Crafoord Prize acceptance speech

Given by Prof. Adam Dziewonski
to the King and Queen of Sweden and assembled guests
September 16, 1998

Your Majesties, Your Excellences, Mr. President, Dr. Margareta Nilsson, Ladies and Gentlemen. Thank you for the great honor bestowed upon me and the opportunity to speak to you.

Because seismic waves penetrate the Earth's deep interior with a relatively small loss of energy, the seismic techniques are unique among the tools of geophysics in the potential to map several important physical parameters not only as a function of depth, but also as a function of geographical coordinates. These results are not only important to seismology, but have impact on many fields of Earth sciences such as mantle convection, geomagnetism, geochemistry, petrology, mineral physics and geodesy. I believe that the principal reason I am here today is the work published in a relatively short period of time, from 1984 through 1987, during which our group at Harvard established the principal elements of the 3-dimensional (3-D) structure of the Earth from the crust to the inner core. From gray images of a 1-D Earth, in which physical properties change only as a function of depth, we advanced to distinguishing lateral variations at all depths. A series of developments both preceded and followed this period, without which this rapid progress would not have been possible and its significance would have been much less clear.

I owe much to support by other people. Most of all to my wife, Sybil, who always believed in me and put up -- particularly during the long early years -- with impossible hours and preoccupation with professional matters. From among my friends, who are many, five have had an especially important impact on my work. In order of decreasing age they are: Anton L. Hales, Freeman Gilbert, Don L. Anderson, John H. Woodhouse and Göran Ekström. The reason for my gratitude will be clear from the rest of my presentation.

All my formal education was in Poland: University of Warsaw (M.S. 1961) and Academy of Mines and Metallurgy in Cracow (D.T.S., 1965). My first place of employment was the Institute of Geophysics of the Polish Academy of Sciences; I am delighted to see in the audience Professor Slawomir Gibowicz, a representative of the Institute.

When I arrived at the Soutwest Center for Advanced Studies in Dallas, Professor Anton L. Hales was the Head of the Geosciences Division. He is one of the few `present at the creation', when the field of geophysics, as we know it today, was formed during the 1930's through the 1950's. It was the encouragement and support of Anton that led to my decision to begin a concerted attack on determining the whole Earth structure.

In 1970 I began my collaboration with Freeman Gilbert on the interpretation of the records from the 1964 Alaskan earthquake and derivation of new Earth models. Freeman, the world's foremost theoretical seismologist, had all the necessary tools; he also taught me how to approach the impossible task of deriving an Earth model from a limited set of inaccurate, and sometimes erroneously identified, data. Our first joint paper was on the solidity of the inner core. The period of our close collaboration, during which I moved from Dallas to Harvard, culminated in a paper which simultaneously attacked the question of the Earth's structure and studies of earthquake mechanism.

In 1974, I became convinced that any further improvement in the knowledge of the Earth's structure must come from investigating it in three dimensions. A pilot study indicated that indeed there are large scale 3-D heterogeneities in the lower mantle. In a paper with Bradford Hager and Richard O'Connell we demonstrated a significant correlation between the velocity anomalies and the large scale geographical variations of the geoid.

The interest in modeling of the Earth brought me together with Don Anderson. In 1977 we were asked to develop a Reference Earth Model (REM) that would satisfy most of the available seismological data and be consistent with other geophysical observations. After three years of effort we constructed PREM, which included not only seismic velocities and density, but also the anelastic parameters and anisotropy in the uppermost mantle. The work with Don taught me the simplest models are not always the best.

The availability of digital data from a global network of seismographic stations, made it possible to approach systematically the problem of studying earthquake mechanism of moderate and large earthquakes. The Harvard Centroid -- Moment Tensor (CMT) project began in 1981 and continues to this day; it now contains data for over 15,000 earthquakes. John Woodhouse, in the beginning, and Göran Ekström, today, were the most essential members of the team. For the last few years, Göran has become the leader of the project.

In a time span of some four years, and in close collaboration with John Woodhouse we have written a series of papers on the three dimensional structure of the Earth, ranging from the uppermost mantle to the core-mantle boundary. The work established the excellent correlation between the surface tectonic regime and lateral variations in the structure down to depths exceeding 200 km. It indicated a relatively deep origin of mid-ocean ridge anomalies and confirmed the existence of `continental roots'. The transition zone structure was significantly different, indicating correlation with the principal subduction zones. The characteristic structure of the lowermost 1000 km of the mantle is the ring of fast velocities circumscribing the Pacific and two very large slow velocity anomalies, one under the Pacific and the other under Africa.

Andrea Morelli and I have recovered the topography of the core-mantle boundary, showing that it has a relief on the order of 10 km. With Morelli and Woodhouse we have shown that an explanation of the observed anomalies of the inner core travel times requires anisotropy of the inner core with the axis of symmetry approximately aligned with the rotation axis.

During the following 10 years, up to the present, this rapid sweep through the Earth's interior has been followed by increasing resolution of the models, gathering sets of new types of data needed to elucidate the structure in the critical regions of the mantle, such as the transition zone and the lowermost mantle. It was a rewarding and satisfying period; most of the work at other institutions strongly supported our findings. But the excitement of being on the leading edge of research continues. Strangely enough, the two regions of our particular current interest are about as far from each other as they can be. In both cases, the phenomenon of anisotropy plays an important role. In a work led by Göran Ekström, we have discovered that not all ocean basins change their properties with age as predicted by plate tectonics. In particular, the Pacific plate seems to be different from any other region, with its interior being especially anomalous. In continuation of the work with Wei-jia Su and Raymond Jeanloz on the inner core, we studied the possible differential rotation of the inner core.

Two resources have been particularly important in my work. The first was the existence of data from global seismographic networks. Without the computerized analysis of seismograms, spanning time intervals from minutes to days, little of what I said tonight could have been accomplished. Now, the land based global broad-band network is nearly completed. Scientific leaders from many countries contributed to this accomplishment, among them --- present here, Terry Wallace from the United States and Enzo Boschi from Italy. The next step is instrumentation of the oceans. Even when the smallest islands are used, there are oceanic areas several thousand kilometers in diameter without seismographic stations. The recent experiments show that seismographs on the ocean bottom can perform as well as, or better, than stations on islands of moderate size. An approximately even receiver coverage is necessary to produce unbiased images of the interior of our planet.

The other resource is the availability of the theory requisite for a problem studied. Some say that the lack of the theory accompanied by the appropriate numerical methods, is now more of the impediment in achieving progress in interpretation of 3-D waveforms than gaps in data coverage. There is a need to devote more attention to the support of novel quantitative methods in seismology. With notable exceptions, such as Jeroen Tromp, my colleague at Harvard, I see few successors to Freeman Gilbert and John Woodhouse.

I am fortunate to have found interesting problems to study. As I said, my work has been immensely helped and influenced by many friends, and all of this has been a great fun. I am very grateful to the Royal Swedish Academy of Sciences for the honor of being selected for the Crafoord Prize.