A strange symbiosis: Seismic ray theory and Saturday touch football — Donald V. Helmberger as a graduate advisor

Terry C. Wallace

Director Emeritus, Los Alamos National Laboratory, Los Alamos, NM 87544, USA

At the end of December of 1978 I was sitting in the office of Caltech Professor Don Helmberger, he was peering out his window towards the San Gabriel Mountains, silent and I supposed, deep in thought about some mystery of seismology. When he finally spoke, he said “Chuck Foreman’s days as a great running back are over”. The previous day the Los Angeles Rams had eliminated the Minnesota Vikings (despite Fran Tarkington’s best efforts) from the NFL playoffs. Don grew up in Minnesota, and was a Vikings fan – something he and I shared. We then had a brief discussion about the Vikings, and finally Don returned to peering out the window. When he next spoke it was to ask about the status of my digitizing LRSM seismic records for a series of large yield nuclear explosions that had been detonated at the Nevada Test Site in the 1960s. He never talked about the Vikings again in his office: however, he always asked me what I thought about them every Saturday when we played a rousing game of touch football on the Caltech field across from the South Mudd building. Don was a deeply passionate man — his curiosity for seismology was profound, but he also loved life in a way that made him a remarkable graduate advisor. In my thesis I had a simple acknowledgement: “I have particularly benefited from the support of, and interaction with my thesis advisor Don Helmberger. Don, besides being a Minnesota Vikings fan,has deluged me with ideas, good and bad!”

In the fall of 1973, I had every intention of attending Caltech after I graduated from Los Alamos High School the following May. However, it was not to be, and I would not arrive in Pasadena until the summer of 1978 when I matriculated in the PhD program after getting my undergraduate degrees at New Mexico Institute of Mining and Technology. On reflection, this delay in arriving at Caltech was one of the best things that ever happened to me; I arrived in what others have coined “the golden age of seismology” and the Seismo Lab was unlike any place else in the world. Only a decade earlier Oliver, Isacks and Sykes had published “Seismology and the New Global Tectonics”, and for a brief period of time Earth physics was the hottest science topic around. Seismo Lab professors, graduate students, and a steady stream of visitors from around the globe created an incredible environment of discovery. Daily coffee breaks — with attendance from nearly everyone – always produce lively (and often heated)discussion on topics ranging from the nature of the core to the predictability of earthquakes. I have seen many attempts at creating a similar environment in the nearly 40 years since I left Caltech, but nothing has come close.

Arriving into such a remarkable place was both exciting and extraordinarily intimidating. All the incoming students had to choose three research topics (with different professors) that would be defended in a qualifying exam that felt like the star chamber trials of 16thcentury England. In retrospect, the pressure to succeed in the qualifying exam was mostly peer pressure; a chance to show your fellow students that “you belonged, intellectually and creatively”. In August of 1978 a moderate sized earthquake occurred near Santa Barbara (and earthquakes seemed to be quite rare in California after the 1971 San Fernando earthquake) that produced some interesting strong motion recordings. This provided the opportunity for one of my research topics, and how I first came to work directly with Don Helmberger. One of Don’s first students,Tom Heaton, had used Don’s work on computing synthetic seismograms with generalized ray theory to model the San Fernando earthquake and showed that the rupture of fault was rough; there were areas, or patches, of large and small displacement, not a smooth rupture from one end of the fault to the other. Using Heaton’s work as a guide I produced my first paper at Caltech, Wallace, Helmberger and Ebel (1981a). When I went to see Don about my proposal for working on the strong motion records, he asked me if I played football, and invited me (or, from my perspective, strongly suggested) to play in the Saturday touch game. I rarely missed a game for the next 4.5 years,and I can say today whenever I smell the pungent aroma of ozone, I get happy tears in my eyes. The LA basin was pretty polluted in the late 1970s, and ozone was one of the most obvious irritants created by 10 million automobiles driving on the freeways. I remember with great fondness the burning eyes and lungs created by chasing down (or at least trying to) Don on a field marked with scattered tee shirts before he could score a touchdown.

The introduction to modeling an earthquake made it obvious to me that I was far more interested in the mathematics of the simulation methods than actually understanding the earthquake. Fortunately, Don was very supportive of exploring improvements to his Cagniard-de Hoop formulization of ray theory. Don suggested I work on regional distance seismograms. These seismograms,recorded at regional distances (2°–12°), are quite “messy”and complicated due to the waveguide nature of the crust.The body wave trains are essentially crustal reverberations.If these complicated waveforms are modeled with synthetic seismograms then significant information can be learned about the seismic source and the structure along the travel path. With certain restrictions, the long-period regional body waves (Pnl) from shallow, continental earthquakes can be modeled with layers (crust) over a halfspace (mantle). Generalized ray theory and the Cagniardde Hoop technique can be streamlined for computing a synthetic seismogram in such a structure. We developed an approximation to the travel-time equation, which results in an analytic inversion for the de Hoop contour. The simplicity of the individual rays required that the displacement potential need only be evaluated at a small number of time points; small changes in structure are, to first order, expressed in terms of the timing of different arrivals. It is possible to "stretch" or "squeeze" the synthetic to simulate a change in structure. Therefore, a single Green's function can be used to investigate a whole suite of structural models.

This mathematical work meant that I had the“opportunity” to rewrite a version of Don’s generalized ray theory computer code, ASERIES. In the late 1970s that meant Fortran coding on punched cards. My new computer code, named DeHoopster, required 3 full boxes of punched cards (with inputs) that I would have to haul across the Caltech campus to submit to an operator sitting behind a high wall at the computer center. It was not simple converting the Helmberger code to DeHoopster because Don was quite fond of “picket” do-loops that he would simply insert an instruction to skip hundreds of cards in his computer stack. Finding these programing treasures always meant days of making sure the modified code could be verified. After the code was done it opened up a plethora of modeling opportunities. I wrote a half dozen papers on modeling regional distance seismograms for seismic source, crustal and upper mantle structure, and the forensic analysis of underground nuclear explosions (these papers are included in the references: Wallace et al., 1981b;Wallace and Helmberger, 1982; Wallace et al., 1983, 1985,1986, Burdick et al., 1984, and Lay et al., 1984).

One of the forgotten aspects of the late 1970s and early 1980s is that the primary funder of seismic research in the US was not the National Science Foundation, but rather it was the Department of Defense through ARPA and DTRA. In the late 1950s there was an intense debate about the viability of monitoring a complete prohibition of nuclear testing. The US decided to conduct a nuclear test that was underground, and fully contained (meaning no release of materials and debris from the explosion). This test, code named RAINIER, was quite modest having a yield of 1.7 kt (less than 8 percent of the size of the first nuclear explosion, TRINITY, which had a nuclear yield of approximately 25 kt). The seismic waves created by RAINIER were quite unexpected; it was detected at least 1000 km from the Nevada Test Site, and there were very“earthquake like looking” S waves. This prompted a US group of experts chaired by Lloyd Berkner to exam the needs of the country to monitor a comprehensive test ban.The so-called Berkner Report was breathtaking in its recommendations: a funded international seismic network,improvements in seismic instrumentation, and perhaps most importantly, a massive program in basic seismology research (not just focused on nuclear explosions). This program came into existence, and ultimately funded the education and research of a generation of seismologists. I am one of those, and Don Helmberger had dozens of research grants that can be traced back to the RAINIER nuclear test.

I wrote several dozen reports and papers on various aspects of what is now called “forensic seismology” trying to understand what seismic and other geophysical signals could tell us about a nuclear weapon explosion (for a few of these papers see the last five references in the bibliography). My early focus with Don’s guidance was to understand the complexity of the body waves, exploring why S waves could be so strong for a purely compressional source. Figure 1 shows how a large explosion at NTS, COLBY, could be used to simulate another large explosion, BOXCAR by the addition of a simulated earthquake. This simulated earthquake-type source became known as tectonic release and is thought to represent the explosively driven fracturing of rock around the explosion source.

Figure 1. A comparison of the P and PL waveforms for BOXCARand COLBY at WWSN station LUB. Shown below is the COLBY waveform summed with a synthetic seismogram to simulated the tectonic release. The synthetic is created with a pure strike slip orientation and a moment of 5×1024 dyne·cm(From Wallace et al., 1983).

Every year the results of the research funded by the DOD were presented at a meeting which became known as the Seismic Research Symposium. Thorne Lay (my classmate) and I would attend these meetings with Don,which indirectly led my career path back to my hometown of Los Alamos, and I eventually became the 11th director of the birthplace of the atomic bomb. At the Research Symposium we would present our results — in those days the preferred presentation media was transparencies projected on an always too small screen in front of an audience of a hundred scientists and officials from the funding agencies. Don would always ask to “help” by placing and changing the transparencies during the course of the talk. This was a major challenge because Don would become interested in some aspect of the talk and would decide to skip whole sections of the presentation to get to the material he was thinking about. The first time Don did this it was quite unnerving; by the fifth meeting I actually looked forward to the improvisation I would be forced into with Don’s whim (or guidance).

Don Helmberger was an extraordinary advisor. It is easy to point to his brilliant insight on the nature of wiggles on a seismogram as his greatest talent. However, in the hindsight of a career, it is obvious Don’s contributions were far, far greater. His insatiable curiosity, passion for understanding, and his hidden playfulness exposed mostly on the gridiron, taught how to be a scientist, how to be a critical thinker. I am sure that Don Helmberger’s thousand papers will be cited as his legacy, but his biggest contribution was actually molding a generation of scientists.