Turning point: David Shelly

Journal name:
Date published:
Published online

Earthquake researcher's open-minded approach brings high-level recognition.


David Shelly is a seismologist with the US Geological Survey (USGS) in Menlo Park, California. In December, he will receive the latest in a string of high-profile awards: the Macelwane Medal, presented at the American Geophysical Union conference in San Francisco, California.

You had a broad-based education at a liberal-arts college. How did that prepare you for earthquake research?

I studied maths and physics at Whitman College in Walla Walla, Washington, but knew that I wanted to do something more applied as a graduate student. I didn't have the background for a geology programme, so I studied geophysics at Stanford University in California. My broad liberal-arts background helped me to learn how to communicate ideas concisely: an important part of the scientific process.

Describe your PhD research.

I went to Tokyo to study subduction-zone tremors in 2005, as part of the East Asia and Pacific Summer Institutes funded by the US National Science Foundation and the Japan Society for the Promotion of Science. A typical earthquake is one sudden rupture of a fault, but low-frequency subduction-zone tremors are weak vibrations resulting from slow slip between tectonic-plate boundaries. They start and end gradually, yet last much longer than a normal earthquake. Such tremors are a challenge to work with: it is hard to distinguish the seismic-wave signals from the background noise.

Your work had a big impact on the field. Why?

After low-frequency events were discovered in 2002, it was unclear whether they were earthquake-like or more like volcanic tremors, with fluids moving below ground. I used a technique to identify low-frequency tremors without knowing the exact onset time of the wave phases, which overcame the signal-to-noise difficulties. My team's results suggested that low-frequency subduction-zone tremors can be generated by similar processes to, and on the same faults as, larger earthquakes (D. R. Shelly et al. Nature 446, 305307 (2007) and S. Ide et al. Nature 447, 7679; 2007). That got attention and was good for my career. I think not having preconceived ideas helped, as did being naive and willing to try untested approaches, and being one of the first people to work in the field.

How did the 2011 Japanese quake affect you?

It was shocking. I thought the early (conservative) reports of a magnitude-8.8 earthquake were a mistake. An hour later, I saw the tsunami footage and realized that this was a wake-up call about scenarios that could exist but haven't been observed for a few hundred years.

How did it affect earthquake research?

It raised the profile of earthquake forecasting in general. It was by far the best-recorded earthquake of that size ever. Having that gold mine of data is driving big parts of the field forward, and helps my group to maintain strong collaborative ties with Japan. There have been a lot of studies based on data from that event, and they will continue for decades.

In the course of your career, will scientists get closer to being able to predict earthquakes?

Some people think it is inherently impossible. I think there is a subset of earthquakes, like those triggered by a slow-slip event with tremors serving as an indicator, that can be predicted. Those that start small and cascade into a large event may be inherently unpredictable. Unfortunately, research funding remains a challenge overall.

What is it like to get so much recognition so early in your career?

It is flattering, and it is almost certainly good attention. I was shocked to get the Macelwane award; I didn't know that two colleagues at the USGS had nominated me. It is a lot to live up to. That said, it is good to have motivation for the future. I don't have any overarching research goals: I plan to keep my focus on the big picture, finding solutions to pressing problems. I also make sure I have a life outside science. Going camping and hiking helps me to avoid research burn-out.

Additional data