Memories of Donald V. Helmberger: Mentor and seismologist extraordinaire

Allison Bent

Natural Resources Canada 7 Observatory Cres., Ottawa, Ontario K1A 0Y3, Canada

I grew up in one of the most aseismic regions of the planet but decided as a child that I wanted to be a seismologist. Ignoring all the naysayers who claimed that no one really did that for a living and who told me I should choose something sensible like neurosurgery or engineering, I showed up at the Caltech Seismo Lab in the summer of 1985. I arrived equipped with the right skills but, as I quickly learned, no real knowledge of the breadth of seismological research. On day one I made the usual rounds of the faculty and soon thereafter began working on a project relocating earthquakes of a swarm in the Coso region.

A couple of months later, Don Helmberger stopped me in the hall and asked if I needed a project. I told him I was working on something but was also looking for a second project. He was wondering whether there could be a relation between stress drop and depth and was looking for a student to work on it. I thought it sounded interesting,agreed to take a stab at it and was thus introduced to the wonderful world of synthetic seismograms. It was one of the best decisions I ever made. Using a bunch of wiggles to unlock the secrets of earthquake sources and the Earth itself had an appeal that the seismicity based work,although interesting, didn’t. I was hooked. By the way, the answer to the stress drop question is a definite maybe.

A decade prior to my arrival at Caltech, Don and his previous students and collaborators had developed methods to study earthquake sources through waveform modeling and the use of synthetic seismograms (e.g.,Helmberger and Engen, 1980; Langston and Helmberger,1975). Don explained the theory to me and sent me to one of his more senior graduate students, Vicky LeFevre, for a crash course on how to use the codes. He then handed me a paper (I’ve forgotten which one) and told me to come back when I could reproduce all of the synthetic seismograms in it exactly as they appeared. One of Don’s strengths was that he didn’t spoon feed his students but encouraged them to work out problems for themselves. At the same time, he provided guidance and acted as a sounding board. He wasn’t constantly looking over our shoulders but if we went too long without talking to him, he would seek us out and ask about progress. His style of supervision was a perfect fit for my style of working. His enthusiasm was infectious.

I dislike asking for help but have learned that sometimes it’s necessary. Any time I went to Don because I was stuck, I had already tried numerous solutions and come up empty. I’d go into his office, explain the problem and say I tried x, y, and z but none of them helped and this was why. More often than not, in the middle of the explanation I’d say “Never mind, I just figured it out”. It happened enough that it wasn’t lost on Don and he once made a comment “What am I, your confessor?” I think that by articulating the problem, it became better organized in my own mind and that led to the solution.

I got through the infamous “orals” and chose Don as my thesis advisor. I never regretted it. Initially, I worked on earthquakes in the “Big Bend” region of California,which, unlike much of California, is dominated by thrust faulting. The research involved modeling both regional(Pnl) and teleseismic body waves from the World-Wide Standardized Seismograph Network (WWSSN) and similar analog stations. Hand digitizing the data is a step I don’t miss. Broadband stations were rarer than they are now, but we modeled both the long- and short-period data to assure that the source parameters fit multiple frequency bands. Don commented that it was easy to find a perfect fit for a single seismogram but that a solution which provided a very good fit to multiple seismograms and wave types recorded at different distance ranges, azimuths and frequencies was more likely to be correct than one that provided an excellent fit to only one seismogram. If the teleseismic and regional waveforms suggest different values for any parameter, there is something wrong and you need to dig deeper. He also taught me some tricks for using the PL part of the Pnl wavetrain to quickly identify whether the source mechanism was likely to be thrust or strike-slip. I saw him apply that knowledge in real time when the Loma Prieta earthquake occurred in 1989. As was typical whenever a large earthquake occurred,students, faculty and United States Geological Survey(USGS) staff rushed to gather around the seismograms on the second floor of the Seismo Lab. Don took one look at them and stated that the earthquake wasn’t on the San Andreas Fault. That later became the subject of some debate but Don was correct in that this was not a typical San Andreas earthquake.

When the Whittier Narrows earthquake occurred on 1 October 1987 I was ready to tackle it. It soon became apparent that the source was complex and a simple triangular or trapezoidal source time function was inadequate to model it. Don, who had an amazing memory for seismograms, remarked that the waves reminded him of the 1971 Friuli earthquake and suggested I consult a paper by one of his former students (Cipar, 1981) and start with that. Sure enough, the same complex source did a great job fitting the Whittier Narrows data and needed only minor tweaking for a better fit. The Whittier Narrows earthquake led to my first experience as first author (Bent and Helmberger, 1989). It came back from review needing major revisions. The comments from one reviewer were quite negative identifying about thirty “major blunders”. I was naturally somewhat concerned and showed the reviews to Don who merely shrugged them off and said the reviewer must have been in a snit that day. He pointed out two or three comments and told me to address those and ignore the rest. I did as he suggested, and the same reviewer subsequently commented on how much the paper had improved. There were no comments about the twentysomething supposedly major errors that had not been addressed.

As I continued to work with Don, we began to explore the idea of applying the same modeling techniques used to study contemporary earthquakes to historical ones. Given the relative paucity of high quality historical seismograms,we expanded on what could be gleaned from basic modeling by comparing the waveforms of the historical earthquakes to those of recent, well-recorded, well-studied events in the same region (Bent and Helmberger, 1991).This work could be considered a crude forbearer to the cut and paste method developed by Don and his students who succeeded me (Zhao LS and Helmberger, 1994; Zhu LP and Helmberger, 1996).

Conversations with Don were never boring. He asked lots of tough questions and I always tried and generally succeeded in coming up with an answer. However, there was one occasion when I was thoroughly stumped and responded with an emphatic “I don’t know”. Don said, “I don’t either but I thought it was a good question”. Ah-ha,it’s OK if I don’t always know the answer. Don had a habit of breaking off in mid sentence and staring off into space.A minute or so later he would pick up the conversation exactly where it left off. I got used to that. Another common occurrence was that he would say he was going for a coffee and would be back in a minute. Ten minutes later he often would not have returned leaving me to wonder whether I should continue to sit and wait or whether it was better to leave. When Don disappeared for days, I knew it was NBA play-off season and not to expect to see much of him until the play-offs were over.

Despite his countless contributions to seismology and his amazing insight into what seismograms were telling us,Don really didn’t have an ego. I always felt that he was proudest of me (Figure 1) when I shot down one of his suggestions. He once told me that “We (meaning the Seismo Lab faculty) are all stupid. We know a lot about one or two things and nothing about anything else”. Then he grinned and added “Except Hiroo. If he doesn’t know,no one does”.

Figure 1. Don, coffee cup in hand, looks on as I open a gift at my final Seismo Lab coffee break.

He also had a wicked sense of humor. I remember literally cutting and pasting data and synthetics to make a figure for my orals. This was the 1980s. Mercifully, some decent plotting codes were soon to be developed; I’m much better at science than arts and crafts. I recall saying in a sarcastic tone to a student who had stopped by to chat that this was stupendous and I was sure to impress the committee with its beauty. What I didn’t know was that Don was in the hall eavesdropping. A few minutes later, he popped in and remarked that it was the ugliest thing he had ever seen.

I was saddened when Doug Dreger gave me the news of Don’s passing. Seismology lost one of the great ones and I lost a mentor. However, his legacy will live on through his students and the subsequent generations of seismology they influence. More than a year later,conversations with other former students of Don always turn to how much we miss him. Don will not be forgotten.

Acknowledgments

I thank Laurel Sinclair for her review of the first draft.This paper is Natural Resources Canada Contribution Number 20210459. I would also like to thank Xiaodong Song and Thorne Lay for putting together this special issue in Don’s memory.