So, what happens when a giant tsunami inundates coastal lakes?

This blog post is by Philipp Kempf, PhD student at Ghent University, who writes about his recently published Sedimentary Geology paper on tsunami deposits in coastal lakes in south central Chile. Philipp is now on twitter too, tweeting as @TsunamiPhil.


Earlier this month, we, a group of geoscientists from Belgium, England, France, Switzerland and Chile, published a paper showing what happened in two coastal lakes, while a giant tsunami flowed into their basins.

The tsunami in question was caused by the strongest earthquake scientists have ever measured. It occurred on May 22nd in 1960 on the south central Chilean coast and had a magnitude of 9.5. The tsunami was strong enough to damage coastlines and kill people as far away as Hawaii and Japan, and of course wrecked the Chilean coast, too, with up to 20 m high tsunami waves. The obvious question in most scientific works is still the same since we all went to school as kids: “Why do we need to know that?” The question is justified, because what a giant tsunami does, since detailed measurements and direct observations exist for this and other giant tsunamis, like the one in 2004 in Sumatra that killed ~230.000 people or the one in 2011 in Japan that caused the Fukushima power plant to have multiple reactor meltdowns. But we need to know what a giant tsunami looks like in the geological record to be able to recognise tsunami deposits from times when there was nobody to write about it. Geoscientists know how often tsunamis occur in some places, because they have identified several tsunami deposits and have measured their age – most commonly with radiocarbon dating. If we were to make mistakes in identifying tsunami deposits, we would under- or overestimate how often tsunamis hit the coast.

A conceptual model of the processes that occurred during
the tsunami at both lakes.

We started by making a detailed underwater map of two coastal lakes and took seismic profiles, which work on lake sediment like large-scale ultrasound pictures of a pregnant woman’s womb. When we saw the data we decided to take sediment cores in strategic places. The cores contained about 50 years worth of normal lake sediment, which accumulated slowly since the earthquake in 1960. Below we found sediment from a violent inundation and we were able to date it and tie it this way to the great earthquake and tsunami in 1960. Looking at the sandy tsunami deposit, the inundation characteristics can be reconstructed.

The two main conclusions of our paper are:
1. It was possible to reconstruct how many waves of the tsunami inundated one of the lakes.
2. The 1960 tsunami inundated both lakes through their outflows to the Pacific, but also over the more than 1 km wide barrier separating the lakes from the ocean, which gives us an indication of how high the tsunami must have been.