As part of the American Meteorological Society's 30th Conference on Hurricanes and Tropical Meteorology in Ponte Vedra Beach, Florida, in April 2012, an academic lineage (“family tree”) of that community was presented to document the history of contributors to the field on the anniversary. For every self-identified or colleague-identified tropical meteorology scientist, the year of the person's most senior degree, major professor or mentors of that degree, and institution of that degree were documented and graphically presented. This information was supplemented through mining of websites, libraries, news and journal articles, obituaries, and other various historical archives. This manuscript documents the genesis of the family tree, the overall history represented by it, some statistics represented by the current incarnation, colorful personal stories that have come forward during its development, and plans for its expansion to the broader meteorology community.
Countless meteorology students enter college having what can quite accurately be described as a romance for weather. If this adolescent professional relationship could also be equated with a crush, then heartbreak is what follows for many. Those intense—but superficial— childhood feelings for meteorology are often later strained by unexpected but necessary courses in calculus, physics, and computing, partially replacing romance with angst. As is the case with life in general, through reflection and perspective, one will hopefully later find a much more profound appreciation. This critical evolution can be made much more difficult by a poor advisor, or much less painful by an exceptional mentor. The latter explicitly or implicitly reminds us why we fell in love with meteorology as a child. This manuscript is dedicated to those mentors who made that process for students profoundly rewarding, and equally to those students who persevered in the face of adversity and can now look up at the sky or down at a map and smile with a deeper and satisfying understanding.
BACKGROUND AND METHODOLOGY.
We all grow up with the weight of history on us. Our ancestors dwell in the attics of our brains as they do in the spiraling chains of knowledge hidden in every cell of our bodies. —Shirley Abbott
We need to haunt the house of history and listen anew to the ancestors' wisdom. —Maya Angelou
Never destroy any aspect of personality, for what you think is the wild branch may be the heart of the tree. —Mrs. Henry George (Agnes DeMille's grandmother)
Why waste your time and money looking up your family tree? Just go into politics and your opponents will do it for you! —Mark Twain
While informal meteorological studies trace back to antiquity, tropical meteorology itself is a very young science— dating back decades rather than centuries. When this short existence is viewed in context of today's state of the field, it implies two important characteristics: 1) it seemingly has its birth from a diverse set of parents, although the academic nature of those parents is unexplored; and 2) a remarkable amount of progress in tropical meteorology has occurred in a relatively short period of time. In light of this, there has been a desire for some time within that community to document the roots of the field, and to see how it interconnects with the classical sciences, before such history is lost. Indeed, colleagues have discussed at numerous conferences the desire to pursue a robust documentation of this academic genealogy, or “family tree.” At the Seventh International Workshop on Tropical Cyclones (IWTC-VII) in La Réunion, France, the authors decided to pursue this project, buoyed by the late-night vigorous discussion with three colleagues (J. McBride, C. Landsea, and N. Shay) and a unanimously appointed moderator/catalyst (J. Walker); see Fig. 1.
The goals of the project are multifaceted: 1) first and foremost, to document the human academic history of the field, noting its diverse roots and continuing evolution today; 2) to give young and senior scientists alike a perspective of their part and relation to others in their professional community; 3) to provide the opportunity (through as yet to be collected information) to trace the genesis and evolution of ideas within the community; and 4) if successful, to allow the tree to serve as a skeleton on which the history of the entire meteorological field could be developed.
It was immediately apparent that difficulties would be encountered in pursuit of these goals. The determination of graduate major professors is relatively easily done today through solicitation of the living. Finding that information for deceased colleagues, however, sometimes takes substantial detective work. Further, members of the community (living or deceased) can have more than one major professor or may consider their “mentor” or “advisor” a different person than the one who signed the thesis or dissertation. Occasionally, a colleague requested being listed as “self-made” without reference to his/her major professor (if any), defeating the goal of a historical connectedness tree, but it is a request that must be honored.
While the response from colleagues on the proposed project was overwhelmingly positive, there was a small minority response who felt it was an unwise idea. The specific complaint was that the mere formation of such a document would constitute a condoning of nepotism or aristocracy. The authors respectfully but strongly disagree with this conclusion, as no such objective measure of value or importance is explicitly given or even attempted in the resulting family tree. Nonetheless, as stated earlier, the overwhelming response was positive to the idea, with several colleagues providing suggestions and feedback as data collection began. Below is one colleague's provocative yet insightful perspective into the uniqueness of such a project:
Academic family trees are unusual in that only progeny who become academics are able to reproduce.1 It would seem that it is actually impossible to produce a bastard academic heir, though in point of fact I know several individuals worthy of that title. —Name redacted.
Academic lineages have been attempted with varying degrees of completion by some other disciplines, such as theoretical computer science (Johnson 1984; Parberry and Johnson 1995), philosophy (https://webspace.utexas.edu/deverj/personal/philtree/philtree.html), and neuroscience (David and Hayden 2012). The Mathematics Genealogy Project (Jackson 2007; http://genealogy.math.ndsu.nodak.edu/index.php) represents perhaps the current largest such undertaking, hosting information for over 150,000 individuals in studies of math and related disciplines. Such a body of academic connections can provide insight into specific traits inherited by the student from the mentor (Malmgren et al. 2010). Indeed, connections between societal networks and the sharing of ideas can be explored in many ways (Bowen and Wilson 2012), such as research collaboration (Newman 2001). One such other connection with mathematics colleagues is also shown in the “Erdös number” (Goffman 1969; Odda 1979), a measure of the “degrees of separation” from coauthorship with Paul Erdös. Finally, no publication on academic genealogies would be complete without a (partially) satirical but (highly) insightful take on the topic (see www.phdcomics.com; Fig. 2).
Depending upon the scope of the above-mentioned family trees, a varying amount of information was requested. The present project was initially targeted to the tropical meteorology community to keep the scope manageable, so the information requested was straight-forward: student's name, major professor's name(s) for the most senior degree, the year of that degree, and the institution granting that degree. Solicitation requests for information were sent to three discussion/mailing lists in 2012: tropical storms (hosted by Colorado State University), Map (hosted by University at Albany, State University of New York), and the American Meteorological Society (AMS)'s registration list for the 30th tropical meteorology conference held in April 2012. The resulting database was entered manually into a spreadsheet, where multiple spellings of names and aliases were identified and resolved.
Finding a suitable medium to visualize the incoming genealogical information proved to be one of the largest logistical issues. Despite the overwhelming number of graphical products that could display such data, Graphviz (http://graphviz.org/) was chosen as a simple yet hierarchical method to depict the family tree as a flowchart. Plotting advice and sample Graphviz code were provided by Mitch Keller of the Mathematics Genealogy Project. The data spreadsheet was converted into a format readable by Graphviz software, and then the data were plotted as a tree within Graphviz. Further, the data were organized such that the tree's vertical axis represents graduation year and each individual box is given a color based on the institution of the degree.
This preliminary version of the tree provided an excellent skeleton; however, it was clear there were major holes and gaps within. Occasionally, it was easy to fill these by simply contacting students or grandstudents of a missing scientist. However, in some cases, there was simply no memory of the detailed historical lineage.2 Consequently, the authors resorted to obituaries, journal articles, online or hard copy theses, and other historical documentation to determine missing information. In some cases, where a branch of the tree connected back to the far more mature Mathematics Genealogy Project, it was easier to fill such holes by simply searching within that project's online database. While Wikipedia often could help in filling these holes, there were cases later discovered where the asserted major professor in Wikipedia was in question based upon feedback from that professor's progeny; thus, other primary and secondary historical sources were searched (such as resorting to original hard copy theses in dusty libraries).
Several procedural hurdles had to be overcome to produce a viable electronic or printed version of the tree. Despite the growing availability of novel presentation mediums (e.g., Prezi, http://prezi.com), the portability and ubiquitous ability to view very large image files made the image file option more attractive. Test raster images were produced, although the inflexible format and large file size made working with them difficult. Vector-based images, notably in portable document format (PDF), proved to be more compact and able to be indexed, but they presented other difficulties. Adobe Acrobat Reader, available for all environments, has a limit of 200 in. (5.1 m) for PDFs it can display. While other software can display much larger PDFs, each has drawbacks: two such options are Evince (http://projects.gnome.org/evince; limited zooming and cannot search for text) and Brava! Reader (www.bravaviewer.com/download-brava-reader; alleviates all prior drawbacks, but it is only available in the Windows operating environment). Additionally, most poster plotters have a finite width of printing, which is generally smaller than needed to make the family tree readable. Consequently, later versions of the tree had to be printed in multiple strips, necessitating splitting the original PDF into multiple files, which can lead to artifacts at splice points. Finally, it was a considerable challenge to create a unique discernible color to correspond with each institution (given the number of institutions represented), while simultaneously avoiding color shading/text–color combinations that made reading the names difficult.
THE CURRENT TREE, STATISTICS, AND ANECDOTES.
The first version of the tree was presented at AMS's 30th Conference on Hurricanes and Tropical Meteorology in April 2012 in Ponte Vedra Beach, Florida. This version displayed approximately 1000 scientists from over 100 institutions. Figure 3 displays that first version of the tree and a subset of the community that was examining it.3 During the conference, colleagues missing from the tree who wished to be added were able to submit their information. During and after the conference, numerous correspondences were received commenting on the project and the nature of the resulting tree. Three of the most memorable comments are as follows:
I heard from colleagues that the poses [of those viewing the tree] reminded them of people looking for relatives on the Vietnam Wall in D.C. (the latter obviously in an infinitely more important subject). Nonetheless, there was clear expression in both body and verbal language of reverence for those who came before with many taking close-up pictures of their branches. —Paraphrased consensus of Ed Zipser and Lance Bosart
The analogies between this tree and [biological] genealogical trees kept me thinking about whether the former has examples of the latter for such possibilities as offshoots, grafts, adoptions and . . . artificial insemination. For the subject at hand, I suppose we meteorologists are “mutations” from the Math, Physics, etc. tree; at least the mathematicians, physicists, etc. seem to think of us as so. We'd probably attribute it to evolution. —Ed Rappaport
Looking at it as an abstract work of art, I was struck by how beautiful it is—like some sort of exotic biblical temple or futuristic edifice in an unknown civilization. —Jeff Rosenfeld
Following the conference, the additional information received was added to the tree, leading to a total of 1136 scientists among 129 institutions. The updated tree was next presented at the 93rd AMS Annual Meeting in Austin, Texas, in January 2013, using larger text to aid reading and spanned 16.7 ft (5.1 m) × 7 ft (2.1 m). Further feedback and entries were provided by attendees, leading to the current version that is necessarily shown without text in Fig. 4 [a high-resolution PDF version of the tree (and future updates) is available online at http://moe.met.fsu.edu/familytree]. The tree can be overwhelming at first glance given the number of colleagues presented, and given that those currently alive are typically in the densest part of the tree (the bottom third). To aid navigation, a “road map” of a sample of early scientists was provided at the conference and is shown in updated form in Fig. 5. Given the much broader and larger audience of the AMS annual meeting, it was common throughout the week to have renowned scientists visit the tree and graciously volunteer invaluable and otherwise hard-to-find additional information (e.g., Fig. 6).
The earliest “root” of the tree shown is a largely arbitrary choice to keep the printed size manageable, as extending farther back in time rapidly compresses the most recent decades (with the most data) into an increasingly small vertical space. Müller (1822), his student Helmholtz (1842), grandstudent Hertz (1880), and great-grandstudent V. Bjerknes (1890) form this first branch of the tree (Fig. 7a). In reality, the tree can be extended backward centuries into the biological, astronomical, mathematics, and physics communities through online searches of the Mathematics Genealogy Project (not shown). Additional early roots into the tropical family tree are shown in Fig. 7b, illustrating not only the wide range of parental study but also the broad diversity of geographic origin.
Despite the admittedly incomplete content of the family tree at this time, some interesting statistics about the current incarnation of the tree are provided. Table 1 shows the institutions with the highest representation in the family tree. Of note, The Florida State University has the advantage of being the authors' institution with easy access to local information, and Colorado State University recently documented online a plethora of historical information for the Atmospheric Science Department's 50th anniversary (www.atmos.colostate.edu/anniversary/anniversary.php). Many institutions are clearly underrepresented—especially those outside the United States. The family tree's bias toward currently active United States–based institutions partially accounts for many discrepancies, as noted in the steep drop-off of information available before the mid-twentieth century (Fig. 8). While the authors could have delayed publication until a hypothetically complete version of the tree was available, it quickly became clear that such a state would likely never exist and that the most effective method for expanding the tree would be publication in its late adolescent state.
During and after the solicitation process for colleagues' information, many anecdotes (serious and humorous) were provided as part of the information sent. We share a selection of these anecdotes not only to express deep and sincere gratitude to those represented but also to illustrate the further human dimensions (including the undeniable humanity of intellectually revered figures) that cannot be communicated by the limited dimensions of the tree. These anecdotes are not limited to student–advisor relationships but extend to insightful perspectives about historical colleagues outside of mentoring.
I finally got Rossby to agree to a final “defense of thesis” time in 1948. I went to the Institute of Meteorology and found that Rossby had been at a party the night before [undoubtedly at Frannie Day's (Ashley) apartment]. At the party Rossby had been pacing up and down insisting that I (Bryson) would definitely not be passed the next day. Hanging onto his arm and reasoning with him had been Irwin Biel, asking him not to judge in advance. Before the exam Biel took me aside and asked if I was aware that I had a hostile major professor. I assured him I knew that. He then said that he would not tell me what he himself was going to ask, but would advise me to make my answers as concise and short as possible. Rossby began the “defense of thesis” by saying that since none of the committee had read the thesis (not true), the examination would be a general exam on the entire field of meteorology. He then leaned back as he always did when he was about to ask a blockbuster. At that point, Biel said “Carl, since Bryson is in my field of meteorology, could I be the first to question him?” Rossby said “Of course, Irwin.” Biel reached in his pocket and took out a sheaf of papers—prepared questions. Every so often as we fired questions and answers at each other Biel would ask Rossby if he knew that or comment that the answer wasn't even in the literature. An hour and a half later Biel folded his papers and said “One hundred questions, one hundred correct answers.” Palmén passed, Fultz asked a simple question, as did Byers, and then Rossby asked his blockbuster. I simply answered that I did not know how to answer. (It is likely that Rossby did not either). I was sent out and recalled in two minutes to hear that I had passed. Thank you forever, Irwin Biel. —Reid Bryson on Carl Rossby and Irwin Biel (Bryson 2000)
In early 1942, the rain gauge at the Univ. of Chicago was located at ground level. If the record shows an all-time record rainfall in the summer of 1942 at that station, don't believe it. The cadets relieved themselves there one night after a long beer-drinking session. The gauge ran over! —Reid Bryson (on the role of anthropogenic forcing?; Bryson 2000)
When I left the Air Force and moved to Chicago to be one of Herbert Riehl's graduate students I was a little concerned if we would be able to get along. I had previously been in a few of Riehl's classes as an Air Force student in 1953–54 and knew he had a temper and could be difficult at times. Chuck Jordan (former Riehl Ph.D. student) was just passing through Chicago at that time and I asked Chuck about working with Riehl. Chuck smiled and told me “Oh don't worry about Herbert, he has mellowed a great deal in recent years—for instance, he no longer throws chairs at his graduate students.” —Bill Gray and Charles Jordan on Herbert Riehl (W. Gray 2013, personal communication)
Bill Gray used to take our completed Ph.D. dissertation, put a cover and report number on it, and mail it out as a technical report to every US academic he could think of. At the time Joanne Simpson was suddenly taken ill and had to spend a week in hospital. Leaving suddenly she grabbed the closest reading matter available on her desk. It was my Ph.D. having arrived in today's mail. Joanne lay in the hospital for several days with my Ph.D. and nothing else to read. Decades later she would see me at a meeting and refer to “your damn Ph.D. dissertation.” —John McBride on Bill Gray and Joanne Simpson (J. McBride 2013, personal communication)
Our student office was located one hallway down from Dr. Gray's office at CSU [Colorado State University]. Dr. Gray's daily routine was to work from home in the morning and come into the office in the afternoon. At least once a week we would hear Dr. Gray running down the hall, keys jangling in his pockets toward our office. His question invariably was: “Show me something new and sexy today!' — Chris Landsea on Bill Gray (C. Landsea 2013, personal communication)
The story about Vic Ooyama's thesis was well known in the Department [New York University]. Vic went into [Bernard] Haurwitz's office to discuss the thesis and Vic became so disgusted with the conversation that he threw the draft in the waste paper basket. After Vic left the office, Bernie picked the draft out of the trash can and had his secretary type it in final form and Vic received his Ph.D. in record time (I believe it was one year after he received his MS).4 —Dick Greenfield on Vic Ooyama and Bernard Haurwitz (R. Greenfield 2012, personal communication)
I was a graduate intern at summer school at NCAR [National Center for Atmospheric Research] in 1976, working under Ed Zipser on GATE [Global Atmospheric Research Program's Atlantic Tropical Experiment] data. I wanted to look at some of the radiosonde data and he told me to go to Vic Ooyama (who was working there at the time). I did not then know Vic or had even heard of him. I went to his office and asked if I could analyze some of the radiosonde data. He gave a sharp and highly sardonic laugh and then told me, in a somewhat hostile way, that he had been working on fixing all the problems with that data for the last 6 months and now some kid just wants to grab it and analyze it. I can't remember whether I ever got that data or not. —Kerry Emanuel on Vic Ooyama (K. Emanuel 2013, personal communication)
The authors welcome additional anecdotes through the contact information given below, and plan to comprise a web page dedicated to their display.
FUTURE EXPANSION: ALL OF METEOROLOGY.
Immediately after beginning this project, it became apparent that titling the family tree as for “tropical meteorology” was inherently fraudulent. The roots of the tree span numerous fields (e.g., mathematics, physics, and physiology) and the community is growing rapidly through “parents” originating outside the conventional tropical designation. Further, many scientists originating in the tropical community have diversified their work. Readers will undoubtedly note that many of the early scientists in the tree (e.g., Fig. 5) would also form similar roots for a family tree on midlatitude/baroclinic meteorology. In short, this genealogy was never really a purely tropical one. It is thus proposed that this genealogy be expanded to all of meteorology, and the authors welcome submissions (either for the reader himself or herself, or for colleagues) from the broader meteorology community (send to email@example.com).
The current version of the tree is available online (http://moe.met.fsu.edu/familytree). Future versions resulting from the broadening to all of meteorology (and for correcting inevitable errors) will be available at the same location. Caution is advised when examining electronic future versions of the tree, however. Many of the figures shown in this paper are going to dramatically change as additional parents and their children are added as a consequence of this expansion, as well as the natural procreation of children from parents already identified. It is expected that in the very near future, the entire tree will become too big to print (or perhaps simply too large to hang in a building hallway of typical 2.5–3-m height). The authors hope that very large screen liquid crystal display (LCD) monitors (e.g., 3–4 m or larger) will become reasonably priced in the near future, such that an interactive, searchable, zoomable, and perhaps instantly modifiable tree can be readily displayed at all upcoming conferences.5
It is not possible to acknowledge individually the hundreds if not thousands of scientists who have submitted information as part of the family tree. Indeed, we extend gratitude to each and every scientist represented here. Particular thanks are given to the following individuals for their invaluable work on this project: Kelly Hirai of The Florida State University for his work on the postprocessing modification and printing of the tree; Mitch Keller of the Mathematics Genealogy Project for his advice and examples for graphical display of family tree information; Carol Wolf of University of California, Berkeley, for her investigative work on Hans Panofsky's lineage from original thesis hard copies; Darlene Oosterhof for T. N. Krishnamurti's student record; Amie Hedstrom and Phil Klotzbach for Bill Gray's student record and Riehl information; Robert Houze, Peter Webster, and Ed Zipser for their respective student records; T. N. Krishnamurti, Howard Friedman, Robert Rogers, Dick Greenfeld and Yoko Ooyama for Vic Ooyama and Bernard Haurwitz's lineages; and the following for providing information or filling numerous gaps in the family tree: John Dutton, Jenni Evans, Bill Frank, Richard Johnson, T. N. Krishnamurti, Gabriel Lau, Frank Marks, Sharon Nicholson, Abraham Oort, Jean Phillips, Robert Rogers, Christine Sherratt, Gabe Vecchi, and Ed Zipser. Several articles served as a wealth of information for filling additional gaps in the tree, including, but not limited to: Lewis (1992, 1993), Lewis et al. (2012), Reiter (1998), and Rittner (2003). We also profusely thank the Department of Atmospheric Science at CSU for their 50th anniversary web page as a valuable source of data for this project. The authors are grateful to those cited therein, those who provided photographs, and others who provided anecdotes. The authors thank Jorge Cham of www.phdcomics.com for the use of that website's cartoon. The manuscript dedication was partially inspired by comments made by Jon Merritt of The Pennsylvania State University during the first author's first week of college. The authors thank J. McBride, C. Landsea, N. Shay, and J. Walker for their enthusiastic support in encouraging this project. The AMS, in particular Marge Huntington and Jennifer Ives, has been generous with its flexibility in displaying a poster of unusual size at the two conferences stated earlier. Finally, the authors appreciate the critical suggestions provided by three reviewers.
1 This is actually not necessarily true, as some colleagues requested their nonacademic mentors be listed as the advisor or a coadvisor for their placement in the tree.
2 The year of the most senior degree was the most commonly missing piece of information, which is critical, as it can significantly influence the vertical spacing of all progeny on the tree (and by extension the horizontal location of colleagues). If one advisor's group of students is missing graduation year information, then they are lined up together horizontally, creating unappealing vertical white space.
3 While magnifying glasses were provided by the authors for those needing visual aid to read the immense detail of the tree, it was noted that some senior colleagues did not wish to be caught in a photograph using the magnifying glasses for fear of later exploitation at retirement celebrations (or roasts).
4 At the first oral presentation of this work at the 93rd Annual AMS Meeting in Austin, Texas (Hart and Cossuth 2013), it was noted to students in the audience at this point that it would be unwise for them, perhaps hoping to accelerate their graduation, to emulate Ooyama in this one regard.
5 The authors are optimistic that existing AMS conference registration pricing can absorb this additional cost without any increase.