Bomford Prize Acceptance
by T.A Herring
IUGG President Moritz, IUSM President James, IAG President Torge, IAG Secretary Boucher, colleagues, it is a great honor to receive the 1995 Guy Bomford Prize. Tradition has it that I speak of my research, and although at times I feel my whole career can be summarized in ten minutes, I would like to review the spectacular progress geodesy has made in the past decade and to comment on its future directions. My research has been focused on the geophysical applications of geodesy and much of this research has been made possible by the advent of space-based geodetic systems.
There has been a dramatic evolution in geodetic measurements in the past three decades with approximately an order of magnitude improvement in precision and accuracy per decade since the early 1970s when meter level baseline results were first obtained. By the early 1980s, ten-centimeter precision baseline determinations over intercontinental distances were being reported, and by the early 1990s, measurements of these lengths with 1 cm precision were common. We are currently well on our way to 1 mm precision determination of intercontinental positions even now in mid-decade. Two examples of the accuracies that can be achieved are illustrated in the Figures 1 and 2.
Figure 1: Estimates of the height at Fairbanks, Alaska, determined with VLBI (squares) and GPS (triangles). Both results are shown with one standard deviation error bars. The solid line is the expected height variations from atmospheric pressure loading computed from local pressure and assuming an admittance of 0.5 mm/mbar (coefficient from vanDam and Herring, Detection of atmospheric pressure loading using very long baseline interferometry measurements, J. Geophys. Res., 99, 4505-4518, 1994).
In Figure 1, estimates of the height at Fairbanks Alaska are shown determined by very long baseline interferometry (VLBI) and the global positioning system (GPS). The correlation between these estimates is evident but even more remarkable is that these variations are also correlated with the variations expected from the elastic deformation of the Earth due to atmosphere pressure changes. Modern geodesy has already reached the stage where temporal variations of stations positions (other than those due to tides and plate tectonics) have to be considered to fully exploit the accuracy of modern systems. Of even more importance are the methods used to judge the accuracy of geodetic systems. In the case of Figure 1, the height variations, if measured with only one measurement system could be considered to be noise, and the error budget of the system set so that variations of this type would be considered to be an unknown, temporally correlated noise source. If this process is followed, with no consideration of these variations being actual signal, then signals of this type will always be below noise level of the measurement system. But it is also clear that there are unmodeled noise sources and therefore all variations are not necessarily signals. Thus with modern geodetic systems, it would seem critical that the noise in the measurements be assessed independently of our perceived notions of the nature of the geodetic results. The only way to do this is ensure that the noise contributions from each part of the geodetic system are understood and quantified. The "art" of the modern geodetic analysis is to separate the noise from signals without artificially turning signals into noise and visa versa.Figure 2: Time evolution of the free excitation of the retrograde free-core nutation (RFCN). The estimates are made in two year intervals except for the interval 1979-1983 which was estimated as one interval (due to the lower accuracy data collected during that interval. The reason for the decay of the free excitation is not clear at the moment especially since the primary excitation mode is thought to be global average diurnal pressure changes [Sasao and Wahr, An excitation mechanism for the free core nutation, Geophys. J. R. Astron. Soc., 64, 729-746, 1981.]
The second case shown, Figure 2, is another example of temporal dependence of geodetic results. In this case, we have no other independent means of assessing the accuracy of these results. Here we must rely on knowing the individual contributions to the error budget. Analysis of the data under different scenarios can also be used to judge the significance of the results. (For example, estimates of time dependence of signals close in frequency to the RFCN mode do show variations of the magnitude seen in Figure 2.)
Fifty years ago, geodesy was determining the size and shape of the Earth. We are now able to monitor its secular changes and soon we will be reliably detecting and hopefully interpreting non-secular motions. In some senses it is amazing how stable the Earth is given that we know the planet can sustain tens of centimeter deformations daily due to Earth tides (strain rates of 10-12 per second) and the geological record shows strains of up to 50% occurring over intervals of millions of years (10-15 per second). Modern geodetic systems are approaching the point where strain rates of 10-14 per second can be measured with one day of data. Given the Earth can accommodate strain rates two order magnitude larger than this, it is curious that the observed non-secular variations are so small.
I would be remiss not to mention colleagues who have helped so many times of the years. Modern geodesy is (and I think has to be) a collaborative effort. I would like to thank my early mentors Mete Nakiboglu and Irwin Shapiro; my colleagues at MIT and the Harvard-Smithsonian Center for Astrophysics that provide a unique combination of geodesists and geophysicists, Bob King, Brad Hager, Rob Reilinger, Peter Molnar, Clark Burchfiel and Jim Davis. I would also like the acknowledge the assistance of colleagues at Scripps Institution of Oceanography, Yehuda Bock and Fang Peng; Goddard Space Flight Center, Tom Clark, Chopo Ma, and Jim Ryan; the Jet Propulsion Laboratory, Mike Watkins and Mike Heflin; the University of Berne, Gehard Beutler and Marcus Rothacher; and the University of Texas, Bob Schutz.
Guy Bomford must be impressed with modern geodesy but in many respects there is still much we need to learn. Geodetic systems have produced some spectacular results but we have still not exploited their full potential either in measurements or their interpretation. However, I am confident that we will continue to see breakthroughs in both these areas.