resulted in the generation of a significant tsunami.
"[The original Krakatoa volcano] exploded and destroyed itself in 1883 and since then has been building again slowly. As volcanoes build, their sides can become unstable and collapse even without any volcanic activity. As the slide displaces water it generates a large wave - the same way as if you enter a bath from one side too quickly."
Using tide gauge data, scientists have already begun to model the tsunami. Early results suggest something happened on the southern side of Anak Krakatau. The area will become the subject of intense scrutiny in the days and weeks ahead.
Landslide- or rockfall-driven tsunamis can be very big indeed. In the geological record, they have been responsible for gargantuan events.
Just recently in Greenland in 2017, a 100m (330ft) wave was produced by a rockslide entering a fjord in the west of the country; and there is still some suspicion that September's damaging tsunami that affected Sulawesi Island in Indonesia was, in part at least, strengthened by the mass movement of sediment, either entering the water from shore or slipping down underwater slopes in Palu Bay.
In the coastal towns that line the Sunda Strait, waves as high as 5m have been reported by eyewitnesses. But although big, these waves can also dissipate quite quickly as they move away from their source.
One of the most distressing things on occasions like this is to see video of people going about their lives completely unaware of what is about to hit them.
If there had been a major earthquake tremor associated with the eruption on Anak Krakatau, this might have been enough to prompt many locals to take evasive action.
But although there was seismicity reported by sensitive instruments, it wasn't large enough to change people's behaviour. And you really have to rely on yourself in an evacuation in cases like this because the distance from the source of the tsunami is so short.
Content is not available
"Tsunami warning buoys are positioned to warn of tsunamis originated by earthquakes at underwater tectonic plate boundaries. Even if there had been such a buoy right next to Anak Krakatau, this is so close to the affected shorelines that warning times would have been minimal given the high speeds at which tsunami waves travel," observed Prof Dave Rothery from the UK's Open University.
What an event like this one (and the one at Palu City which also caught the population unawares) teaches us is that there needs to be far more investigation into the hazards that exist away from the expected dangers in the region.
Enormous research effort has gone into understanding what are called subduction earthquakes and tsunamis, such as the 2004 disaster which originated on the Sunda Trench, where one tectonic plate dives under another. The science now needs to encompass more of the wider problems in the region.
This recognition was voiced strongly at this month's American Geophysical Union Meeting - the world's largest annual gathering of Earth scientists.
"Focus is always where the light is," Prof Hermann Fritz, from the Georgia Institute of Technology in the US, told the meeting.
"The focus has been on Sumatra and Java - on the big subduction trenches. The warning centres have also been focussing on that - because we've had big events such as Japan (2011), Chile in 2010 and Sumatra in 2004. These are all classic subduction zone events, so everything has been geared towards that - the science from the scientists and also the warnings from warning centres."