Norman A. Phillips: Institutional Arc

Research notes on the four institutions that defined Norman Phillips’s career: the University of Chicago (1940-1951, with a war interruption), the Institute for Advanced Study at Princeton (August 1951 - April 1956), the MIT Department of Meteorology (summer 1956 - July 1974), and the National Meteorological Center at Suitland and Camp Springs (July 1974 - 1984/1988). Also covers his retirement years (1984/1988 to 2019).

Cross-link with research/people/Phillips.md for biographical detail and research/institutions/MIT_meteorology_Lorenz_era.md for the broader MIT context.

I. The University of Chicago, 1940-1951

The pre-war department

Norman Phillips enrolled at the University of Chicago in autumn 1940, intending to major in chemistry. He was seventeen. Pearl Harbor was fourteen months away, and the Department of Meteorology that would later define his life was not yet a department – it was a meteorology section inside Horace Byers’s older Department of Geophysical Sciences, and Carl-Gustaf Rossby had only just arrived in 1940 to chair it. The independent Department of Meteorology was constituted in 1942, with Rossby as its founding chair.

What made Chicago meteorology different from other US programmes by the mid-1940s was that Rossby had brought a particular synthesis with him. He had been at MIT (1928 onwards), at Caltech briefly, at the Weather Bureau in Washington, and at Woods Hole. He carried the Bergen School’s frontal concepts in his head but distrusted Bergen-style descriptive meteorology as the foundation of the science. Atmospheric motion, in his view, was a problem in fluid dynamics on a rotating sphere, and large-scale dynamics could be reduced to a small number of mathematical relations – most famously the planetary wave equation (1939) that bears his name. The Chicago programme he built between 1940 and 1947 was the institutional embodiment of that conviction.

Phillips’s interrupted undergraduate years (1940-1946)

The war removed Phillips from Chicago for four years. He enlisted on 15 March 1942 in the “B” school at the University of Michigan, then went through the cadet “A” school at Chanute Field, Illinois, graduating fourth in a class of 391 on 5 June 1944 (the day before D-Day). Commissioned a second lieutenant in the Army Air Force, he was assigned to the 8th Weather Region at Grenier Field outside Manchester, New Hampshire, and then deployed to the Azores. There he ran briefings for transatlantic flights crossing the entire ocean, learned to draw weather maps under operational pressure (and survived the embarrassment of forecasts that ended badly – a plane he routed to Newfoundland with a marginal fuel margin he was “not sure” could make it to Stephenville), and absorbed the Bergen frontal models from the inside. He was discharged in August 1946 as a first lieutenant and returned to Chicago that autumn, now an undergraduate in meteorology rather than chemistry.

Rossby’s departure and Byers’s chair (1947-1951)

By the time Phillips received his BSc in spring 1947, Rossby had already begun the long withdrawal that would take him back to Stockholm. In 1947 Rossby was named founding director of the new Meteorological Institute (MISU) at Stockholm University, and from that point his role at Chicago became one of frequent but no longer continuous visits. The formal succession in the Chicago department fell to Horace R. Byers, who served as chair from 1948 through 1960 – the entire span of Phillips’s graduate study and well beyond.

This matters for the texture of Phillips’s graduate experience. Phillips never described Byers as a teacher in his oral history; the dynamical figures he named were Rossby (when in residence), Erik Palmen (visiting from Helsinki 1946-1948 and recurrent thereafter), Herbert Riehl (assistant professor by 1948, who took over the ONR-funded jet stream project after Palmen’s return to Finland), Hsiao-Lan Kuo (mathematical theorist), and George Platzman (a Chicago meteorology PhD himself, class of 1947, and a junior faculty member by 1948). Phillips wrote his master’s thesis – a synoptic study of the subsidence of a cold dome – “under” Palmen, though Palmen’s pattern of long visits rather than full residence meant the supervision was looser than in conventional departments.

For the doctoral work, Phillips’s advisor of record is conventionally said to be George Platzman, and Lewis’s 1998 BAMS reconstruction confirms this: a footnote credits Platzman, along with Charney and Rossby, for help with the mathematical derivation of two-layer baroclinic instability. The dissertation itself, however, Phillips chose entirely on his own initiative. He had read Charney’s 1948 Geofysiske Publikasjoner paper “On the Scale of Atmospheric Motions” and decided that a simple two-layer model could capture extratropical cyclogenesis with a minimum of vertical structure. In his own retrospective: “I had seen Rossby use such kind of models in his lectures.” The thesis – A Simple Three-Dimensional Model for the Study of Large-Scale Extratropical Flow Patterns (1951) – ran to twenty-five pages and contained both a derivation of the baroclinic instability criterion (which Phillips later discovered Eric Eady had derived independently the year before) and a hand-computed forecast of the Thanksgiving Day 1950 storm using Southwell’s iterative relaxation method on a quasi-geostrophic two-layer model. It was the first numerical baroclinic forecast.

The Chicago intellectual atmosphere

The Chicago meteorology PhD programme in the late 1940s was small, hand-picked, and intensely theoretical. The “Chicago school” name – coined retroactively – captured several distinct features:

Daily map briefings led by George Cressman, drawing synoptic charts across 240 degrees of longitude (mid-Pacific to the Urals), the largest hemispheric synthesis any university programme attempted. The briefings turned into open arguments over the existence and structure of the jet stream, the role of meridional circulation in maintaining the westerlies, and the merits of Bergen frontal analysis versus Chicago wave dynamics. Phillips attended these as a graduate student and was “exposed to much of the controversy on general circulation theory” through them.
The ONR jet stream project. After WWII Rossby had secured Office of Naval Research funding for a comprehensive general circulation study, with Palmen as the founding scientific leader and Riehl taking over after Palmen returned to Helsinki in 1948. As a research assistant under Kuo on Rossby’s ONR contract, Phillips was tasked with drawing whole-winter cross-sections at the 80th meridian. His earliest hand-drawn sections were sketchy enough to make Rossby visibly shudder; Phillips quickly learned to draw them prettier. The work taught him both the observational reality of the jet (more complicated than the simple single-stream picture) and the politics of working under Rossby (publish quickly to satisfy ONR).
A small cohort of contemporaries who would dominate post-war NWP. George Platzman (PhD 1947, joined faculty 1948), Chester Newton (graduate student under Palmen and longtime collaborator), Riehl (PhD 1947, faculty), Phillips himself (PhD 1951). Joseph Smagorinsky is sometimes assumed to have been a Chicago classmate but in fact came through NYU under Bernhard Haurwitz; the link to the Chicago/Princeton axis came through Charney rather than direct co-residence. Aksel Wiin-Nielsen, soon to be a key figure in Stockholm and Copenhagen, was also passing through.
Charney’s visits. In late 1950 or early 1951 Charney passed through Chicago, and Phillips presented his dissertation work on the blackboard to Rossby, Charney, and Platzman – the encounter that secured Phillips an offer to join the IAS as soon as he defended. Charney’s later recollection in the Lindzen-Lorenz-Platzman Atmosphere volume: “Phillips just outlined very briefly on the blackboard what he had done. And what he had done, of course, was to develop a two-layer model… I recognized that right away as the next step. I hadn’t proposed to do it in that way, but it turned out that the finite difference formulation in the simplest way, by vertical differencing of the quasi-geostrophic equations, one arrives at the same equations as one gets for a two-layer model.” That recognition was a hiring decision in real time.

Open question: did Phillips know Rossby personally?

The answer from Phillips’s own oral history is: yes, but at a distance. Phillips arrived back at Chicago as an undergraduate in late 1946, just as Rossby was beginning his long-distance arrangement with Stockholm. The two overlapped most concretely from autumn 1947 (when Phillips entered the graduate programme) through about 1949-1950, with Rossby in residence enough to give the lectures Phillips remembered and to react personally to Phillips’s first sketchy ONR cross-sections. Phillips later said he had attended Rossby’s lectures and seen Rossby use the kind of two-layer construction that ultimately became his thesis model. After 1950 the relationship was more episodic, but it was Rossby who arranged the invitation that took Phillips to Stockholm in 1953 – evidence that the mentor relationship persisted long after Rossby’s centre of gravity had shifted to Stockholm.

II. The Institute for Advanced Study Meteorology Project, August 1951 - April 1956

Origins and Charney’s leadership

The Meteorology Project at the IAS in Princeton was officially launched on 1 July 1946 as a subsidiary of John von Neumann’s Electronic Computer Project (ECP), funded jointly by the US Navy’s Office of Naval Research, the US Atomic Energy Commission, and (later, more episodically) the US Weather Bureau. Von Neumann’s August 1946 organisational meeting at the Institute – the meeting that put Charney on track to leave Chicago for Princeton – had been suggested to von Neumann by Rossby himself, according to Frederik Nebeker’s archival reconstruction.

After a fitful start staffed by theoreticians like J. Freeman, G. Hunt, P. Queney, and Philip Thompson, Charney came to head the Meteorology Project formally in early 1948 (after his 1947-1948 Oslo fellowship year), at age 31. Eliassen and Fjortoft came over from Oslo in succession to anchor the dynamics side; Smagorinsky came from NYU in 1950 as a doctoral student looking for a problem.

The ENIAC integration and its aftermath

The April 1950 ENIAC integration at Aberdeen Proving Ground – the celebrated Charney-Fjortoft-von Neumann paper – predates Phillips’s arrival, but it is the experiment that made the Meteorology Project famous and provided the operational template that everything Phillips did at the IAS extended. (See Post 1 of the NWP blog series for the full story.) The IAS’s own computer, the MANIAC I (sometimes called the Institute machine, built to von Neumann’s specifications by Bigelow and the engineering team), was still under construction; the ENIAC was a borrowed instrument.

Phillips arrives, August 1951

Phillips defended his Chicago PhD in summer 1951 and joined the IAS that August. His IAS scholar record gives the formal dates as October 1951 - April 1956, but his oral history places his arrival a month or two earlier, with formal membership beginning at the start of the IAS academic year.

The MANIAC was nearly operational. By late 1951 the machine was running tests, and Phillips’s first IAS assignment was to write the program for a barotropic model – a direct continuation of the ENIAC work. He soon extended it to the two-level baroclinic version reported in Charney and Phillips (1953), “Numerical integration of the quasi-geostrophic equations for barotropic and simple baroclinic flow,” Journal of Meteorology 10. The test case for both models was the same Thanksgiving Day 1950 storm Phillips had used in his Chicago dissertation – a deliberate continuity. Nebeker’s archival work documents that the IAS baroclinic forecasts were “widely seen as evidence that numerical forecasting could outperform subjective forecasting.”

Who was at the IAS while Phillips was there?

The Smagorinsky photograph from 1952 (reproduced in Lewis’s 1998 BAMS retrospective) captures the core team at the MANIAC: Jule Charney, Norman Phillips, Glenn Lewis, Norma Gilbarg, and George Platzman. Smagorinsky himself was behind the camera. The longer roster across Phillips’s tenure (1951-1956) included:

Jule Charney – scientific leader of the Project, 1948-1956.
Norman Phillips – research staff, August 1951 - April 1956.
George Platzman – repeat visitor; primary appointment at Chicago.
Joseph Smagorinsky – 1950-1953, then to the US Weather Bureau in 1953 (age 29) to seed what would become the Joint Numerical Weather Prediction Unit and later GFDL.
Glenn Lewis – computational staff.
Norma Gilbarg – mathematician and programmer; left in 1953.
Ragnar Fjortoft – Norwegian, on a one-year arrangement that overlapped with Eliassen’s; both were part of Charney’s Norwegian rotation.
Arnt Eliassen – Norwegian, the year before Phillips arrived; back as visitor.
Bert Bolin – spent a year at the IAS (and Chicago) in 1950-1951 before returning to Stockholm; he and Phillips first met during this overlap, beginning a friendship that would matter for Phillips’s 1953 Stockholm visit.
Yale Mintz – not formally on the IAS staff; based at UCLA but a regular visitor and collaborator on general circulation work.
John von Neumann – not a meteorologist but actively engaged through the entire period; named AEC commissioner in 1955, which began the unraveling of the Project.
Philip Thompson – early Princeton team, departed before Phillips arrived.

The Stockholm and Oslo interlude, August 1953 - April 1954

In 1953 Rossby sent Phillips a personal invitation to spend time at the Stockholm Institute. The pretext was operational: the Swedes had built BESK (Binär Elektronisk SekvensKalkylator), modelled on the Princeton IAS computer, and wanted help getting NWP code running on it. Phillips arrived in August 1953 and stayed six months, with two further months at the Geophysical Institute in Oslo on the way back – the precise period when Eliassen had just been appointed Lecturer at the University of Oslo. The Stockholm experiment included Phillips, the Swedish team (Bolin, Wiin-Nielsen, Bert Doos, Germund Dahlquist), and the Icelander Gudmundur Arnason, and on 23-24 March 1954 it produced what Wiin-Nielsen calls “the first forecasts finished in time to be of use in operation” – preceding the JNWPU’s own operational forecasts in the US by more than a year.

Phillips also used the Stockholm year to begin thinking about general circulation modelling rather than just short-range forecasting. Back at the IAS in April 1954, the 1956 general circulation experiment was already on the agenda.

The Project closes

By 1955 the structural pressures on the Meteorology Project were converging. Von Neumann had become an Atomic Energy Commissioner and was effectively absent in Washington. He was diagnosed with cancer in late 1955 (died February 1957). Oppenheimer, as Institute director, was unable to promise the meteorology group a permanent home; “pure mathematics was the preferred science there,” in Phillips’s own later phrasing in his 1995 Charney memoir. Charney, now thirty-eight and with a five-year IAS appointment expiring in 1956, began looking elsewhere.

Charney accepted MIT’s offer in summer 1956. The package included Phillips and the retention at MIT of Edward Lorenz. Smagorinsky had already left in 1953 for what became (1955) the US Weather Bureau’s General Circulation Research Section in Suitland – which would migrate via DC (1959 rename to General Circulation Research Laboratory) and finally to Princeton (1968) as the Geophysical Fluid Dynamics Laboratory. Manabe joined Smagorinsky in 1958.

After Charney and Phillips’s departures the IAS Meteorology Project effectively wound down. The MANIAC continued at the Institute for some years for other projects but no longer for meteorology; the meteorology mission had migrated to JNWPU/NMC (operational) and GFDL (research). The closure was institutional rather than ceremonial – no formal end date was announced.

III. MIT Department of Meteorology, summer 1956 - July 1974

The Houghton department

When Charney and Phillips arrived at MIT in summer 1956, the Department of Meteorology was already the strongest in the US. The institutional facts:

Founded 1941 as Course 14 in the School of Engineering, the first independent meteorology department in the United States. Renamed Course 19 in 1946.
Departmental heads: Sverre Petterssen (1941-1942), Henry Garrett Houghton (1942 acting, 1945 confirmed - 1969), Norman A. Phillips (1970-1974), Jule Charney (1974-1977), Edward Lorenz (1977-1981), Peter Stone (1981-1983).
Moved from the School of Engineering to the School of Science in 1957, just after Phillips arrived.

Houghton had become assistant professor in 1939, executive officer in 1942, professor and head in 1945, and would chair the department until his retirement in 1969 (some sources say 1970; the AMS Houghton Award page gives 1970). His own scientific work was on physical (rather than dynamical) meteorology – cloud microphysics, fog dissipation, radiative transfer – but he was institutionally an empire-builder, and the 1956 hiring of Charney and Phillips was the deal that made the department what it became. Houghton was first chairman of the board of the new University Corporation for Atmospheric Research (1959).

Faculty during Phillips’s MIT years (1956-1974)

Drawing on the MIT PAOC departmental history and the period’s recollections:

Faculty member	MIT tenure	Notes
Henry G. Houghton	1939-1969 (chair 1942-1969)	Cloud physics, radiation; founder UCAR chair
Victor P. Starr	1947-1976	General circulation statistics, eddy fluxes
Edward N. Lorenz	1948-2008	Chair 1977-1981; chaos 1961
Frederick Sanders	1954-1984	Synoptic; mentored Lorenz students
Jule G. Charney	1956-1981	Chair 1974-1977; theoretical leadership
Norman A. Phillips	1956-1974	Chair 1970-1974
Willem V. R. Malkus	1959-1983	Geophysical fluid dynamics seminars
Reginald E. Newell	1961-2002	Started as assistant professor 1961; later department head briefly
Erik Mollo-Christensen	1955-1984	Fluid dynamics, turbulence
Henry Stommel	1963-1992	Joint MIT-WHOI; physical oceanography

(The MIT history page also lists Houghton’s chair as 1942-1969, while the AMS Houghton Award page says 1945-1970. Both could be correct under different definitions – acting vs confirmed, retirement announced vs effective. The 1970 chair transition to Phillips is firm.)

Buildings and the move to the Green Building

The meteorology department’s home during Phillips’s first years was Building 24 – a low brick structure on the MIT main campus along Massachusetts Avenue. Lorenz’s January 1961 discovery of sensitive dependence on initial conditions happened in Building 24 on the Royal McBee LGP-30. The department’s heroic post-war years – Houghton’s recruitment of Starr in 1947, the arrival of Charney and Phillips in 1956, Reginald Newell as assistant professor in 1961 – all happened in Building 24.

The Cecil and Ida Green Building (Building 54), designed by I. M. Pei, was constructed 1962-1964 of reinforced concrete and at 295 feet (90m) was, until 2019, the tallest building in Cambridge. The department moved in in 1964. Phillips’s MIT chair (1970-1974) was conducted from the Green Building, as was Lorenz’s office for the rest of his life. The chaos discovery (1961) thus belongs to Building 24; the GCM work, Phillips’s chairmanship, and most of Lorenz’s later career belong to the Green Building.

Phillips’s MIT scientific output

Phillips’s MIT period produced his most theoretically significant work after the 1956 GCM. The discovery of nonlinear computational instability (NCI, or “aliasing”) was published in the 1959 Rossby Memorial Volume but the analysis was performed at MIT in 1957-1958, and the corollary – the necessity of de-aliasing transformations – shaped every subsequent numerical model. The sigma-coordinate formulation, in which the vertical coordinate is pressure normalised by surface pressure, appeared in 1957 (Journal of Meteorology 14: 184-185); it became the standard vertical coordinate for primitive-equation models worldwide.

He held a Sloan Fellowship in 1961 (an Alfred P. Sloan Foundation Research Fellowship, then a newish programme; not the 1956 award sometimes cited in retrospectives). The Rossby Medal – the AMS’s highest – came in 1971, during the year before he became department head. The 1956 GCM experiment had won the first Napier Shaw Memorial Prize in 1956, awarded by the Royal Meteorological Society on the centenary of Shaw’s birth.

The chairmanship, 1970-1974

Houghton’s retirement in 1969 (or 1970, depending on source) created the chair vacancy. Phillips became head in 1970. Charney, the obvious internal candidate, was at this point deeply committed to GARP – the Global Atmospheric Research Programme – and had been head of its NAS Committee since 1966. He was not available for administrative work. Phillips was the natural alternative: senior enough, sufficiently established scientifically (Rossby Medal in 1971), and willing to do the work.

Why did Phillips leave MIT in 1974?

This is the question the obituaries handle most thinly. Phillips’s own oral history records the move pragmatically: he was approached by the NMC, the offer was attractive, and the operational problem he wanted to work on – data assimilation as a coherent statistical-dynamical inverse problem – was best pursued from inside an operational centre rather than from a university chair. The MIT department was about to be handed over to Charney (who served 1974-1977), and it was an opportune moment to leave.

There is no documented evidence of internal MIT politics driving the departure. The most reasonable reconstruction is that Phillips had served his chair term, the department was in good hands, and the operational pull of NMC was real. He was 51; the chair work had taken him away from research for four years; the data assimilation problem was technical and tractable; and the NMC operational platform offered both compute resources and routine global data streams that no university could match. His oral history makes the operational pull clear: he wanted to be where the real-time data lived.

The MIT chair passed to Charney (1974-1977), then to Lorenz (1977-1981).

IV. NMC at Suitland and Camp Springs, July 1974 - 1984/1988

The NMC of 1974

When Phillips joined NMC in July 1974, the organisation was sixteen years old (founded January 1958 from the merger of the Joint Numerical Weather Prediction Unit and the National Weather Analysis Center) and entering one of its most ambitious transition periods.

Directors during Phillips’s NMC tenure:

Director	Years	Notes
George P. Cressman	1954-1964	First director; left to be Chief of Weather Bureau
Frederick G. Shuman	April 1964 - January 1981	Started as Chief, Development Division 1958
William Bonner	1981-1990
(James Howcroft)	1990	Acting
Ronald D. McPherson	1990-1998	Renamed NMC -> NCEP in 1995

Phillips overlapped fully with Shuman (1974-1981) and partly with Bonner (1981-1988). His role was Principal Scientist of the Development Division, the same Development Division that Shuman himself had founded and chaired before becoming director. The Development Division was the in-house research arm: it built and maintained the models that the operational shifts ran. Phillips was the senior scientific voice in that group – not its administrator (who was, at various points, Joseph Sela, John Brown, and others) but its theoretical anchor.

Suitland, Camp Springs, and the World Weather Building

Phillips arrived at NMC’s original location: Federal Office Building 4 (FOB-4), Suitland, Maryland. NMC had been there since its 1958 founding. The decision to move had been made before Phillips arrived: a new building had been authorised, sited in Camp Springs about two miles south of FOB-4, and dedicated on 22 October 1974.

The full NMC move began in early 1975. By press accounts of the day, “in January 1975, NMC (except for the supercomputer facility) moved to the World Weather Building,” at 5200 Old Auth Road, Camp Springs, Maryland, occupying the second, third, fourth, and sixth floors. The dedicated date in NWS heritage materials is 19 February 1975 for the forecaster move. The computers were too large to move; they stayed at FOB-4 in Suitland, with output transmitted electronically two miles to Camp Springs. (FOB-4 housed NMC’s mainframes until 1999, when the IBM SP was installed at a new site in Bowie.)

This split site – forecasters and developers at the World Weather Building, computers at Suitland – was the operating arrangement for all of Phillips’s NMC years and beyond. Phillips’s office was at the World Weather Building in Camp Springs. His residence was in Camp Springs.

The operational landscape Phillips inherited

When Phillips arrived in July 1974, NMC was running:

The Limited Fine Mesh (LFM) regional model for North America (operational since 1971), an early primitive-equation grid-point model.
A 6-layer primitive-equation grid-point global model (the “PE model”), in use through the 1970s. The grid-point design was about to be superseded.
Limited objective analysis using successive corrections (Cressman’s method, refined since 1959).

The big imminent transitions:

The spectral GFS – the Global Forecast System – went operational in 1980 under Sela’s leadership in the Development Division, transitioning NMC’s global model from finite-difference grid-point to spherical-harmonic spectral representation. Bourke’s Australian work (1972) was the methodological template; ECMWF was moving spectral around the same time.
Operational optimum interpolation (OI) data assimilation. NMC’s hand-tuned successive-corrections analysis was being replaced by a coherent statistical scheme. The OI literature was European (Bratseth, Lorenc, Hollingsworth at ECMWF). Phillips’s role at NMC during the late 1970s and 1980s was as the senior theoretical voice making sure the OI implementation respected the statistical structure of geostrophic adjustment – his 1986 Tellus A paper “The spatial statistics of random geostrophic modes and first-guess errors” was the most cited contribution of his NMC years and is the bridge between the 1956 GCM Phillips and the operational data-assimilation Phillips. (See research/concepts/four-dimensional-data-assimilation.md for cross-reference once Post 44 is researched.)
The Nested Grid Model (NGM), Phillips’s signature operational model at NMC. The NGM was a stereographic-projection hemispheric primitive-equation model with optional one or two nested fine-mesh inner grids over North America, displacing the LFM as the regional model. The model went operational in 1987. By the time Phillips retired from NMC – formally in 1984, though he continued in consulting capacity for some years, with some sources citing a 1988 effective retirement – the NGM was known in the NMC corridors as “Norm’s Great Model”.

Phillips’s role as operational-scientific bridge

Phillips’s transition from the theoretical academic of the 1956 GCM and 1959 NCI papers to the operational scientist of the 1980s NMC was distinctive in the field. Most of the IAS/Princeton cohort had gone the other way: Smagorinsky to GFDL, Charney to MIT’s theoretical seminars, Mintz at UCLA, Manabe at GFDL Princeton. Phillips alone reversed direction in mid-career, taking the operational job at age 51 and putting his theoretical authority in service of the daily forecast suite.

The professional consequence was that NMC’s data-assimilation transition through the late 1970s and 1980s – the move from successive corrections to OI to the spectral statistical interpolation (SSI) of the early 1990s – was scientifically anchored by a man who had been at Charney’s elbow in 1951 and who had personally written the first numerical baroclinic forecast in 1951. That continuity of authority is rare. It is the reason Phillips’s name appears repeatedly in the NMC technical reports of the 1980s on the foundations of the analysis system.

When did Phillips retire?

The sources disagree.

The Rivet Funeral Home obituary (the primary obituary source) says “1984” for his retirement from NMC. The same obituary says he then retired to Merrimack, New Hampshire.
The internal Phillips research note in this repository (research/people/Phillips.md) records 1988 as the end of the principal-scientist appointment, citing the MIT News obituary’s “stayed four years before retiring in 1988.”
The MIT News obituary itself reads: “He served as principal scientist of the NMC Development Division until his retirement in 1988.” – placing the formal retirement at 1988, not 1984.

The most plausible reconciliation: Phillips formally retired from his full-time NMC appointment in 1984 (the obituary’s “retirement” year, which fits with a 65th birthday in July 1988 also) but continued in some research-emeritus or consulting capacity through 1988, and the MIT obituary collapsed those into the single year 1988. His 1986 Tellus A paper carries the NMC affiliation, supporting the view that he was still meaningfully active at NMC through the mid-1980s. The Nested Grid Model becoming operational in 1987 with the “Norm’s Great Model” nickname suggests Phillips was still scientifically present at NMC at least through the implementation date.

Open question: An accurate retirement date will require either the NWS personnel file (FOIA-able) or a clearer statement from Phillips’s own correspondence than has been surfaced in published sources. For purposes of the blog series, treating 1984 as the formal retirement and 1988 as the effective end of NMC engagement is defensible.

V. Retirement years, 1984/1988 - 15 March 2019

Where Phillips lived

The retirement note in the user’s prompt says “Falmouth, Maine.” This is not supported by the obituaries. The Rivet Funeral Home obituary places retirement residence in Merrimack, New Hampshire. The death location was Grace House in Windham, New Hampshire (15 March 2019). The MIT News obituary corroborates Windham. No published obituary places Phillips in Falmouth, Maine. The Falmouth Maine attribution in the user’s research notes may be a confusion with another Norman Phillips (an obituary search returns multiple Norman Phillipses), or may derive from an unrelated AMS service address. Treating Phillips as a New Hampshire retiree (Merrimack, then Windham) is what the evidence supports.

AMS service and the Phillips Award

The user’s prompt asks about an AMS award named for Phillips. There is no AMS “Phillips Award” as a standing prize on the AMS scientific awards roster (the closest are the Carl-Gustaf Rossby Research Medal, which Phillips received in 1971, and the Henry G. Houghton Award for early-career meteorologists, named for Phillips’s MIT chair). It is possible that the prompt is conflating Phillips’s role as honoured recipient with a hypothetical eponymous award; this should be flagged as unverified.

Phillips’s AMS service included: Carl-Gustaf Rossby Research Medal (1971), Meisinger Award, Editor’s Award, Award for Outstanding Contribution to Applied Meteorology, distinguished lecturer designation, honorary member. He gave the 6th WMO Lecture. He was elected to the National Academy of Sciences in 1976.

Memorial writing

Phillips wrote the NAS biographical memoir for Jule Charney, published in Biographical Memoirs Volume 66 (1995), pages 80-113. The memoir is the most detailed Charney biography in print and has been heavily mined by subsequent historians (Nebeker, Edwards, Lewis, Lynch). Phillips drew on his personal experience as Charney’s IAS colleague, his Chicago graduate-school overlap, and decades of correspondence. The memoir is a primary source for the Charney material in this blog series (Post 1 and Post 3).

Phillips published other shorter tributes for AMS journals over the decades but the Charney memoir is the major one.

Late publications

Phillips’s last academic paper was “What Makes the Foucault Pendulum Move among the Stars?” published in Science & Education 13:387-401 (2004), when he was 81 (not 90, as one obituary states; the age-90 detail may refer to a later short note or conference proceedings). The Foucault pendulum interest reflects his lifelong fascination with the rotational physics that underlies both meteorology and classical mechanics.

Death

Norman Phillips died at Grace House, Windham, New Hampshire on 15 March 2019, aged 95. The MIT News obituary (16 May 2019) and the Rivet Funeral Home obituary are the primary published sources. A memorial service was held on 29 June 2019 at the Rivet Funeral Home in Merrimack.

Source notes and open questions

Sources consulted

Primary

Phillips, N. A. (1989). Oral History Interview with Norman A. Phillips, conducted by Anthony Hollingsworth, Warren Washington, Joseph Tribbia, and Akira Kasahara, NCAR Archives. Available via OpenSky at NCAR. This is the principal Phillips primary source for the Chicago/IAS/Stockholm years.
Phillips, N. A. (1995). “Jule Gregory Charney, 1917-1981, a Biographical Memoir.” Biographical Memoirs Volume 66, National Academies Press, pp. 80-113. Available at https://www.nasonline.org/wp-content/uploads/2024/06/charney-jule-g.pdf. Phillips’s own writing on the IAS years.
Phillips, N. A. (1956). “The general circulation of the atmosphere: a numerical experiment.” Quarterly Journal of the Royal Meteorological Society 82: 123-164.
Phillips, N. A. (1951). “A Simple Three-Dimensional Model for the Study of Large-Scale Extratropical Flow Patterns.” University of Chicago doctoral dissertation, 25 pp.
Phillips, N. A. (1986). “The spatial statistics of random geostrophic modes and first-guess errors.” Tellus A 38A: 314-332.

Secondary

Lewis, J. M. (1998). “Clarifying the Dynamics of the General Circulation: Phillips’s 1956 Experiment.” Bulletin of the American Meteorological Society 79(1): 39-60. The fullest biographical reconstruction of Phillips’s path to the 1956 GCM, drawn heavily from the 1989 oral history.
“Norman Phillips, former meteorology department head, dies at 95.” MIT News, 16 May 2019. https://news.mit.edu/2019/norman-phillips-former-meteorology-department-dead-dies-0516
Rivet Funeral Home obituary (Merrimack, NH), 2019: https://www.rivetfuneralhome.com/obituary/norman-phillips
Lynch, P. (2019). “A Pioneer of Climate Modelling and Prediction.” ThatsMaths, 4 April 2019, and The Irish Times obituary.
Nebeker, F. (1995). Calculating the Weather: Meteorology in the 20th Century. Academic Press. Especially chapter 5 on the IAS Project and chapter 6 on the move to operational NWP.
Wiin-Nielsen, A. (1991). “The birth of numerical weather prediction.” Tellus B 43(4): 36-52. The Stockholm 1953-54 forecasts.
“Norman Phillips,” IAS Scholar profile: https://www.ias.edu/scholars/norman-phillips
MIT PAOC department history: http://paocweb.mit.edu/about/history
Wikipedia entry on Norman A. Phillips: https://en.wikipedia.org/wiki/Norman_A._Phillips
Shuman, F. G. (1989). “History of Numerical Weather Prediction at the National Meteorological Center.” Weather and Forecasting 4(3): 286-296. The NMC director’s own retrospective.
World Weather Building history, NWS Heritage Project: https://vlab.noaa.gov/web/nws-heritage/-/the-world-weather-building
NCEP history page: https://www.emc.ncep.noaa.gov/emc/pages/ourhistory.php
Mariners Weather Log Vol. 51 No. 3 (December 2007): NCEP directors retrospective. https://vos.noaa.gov/MWL/dec_07/legacy.shtml
Lindzen, R. S., Lorenz, E. N., and Platzman, G. W. (eds.) (1990). The Atmosphere – A Challenge: The Science of Jule Gregory Charney. American Meteorological Society. Contains Charney’s 1980 oral history with Platzman and Phillips’s own reminiscences.

Open questions and source disagreements

Exact retirement date from NMC: 1984 (Rivet obituary) vs 1988 (MIT News obituary). Most likely formal retirement at 1984 with continuing affiliation through 1988. Need NMC/NWS personnel records or Phillips’s own correspondence.
PhD advisor at Chicago: George Platzman (per Lewis 1998 footnote 6) is the most commonly cited, but it has been noted in passing in older sources as “self-supervised” or as “under Rossby” depending on the year. Platzman was a young assistant professor 1948-1951 and would have been Phillips’s de facto advisor for the mathematical work; Rossby was the senior personality of the department; Palmen had been Phillips’s master’s advisor. The Chicago dissertation footnote remains the best evidence.
Houghton chair end date: 1969 (MIT PAOC history) vs 1970 (AMS Houghton Award page). Possibly acting vs confirmed end. Phillips’s chair start is firmly 1970.
Retirement residence: The user’s prompt says “Falmouth, Maine”; all published obituaries say Merrimack, New Hampshire (retirement residence) and Windham, New Hampshire (death location). Falmouth Maine attribution should be corrected unless an unpublished source surfaces.
AMS Phillips Award namesake: No standing AMS prize bears Phillips’s name. The Henry G. Houghton Award is named for Phillips’s MIT chair. The user’s prompt may be conflating this; should be flagged.
Last paper at age 90 vs 81: The 2004 Foucault pendulum paper (Science & Education 13) puts Phillips at 81. The “age 90” formulation in some obituaries may refer to a later short note or proceedings contribution not yet identified.
IAS Meteorology Project formal closure date: The Project effectively ended with Charney and Phillips’s departure in 1956; no formal closure announcement is documented in the sources consulted. The MANIAC continued at IAS for other projects.
Phillips-Rossby relationship intensity: Phillips arrived at Chicago as an undergraduate in autumn 1946; Rossby moved his primary base to Stockholm in 1947 but visited often. Phillips’s master’s was nominally under Palmen rather than Rossby. The intensity of personal mentorship between Rossby and Phillips at Chicago is documented as real but episodic; Rossby’s 1953 invitation to Stockholm was the strongest single piece of evidence that the mentor relationship persisted.

Word count: approximately 5,800 words. Notes prepared 2026-05-17 for Phillips-focused NWP/HPC history blog post.

Michał Brennek