Deep sea tephra layers – Sumatra

In 2008 I began working with Dr. Adam Kent to figure out the provenance and correlation of some deep-sea tephra layer that were dredged from within, and adjacent to, the Sunda trench off the coast of Sumatra, Indonesia. Fellow PhD student Jason (Jay) Patton, and his adviser Dr. Chris Goldfinger, were working on correlating deep-sea earthquake deposits (turbidites) in order to determine the recurrence interval of major subduction zone earthquakes in the region, the most recent being the devastating 2004 quake and associated tsunami. Jay, Chris, and crew had found some volcanic ashes within the sediment cores and Adam and I got involved to help the correlations through geochemical analyses of tephra glass shards using the electron microprobe and laser ablation ICP-MS at Oregon State University.

Although I wasn’t involved in the research crew that collected the sediment cores, I did make a number of excursions across campus to the refrigerated core lab to collect samples.
core lab

The first goal was to determine if any of the tephra layers could be correlated geochemically. This involved sample collection, cleaning, and mounting in epoxy prior to analysis with the electron microprobe and ICP-MS.
sumatra deep sea tephra
Core localities, stratigraphy, and backscatter images of individual glass shards.

Major element analysis of glass shards was able to clearly distinguish three separate compositional groups.

wt.% SiO2 vs. wt.% K2O

Once these groups were established, it was necessary to look at the trace elements to look for correlations within the three groups.traces_sumatra

Trace element analysis revealed 6 distinct tephra horizons within the cores that we interpreted to represent 6 distinct volcanic eruptions.

The next question was to establish the age of the tephra layers. Jay took the lead on this and separated foraminifera shells within the hemipelagic sediment above and below the tephra layers. These shells were then analyzed for <sup>14</sup>C-dating.

Four of the eruptions were determined to be less than 15 thousand years old, supporting the idea that Sumatran volcanoes are young and active.

The final step was to figure out where the eruptions came from. This proved to be a dead end for the time being. Very few of Sumatra’s volcanoes have been properly explored and analyzed, so there is very little data for which to compare. The thick vegetation and high erosion rates of the island don’t help much either.

Instead of matching each eruption to its source, we determined the minimum volume necessary to deposit the tephras, assuming the nearest on-land volcano as a potential source.It turns out, these eruptions must have been quite large (similar to the 1980 eruption of Mt. St. Helens) to produce these tephras.


For the complete study, see Salisbury et al., 2012.

Leave a Reply