EOS VOLUME 92 NUMBER 8 22 FEBRUARY 2011 Geophysicists was elected Fellow of the Royal Society at during that period included the Sverdrup in 2006. In April 1998, hundreds of friends cont. from page 63 the age of 37. By way of decompressing after Gold Medal in 1974; president of the Mary- and colleagues packed Johns Hopkins’s the publication of Upper Ocean, he published land Academy of Sciences, 1979–1985; Shriver Auditorium to hear the world’s most The Heart of the Earth in 1968, a highly origi- and Fellowship in the American Meteo- distinguished fluid dynamics experts from career of innovative contributions to geo- nal introduction to solid Earth geophysics rological Society in 1980. As the second around the world pay tribute to Phillips on physics through fluid mechanics. intended for beginning scientists. Like Upper energy crisis hit the United States in the late the occasion of his retirement from active In 1957, Phillips joined the Department of Ocean, Heart went through numerous reprint- 1970s, Phillips saw how cavalierly Ameri- teaching duties. Mechanical Engineering at Johns Hopkins ings in multiple languages and is unique in its cans were behaving toward an issue that Phillips was much revered for his con- University. He returned briefly to the United emphasis on the scientific method as applied directly affected their wallets and their siderable diplomatic skills and rock solid Kingdom as assistant director of research to geophysics—it clearly explains how geo- security. In 1979 the Johns Hopkins Univer- integrity as well as his professional accom- in the Department of Applied Mathematics physics is done, in addition to what is pres- sity Press published his whimsically titled plishments. On several occasions, univer- and Theoretical Physics at Cambridge, then ently known about the Earth. Last Chance Energy Book, a lively yet sober- sity presidents turned to him for leadership came back to the United States permanently In addition to wave generation, the action ing account of the true costs and risks of during times of crisis. Although he was fun- in 1963 as a full professor at Johns Hopkins, of wind blowing over the ocean surface pro- energy policy negligence. It is now clear damentally a private man, he possessed becoming a U.S. citizen shortly thereafter. duces turbulence in a layer that grows with how far ahead of its time this book was. a gracious charm and a quick sense of By the early 1960s it became evident that time as the result of entrainment mixing with The next chapter in Phillips’s research humor. He had limited patience for trivial further progress on upper ocean dynam- the denser water below. In a famous laboratory began in the late 1980s, when he started things but always seemed to have time for ics required solving the thorny problems of experiment, H. Kato and Phillips measured collaborating with several Johns Hopkins colleagues and students. He was a person nonlinear wave interaction and ocean mix- the rate of entrainment in a stratified fluid as geologists including Lawrence Hardie and one could talk with. Apart from his own ing. Anticipating that these topics would a function of the applied surface stress and John Ferry on the problem of how aque- research and teaching, Phillips drew his have broad theoretical interest as well as derived the first scaling law for entrainment ous fluids infiltrate and react within per- greatest pleasures from the simple rituals many practical consequences in ocean sci- that was directly applicable to wind-induced meable sedimentary and metamorphic in his life: family; woodworking; skippering ence, Phillips began to collect material for a mixing in the upper ocean, now understood to rocks. This work culminated in 1991 in a a small sailboat across the Quissett, Mass., definitive book on ocean waves and turbu- be a critical process for climate regulation (see Cambridge University Press monograph, harbor; or just chatting with a colleague in lence. In 1965 his monograph Dynamics of J. Fluid Mech., 37(4), 643–655, 1969). Flow and Reactions in Permeable Rocks, that familiar, hallmark pose— leaning back the Upper Ocean was awarded the Adams In 1967, Johns Hopkins formed the and in 2009 in a second monograph on this in his chair and quietly puffing his cigar. Prize by the Royal Society of London. Upon Department of Earth and Planetary Sci- subject, Geological Fluid Dynamics: Sub- His departure marks the end of an era, and its official publication a year later, it quickly ences by combining the former Geology Surface Flow and Reactions. More honors his presence will be sorely missed. became a fixture on desktops throughout Department with faculty from the Oceanog- came during this period, including elec- the oceanographic world, was republished raphy and Mechanics departments. Phillips tion to the U.S. National Academy of Engi- —PETER OLSON, Department of Earth and Planetary three times, and was translated into Russian became its first and longest-serving chair neering in 1996, Honorary Fellow of Trinity Sciences, Johns Hopkins University, Baltimore, Md.; and Chinese each twice. In 1968, Phillips and its main visionary. Personal accolades College in 1997, and finally, Fellow of AGU E-mail: olson@ jhu .edu Applied Science at the California Institute of Honors Technology; and Mark D. Zoback, Benja- Ralph Cicerone has been elected to min M. Page Professor of Geophysics, Stan- LETTERS a second 6-year term as president of the ford University, Stanford, Calif. U.S. National Academy of Sciences begin- Steven Koch has been appointed the ning 1 July 2011. As president, Cicerone also new director of the National Oceanic and serves as chair of the National Research Atmospheric Administration’s (NOAA) On the Correct Use of Statistical Tests Council, which conducts independent sci- National Severe Storms Laboratory in Nor- Comment on “Lies, Damned Lies, and Statistics (in Geology)” ence, engineering, and health policy studies man, Okla. Koch, who begins his new under a congressional charter. assignment in April, has been director of The U.S. National Academy of Engineer- the Global Systems Division at NOAA’s Earth Taking the distribution of global seismic- statistical tests used in the geosciences to ing (NAE) has elected 68 new members, System Research Laboratory in Boulder, ity over weekdays as an illustration, Pieter “make deductions more ‘objective’” are at including five AGU members: Michael R. Colo. Prior to joining NOAA, Koch was an Vermeesch (Eos, 90(47), 443, doi:10:1029/ best useless, if not misleading. Hoffmann, James Irvine Professor of Envi- associate professor at North Carolina State 2009EO470004, 2009) in his Forum pre- In complete contradiction, we affirm that ronmental Science, California Institute of University in Raleigh and a meteorologist sented an argument in which a standard statistical tests, if they are used properly, are Technology, Pasadena; former AGU presi- at NASA Goddard Space Flight Center in chi-square test is found to be so sensitively always informative. Vermeesch’s error is to dent John Orcutt, professor of geophysics Greenbelt, Md. dependent on the sample size that probabili- assume that it is possible to reduce in the and Distinguished Researcher, San Diego ties of earthquake occurrence from these chi-square test simultaneously both the num- Supercomputer Center, University of Califor- In Memoriam tests are uninterpretable. He suggests that ber of earthquakes in each weekday and nia, San Diego, La Jolla; Karsten Pruess, the sample size by 10. Instead, Vermeesch senior scientist, Earth Sciences Division, Thomas J. Ahrens, 74, 24 November should have taken 10% of the original data Lawrence Berkeley National Laboratory, 2010, Fellow, Tectonophysics, 1959 set and then again grouped it into 7 days. Berkeley, Calif.; Ares J. Rosakis, Theo- Luis Gomberoff, 68, 13 September 2010, Without doing this, it was inevitable that Ver- dore von Kármán Professor of Aeronautics Space Physics and Aeronomy, 1993 meesch reached his erroneous conclusion. and professor of mechanical engineering, Benoît Mandelbrot, 85, 14 October 2010, Under the conditions that the sample is and chair of the Division of Engineering and Fellow, Nonlinear Geophysics, 1986 large enough, that the observations are inde- pendent and identically distributed, and that the zero hypothesis holds true, the p value should not depend essentially on sample size, as long as there are at least 10 observa- tions per bin. The above mentioned change of the p value with sample size should thus signal the existence of some violation of the conditions for the chi-square test to be valid and not that statistics lie. is is an opportunity to nominate your colleagues for the recognition of a There are many possible causes for such lifetime! AGU Union medals and awards are among the most respected in violation, including aftershocks and catalog the Earth and space sciences communities worldwide. Past honorees can be incompleteness. By removing aftershocks using a standard declustering method and found on AGU’s Web site: www.agu.org/about/honors/union/. using a more complete catalog with events of magnitude greater than 5.0, we obtain Nominations must be complete and received at AGU headquarters by: a p value of 0.46 for any given earthquake to fall on any day of the week. This large 15 March 2011 p value indicates that there is no preferen- tial day that earthquakes occur. We can thus e following medals will be presented in 2011 affirm that the main earthquake shocks with M ≥ 5.0 are distributed uniformly over the William Bowie Medal seven weekdays, as expected from seismo- James B. Macelwane Medal logical intuition. John Adam Fleming Medal When properly used and interpreted, sta- tistical tests always reveal useful informa- Maurice Ewing Medal tion. Ridiculously small p values as found by Robert E.