Brian Savage

University of Rhode Island, Kingston, RI 02879, USA

My first exposure to Don Helmberger’s work was during a seismology class in my 2nd year in college where several students, including myself, were to present on Grand and Helmberger (1984). It was over my head, but seeing the beauty of the seismograms, tracking the upper mantle triplication, and what it revealed about the Earth was just transformative. During my time as an undergraduate at Berkeley, I also digitized (Dreger and Savage, 1999) and forward modeled seismic data under the advisement of Dr. Doug Dreger (a Helmberger student) for about two years, had an academic advisor who went to Caltech (Dr. Mark Richards) and took a couple of courses from Dr. Lane Johnson (Don’s friend from Minnesota).Little did I know the path I needed to take was already laid out -- I just had to follow it.

During the graduate school “decision” period, I had the opportunity to visit the Seismo Lab in Pasadena. Don picked me up in his blue Peugeot from the Burbank airport. I think he purchased the car in Europe many years before; it was well-loved. On that particular day, there was a torrential rainstorm in southern California that flooded the highway (I-5). As a result, we had to take surface streets and the drive to Pasadena took what seemed like forever. At some point, I asked Don, “Why is that pen stuck there?” while pointing to where the windshield met the roof of his car, to which he responded: “To stop it from leaking”. The pen didn’t work, it continually fell out, and the windshield slowly leaked onto my pants. No worries, I was already soaked from the short scamper on the tarmac from the airplane to the terminal. We eventually arrived at South Mudd, where Don kindly pulled up under the walkway, told me to run inside, go to the 2nd floor and tell them I was here while he went and parked the car.

After nonetheless deciding to go to Caltech, I started as a graduate student working with Don in 1998. Whatever the project, Don was always there to assist in the seismogram perusal, hard thinking, derivations, or cutting and tapping of figures (Don liked to do it that way, I normally just remade them digitally). He was always there to guide me and fellow grad students through the complexities of getting research done from the basic work, the writing process, to the inevitable and unwelcome return of the paper from review “l(fā)ike an ex-girlfriend”. The first paper I wrote with Don (Savage and Helmberger, 2001)was about an explosion on a Russian submarine. We were asked to get a quantitative measure of the size of the explosion; Don used his Air Force connections to get the data. He also worked diligently with me on the theory for an underwater, explosive source and how to model the data; the manuscript was sent out 4 months later, I was thrilled.

Typically, I was left to my own devices for a several weeks to work up a set of seismic data, from the newly installed regional network in southern California. Local and regional earthquake data were utilized to examine the crust and mantle structure. I would look for oddities in the data. The seismic data were always treated with special care as the amplitude and phasing were what we tried to reproduce. I was given the freedom to look at interesting responses and try to model them in the absence of Don’s advisement. It was a great learning experience, but the modeling aspect always left me lost, and from what I was told, something called “the Fog”; grasping around for something, really anything to improve the fit between the data and synthetics. This was where Don stepped in to provide needed guidance.

Eventually, I would get Don’s attention on the seismograms I was struggling with. The conversation often started with, “Hey this looks great! Who are you again?” I would normally play along and tell him my name. One day I told him my name was Pete. Don proceeded to call me Pete for the next six months; my Seismo Lab coffee mug with Pete on it broke a decade after I graduated. After capturing Don’s attention, we would work on the data and think about the possible structures (I had chosen structure over source) and the models that would best match the data in question.

In the early years, I would ask many questions while we talked, but I would never get much of a response. I would ask a question and Don would either stare out the window to his beautiful view of the mountains just north of Pasadena or he would look up and stare at the ceiling. I would think and ask another question before he had a chance to respond to the first. At some point, I decided to just let Don think. This worked much better. I have a sneaky suspicion that every time I asked a question, he had to reboot his brain to address the most recent inquiry. I now find myself doing similar things with students in my office, thinking intently about the problem at hand to provide the best guidance.

Eventually, a possible solution, or multiple solutions,would be created by Don and myself and we would talk about why the solution would or would not work, Don always drawing on his engineering notepad and throwing broken pens behind his map case. I always thought my ideas were good and workable. After our long meetings, I would run the simulations (usually 1D or 2D at this point),mine first, of course. Eventually realizing my ideas were not feasible, I ran Don’s structures. To my surprise then(but not now) Don’s structures were good, always. Don learned so much by looking at and modeling lots of data,helping him generate the appropriate response. Sometimes that response took a while, in more way than one

Quite early in my graduate work, Don took me to an Air Force meeting in New Orleans, LA. He was doing his best to introduce me to important people, guiding me, and kept buying me beers during the initial night’s social event; he also did this during a meeting in Jackson Hole.When the event was winding down, he indicated we should go get dinner, “somewhere with good food”. We wandered around a bit, me mostly following Don, and ended up at a place that served deep-fried seafood where we each ordered a large platter and more beer. Sometime in the middle of the meal, Don off-handedly said, “You are going to get me in trouble with my wife for eating all this”.Early the next morning, I barely caught the image of Don zipping past my hotel window on a morning run.

At a later stage as a graduate student, I took a lengthy research cruise across the Pacific. Around the same time,Don told me a story about a research cruise he took during his time at Scripps. This would have been sometime in the 1960s, pre-GPS, and the cruise was in and around the Aleutian Islands. During the cruise they got caught in what Don described as a bad storm which carried the research vessel to the opposite side of the island chain, without the crew’s knowledge. At that point, Don said he was never going on another cruise again. He probably gave me this guidance before I left, thanks, Don. He was still kind enough to let me be gone for as long as I was, it was a good experience and I was able to focus and get work done and some papers written.

Following graduation from Caltech, I had an impromptu discussion with Don at AGU about an incredibly odd piece of data in the late afternoon poster session. Of course, Don was wearing his leather jacket. I described the problem to him and drew out the seismograms on the back of some paper scraps. I knew what they looked like as I had been staring at them for a long time, probably months.Don took a quick look at them and said “Why do we look at seismograms? Because they are pretty”, he then thought for a few moments, and gave me some options of possible models to try, including models he and previous students had tried some decades before. Again, Don was darn close,but as always guiding and supportive. But this time Don’s models were also similar to those I had come up with(Savage, 2012).

There was one more interaction between Don and myself that I remember fondly that happened at an Air Force meeting in New Mexico. I had just wrapped up my talk about a tomography project and following a couple of questions from other attendees, Don started asking question after question about the methodology, what I was seeing, and if I believed the results. This seemed like the usual “Don is interested, this is fun, let’s go.” The questioning continued for 20 minutes or so and it was as if we were back in Don’s office working through a dataset with him. Following the talk, many attendees asked me if I had enjoyed my 2nd oral examination or not. Don always asked the hardest questions but, in a way, to get at the source of an issue or data set.

Thank you Don for always being there to lend a guiding hand as I stumbled around in the fog “putting yellow ribbons on every single one of those trees in the forest”.