Meteorologist, oceanographer, Python solvem probler, experimenter, Raspberry Pi tinkerer, geek dad, manager, space enthusiast, ham radio lover, science columnist, ADHDer, synthetic intellect curious, linux admin cosplayer.

MIT Meteorology in the Lorenz Era: Institutional Context

Research notes for a long-form NWP-history blog post on Edward Lorenz, the MIT Department of Meteorology, and the wider community that received – and slowly absorbed – the chaos result.

Scope. This file covers institutions and the people who shaped Lorenz’s working environment. It deliberately complements the existing biographical files: research/people/Lorenz.md, Charney.md, Phillips.md, Emanuel.md, Henry_Stommel.md, and Rossby.md. Where those files already contain biographical detail, this document focuses on institutional structure, intellectual neighbourhood, and the reception of Lorenz’s predictability work in the operational NWP community.

Date assembled: 2026-05-15. Status: draft research notes, ~5000 words. Open questions flagged at the end.

1. The MIT Department of Meteorology: from a course to a department

1.1 Rossby’s founding (1928-1941)

The MIT meteorology programme was the first university programme of its kind in the United States. Carl-Gustaf Rossby – a young Swede who had come to the U.S. Weather Bureau in 1926 and then to MIT in 1928 – established what was initially a “Course” in meteorology within the Department of Aeronautical Engineering, with backing from the Daniel Guggenheim Fund for the Promotion of Aeronautics. The motivation was practical aviation safety: commercial aviation was just emerging, the 1925 Shenandoah airship had been lost in a thunderstorm, and trained meteorologists were needed.

Rossby ran the programme aggressively: by 1931 his group had purchased a Cessna monoplane and Professor Daniel Sayre was conducting daily upper-air flights from East Boston Airport, gathering upper-troposphere observations that were otherwise unavailable in North America. Through the 1930s the programme grew, eventually achieving departmental status. The official formation of the Department of Meteorology – the first such named department in the United States – is conventionally dated to 1941, although Rossby himself had already left for the University of Chicago in 1941 (1939 in the MIT News timeline, but most other accounts place his departure in 1941 when Chicago set up the parallel programme).

Source: MIT News, Wind, war and weathermen (2011); EAPS history page; Rossby NAS memoir.

1.2 The Willett-Houghton transition (1941-1945)

When Rossby left, the senior figure remaining was Hurd C. Willett, a long-range forecaster and sun-spot/climate enthusiast who had come to MIT with Rossby in 1928. Willett was a popular teacher and pioneer of extended-range weather forecasting (his “Willett curves” of solar activity and weather were widely used through the 1940s), but he was not an institution-builder. The wartime expansion – by 1944, 994 military officers had taken graduate-level meteorology training at MIT, including 25 women – required administrative leadership rather than scientific celebrity.

Henry G. Houghton, Jr. filled that role. Born 1905, MIT SM in electrical engineering 1927, he had been at MIT’s Round Hill Research Station 1928-1938 studying fog, cloud droplets, and atmospheric optics – effectively founding cloud physics as a discipline. He became associate professor and executive officer of the Department of Meteorology in 1942 and professor and Head in 1945, serving until 1970. (Some sources give 1942-1969; AMS sources say 1945-1970. Either way: roughly a 25-year reign that covered Lorenz’s entire formative period at MIT.)

Houghton was not flashy. His authored monograph Physical Meteorology (MIT Press, posthumously revised 1985) reads like a careful engineering text. But under his leadership – and crucially, his willingness to recruit theorists despite his own experimental bent – MIT meteorology became “preeminent.”

Source: Houghton AMS award page; MIT ArchivesSpace Houghton papers; Lifting the Fog (MIT Technology Review).

1.3 What made MIT distinctive

A useful contrast point: NCAR (founded 1960 under Walter Orr Roberts) and GFDL (Princeton, founded 1955 under Smagorinsky) were institution-building projects with explicit policy missions. MIT meteorology was the opposite: a small, departmentally autonomous group of theoretically minded researchers inside a private engineering university, with no operational forecasting mandate and no aspirations toward bigness. The culture was technically deep, modest, and somewhat indifferent to the politics that consumed Charney’s later GARP and CO2 Report work.

Lorenz’s biographer Kerry Emanuel captured this: “Those of us privileged to have known Ed Lorenz will remember him as a gentle, quiet soul, almost painfully shy and modest to a fault.” That description could be extended to the department itself. Phillips, Sanders, Newell, Mollo-Christensen, Stommel all fit a similar mould: serious, technically rigorous, uninterested in self-promotion.

2. Faculty in the Lorenz years (1948-1987)

2.1 Victor Starr and the General Circulation Project

Victor Paul Starr (1909-1976) was the formative scientific influence on Lorenz. Starr ran the MIT Planetary Circulation Project (sometimes called the General Circulation Project) from the late 1940s through to his death. The project’s goal was to understand the energetics of the global atmospheric circulation – a topic where Rossby had pointed the way but where systematic quantitative work was still needed. Starr’s contribution was the theory of negative viscosity: the recognition that mid-latitude eddies systematically pump westerly momentum into the jet stream, contrary to the usual diffusive intuition.

Lorenz worked under Starr from 1948 (the year of his ScD) until his 1955 promotion to assistant professor, and continued to share the General Circulation Project intellectually for another two decades. In Lorenz’s own words: “The things I remember best and cherish most, in looking back over my scientific career, are the almost daily conversations with Victor Starr during the more than twenty-five years that I worked with him.” Starr died in 1976.

Lorenz’s 1955 Tellus paper on Available Potential Energy – the foundational result of his energetic framework for the general circulation – emerged directly from this Starr collaboration. So did his 1967 WMO monograph The Nature and Theory of the General Circulation of the Atmosphere.

Another Starr protege: Robert White, who left MIT in 1960 to become president of the Travelers Research Center (TRC) in Hartford and was named Chief of the U.S. Weather Bureau by President Kennedy in 1963. White’s departure from MIT in 1955 had created the assistant-professor vacancy that Lorenz was promoted into – though the conventional story is that Charney made Lorenz’s promotion a condition of his own move to MIT in 1956. (The two accounts can be reconciled: Tom Malone resigned the tenured chair, White’s lower-ranked slot then went to Lorenz, and Charney secured the senior promotion.)

Source: Lorenz NAS memoir (Emanuel 2011); MIT ArchivesSpace Starr papers; UCAR memorials.

2.2 Charney arrives (1956)

Jule Charney moved from the Institute for Advanced Study in Princeton to MIT in 1956 after the Electronic Computer Project was wound down. (Von Neumann’s death in February 1957 effectively ended the IAS meteorology programme, but Charney had already made the decision to leave.) Houghton offered Charney a senior chair, and Charney accepted on two conditions, both well documented: (a) that Norman Phillips come with him as a “package deal,” and (b) that Lorenz be promoted to assistant professor. Houghton agreed to both.

This is the moment when MIT meteorology pivots into world preeminence in dynamical theory. Within five years it housed:

  • Charney (quasi-geostrophic theory, baroclinic instability, ENIAC veteran)
  • Phillips (first GCM experiment, 1956)
  • Lorenz (energetics, soon to discover chaos)
  • Starr (general circulation observational programme)
  • Houghton (cloud physics, department head)

In 1966 Charney became the Alfred P. Sloan Professor of Meteorology – the first holder of that chair. He served as Department Chairman 1974-1977.

The Charney-Lorenz coexistence (1956-1981). They were colleagues for 25 years, shared offices in Building 54 on different floors after 1964, and reportedly enjoyed each other’s company. Charney is the one who famously called Lorenz a man with “the soul of an artist.” They did not co-author papers (see Section 5.2 below). Their intellectual relationship was one of complementarity: Charney built the deterministic NWP programme; Lorenz quietly demonstrated its fundamental limits. There is no record of public friction between them on the predictability question – Charney accepted Lorenz’s results and incorporated them into his own GARP-era thinking by the late 1960s.

Charney died of lung cancer on 16 June 1981, aged 64, after a 20-month illness. Lorenz outlived him by 27 years.

2.3 Norman Phillips at MIT (1956-1974)

Norman A. Phillips (1923-2019) is the under-appreciated figure of this group. He came from Princeton in 1956 with Charney, served as research associate, then associate professor, then full professor, then Department Head 1970-1974 – exactly the years when the department had to absorb GARP responsibilities and start the planning that would lead to EAPS.

Phillips left MIT in 1974 to become Principal Scientist at the National Meteorological Center (NMC) in Washington, where he stayed until 1988. (This was a significant move: Phillips, the GCM pioneer, deliberately leaving academic theory for an operational role.) His 1974-1988 NMC tenure overlapped with Lorenz’s 1977-1981 chairmanship at MIT but bridged the academic and operational worlds in a way nobody else in the group did. When Eugenia Kalnay took over NMC modelling in the early 1990s and began operational ensemble forecasting, the institutional memory of Phillips was still present.

Phillips wrote Charney’s NAS biographical memoir (1995); the kind of careful, detailed insider account that is now an essential primary source.

2.4 Fred Sanders (1954-1984)

Frederick Sanders (1923-2006) joined the MIT faculty in 1954 – a year before Lorenz’s promotion – and stayed until retirement in 1984. He was the synoptic meteorologist of the group: a forecaster’s forecaster who taught the practical, map-based weather analysis that grad students needed in addition to the dynamical theory they got from Charney and Phillips. Sanders is best known for coining the term “bomb” (1980) to describe explosively deepening extratropical cyclones, and for helping develop early operational hurricane track models.

Born in Detroit 1923, he had enlisted in the Army Air Corps in 1941 and gone through MIT’s wartime meteorology programme; returned after the war; earned his ScD in 1954 (with a thesis on cold-frontal dynamics under James Austin – the same advisor who had supervised Lorenz). Sanders is often described as having been Charney’s student, but the more accurate description is that he was a senior colleague who taught with Charney for decades; his actual ScD predates Charney’s arrival.

Kerry Emanuel cites Sanders as his mentor in synoptic meteorology at MIT in the 1970s, alongside Charney as his theoretical advisor and Lorenz as a senior presence.

2.5 Reginald Newell (1961-2002)

Reginald E. Newell (1931-2002) came to MIT from Birmingham, UK, in 1954 as a research assistant; SM 1956, ScD 1960. Assistant professor 1961; associate professor 1966; full professor 1969 – he spent his entire 48-year career at MIT. His work was on global-scale tropospheric and stratospheric transport: water vapour, ozone, CO and CO2 budgets, sea-surface temperature/atmosphere coupling. He was a relatively early adopter of satellite data for climate work, and in 1969 (when he was still an associate professor) publicly warned the Massachusetts legislature about smog problems.

Newell was president of the International Commission on Climate 1977-1983 – exactly the period when CO2 climate science was crystallising into the Charney Report and IPCC trajectory. He provided MIT’s climate-observational presence while Lorenz did the theoretical work.

2.6 Erik Mollo-Christensen (1955-1984)

Erik Mollo-Christensen (1923-2009) is one of the more remarkable biographies in the group. Born in Bergen, Norway, he was a Norwegian-resistance member during the war, captured by the Nazis and sent to Buchenwald. He survived, came to MIT in 1946, and earned BSc/MSc/ScD there. He held three overlapping MIT appointments: professor of aeronautics 1955-1984, professor of meteorology 1964-1973, professor of oceanography 1973-1984. His scientific work was on turbulence physics, jet noise, and ocean tides and currents. He later served as Chief of NASA Goddard’s Laboratory for Oceans (post-MIT).

The Bergen connection is worth noting: Mollo-Christensen brought a direct line to the Bjerknes school via his Norwegian background and his collaboration with Arnt Eliassen. (Lorenz himself had a separate Norwegian connection through his 1953 UCLA visit, where he met both Jacob Bjerknes and Eliassen and formed a lifelong friendship with the latter.)

2.7 Henry Stommel (joint MIT-WHOI, 1959-1978)

Henry M. Stommel (1920-1992) shared the MIT/Woods Hole world. He held a Harvard appointment 1960-1963, then joined the MIT Department of Meteorology in 1963, remaining for 16 years until 1978 when he rejoined the WHOI staff full-time. Stommel was the foundational figure of dynamical physical oceanography (Stommel gyre, Sverdrup transport, deep western boundary currents) and the Lorenz-Stommel pairing matters institutionally for two reasons:

  1. The 1983 Crafoord Prize of the Royal Swedish Academy was awarded jointly to Edward Lorenz and Henry Stommel “for their fundamental contributions to the field of geophysical hydrodynamics, which in a unique way have contributed to a deeper understanding of the large-scale motions of the atmosphere and the sea.” The Crafoord is explicitly designed to honour fields not covered by the Nobels; the joint citation reflected the parallel revolutions Lorenz had made in atmospheric science and Stommel in oceanography.
  2. The MIT/WHOI joint program in physical oceanography was effectively a Stommel-led initiative that brought the atmosphere and ocean communities together long before EAPS formalised the merger.

2.8 The wider cast

Through the 1960s-1980s the department also included James Austin (Lorenz’s PhD advisor, still teaching synoptics into the 1970s); Edward “Ned” Lorenz Jr. (no relation to Edward Norton Lorenz, occasionally a source of confusion); Peter Stone (climate dynamics, joined in 1968); Paola Rizzoli (joined later, oceans); and after 1981 Kerry Emanuel (hurricanes, who had just finished his PhD with Charney).

3. Building 54: the Green Building (1962-1964)

The physical home of MIT meteorology from 1964 onwards was the Cecil and Ida Green Building, MIT Building 54, designed by I. M. Pei (MIT BArch 1940) with Araldo Cossutta. Construction: 1962-1964. Reinforced concrete, 18 floors, 84 metres (277 feet), the tallest building in Cambridge from completion until 2019 (when Site 4 in Kendall Square surpassed it). The first floor sits on stilts roughly 30 feet above ground, a deliberate move to circumvent Cambridge zoning.

The Green Building was designed specifically to consolidate the earth sciences and meteorology at MIT into a single tower – a major institutional commitment that signalled MIT’s view of meteorology, oceanography, geology and geophysics as one connected enterprise. Before 1964 the meteorology department had been scattered across older MIT buildings.

The Lorenz Center now occupies the top floor of the Green Building. The symbolism is deliberate: Lorenz himself had worked in Building 54 since its opening, and the top-floor placement gives the center a literal high-altitude vantage that staff and visitors comment on. Lorenz’s own office for many years was on the 16th floor.

Source: Wikipedia “Green Building (MIT)”; SAH Archipedia; MIT Capital Projects.

4. From Meteorology to EAPS (1983)

In 1983, the Department of Meteorology and Physical Oceanography (Course XIX, the descendant of Rossby’s 1928 programme) merged with the Department of Earth and Planetary Sciences (Course XII, which traced back to MIT’s original 1861 geology and mining engineering offering established by founder William Barton Rogers) to form the Department of Earth, Atmospheric and Planetary Sciences (EAPS), retaining the Course XII number.

Reasons commonly cited:

  • Building 54 had already physically co-located the relevant faculties for almost two decades.
  • Growth in planetary science (the Apollo lunar samples programme, Voyager, comet/asteroid work) had blurred the line between earth and atmospheric science.
  • Increasing climate science work straddled atmosphere, ocean, and solid-earth carbon cycle work in ways that needed institutional unification.
  • A general MIT pattern of consolidating small science departments.

The new EAPS housed the Program in Atmospheres, Oceans and Climate (PAOC), which today is effectively the descendant of the old Department of Meteorology. Course XIX as a number disappears with the merger; current students take Course XII.

Frank Press, the geophysicist who had chaired Earth and Planetary Sciences in the 1960s before becoming Carter’s science advisor in 1977, is often (informally) cited as having shaped the conditions for the eventual merger; but he was not at MIT for the 1983 merger itself. The merger occurred under Dean Marc Kastner’s predecessor (probably John Deutch’s brief tenure) and was administered through then-current EPS chair and meteorology chair (need to verify exact administrators – see Open Questions).

Lorenz was Emeritus by 1987, having stepped down as Department Head in 1981 (and from the chair Charney had vacated by death the same year). The 1983 merger therefore happened in his post-chairmanship but pre-emeritus years; he continued doing research throughout.

Sources: MIT EAPS About page; MIT Chaos and Climate news; MIT Course Catalog.

5. Reception of chaos in the wider NWP community

5.1 Charney’s predictability conferences and the 1969 GARP context

There is no single “Predictability Conference 1969-1971” – the user’s question conflates several events. What actually happened:

  • The Charney committee (1966). Charney chaired a National Academy of Sciences panel that produced The Feasibility of a Global Observation and Analysis Experiment. This panel, using results from the Mintz-Arakawa GCM at UCLA and the Smagorinsky GCM at GFDL, established the ~5-day error-doubling time estimate that became the foundation of the “two-week predictability limit.” Members included Lorenz, Mintz, Arakawa, Smagorinsky, Leith, and others. The Charney report is the canonical statement that the predictability limit is observed empirically in GCMs, complementing Lorenz’s earlier theoretical demonstration of sensitive dependence in his simple system.

  • GFDL dedication / GARP symposium, January 1969. A two-day event on the Forrestal Campus at Princeton, where Charney (as Chairman of US Committee for GARP) presided over the dedication of GFDL and the early planning meetings for what would become the Global Atmospheric Research Programme (GARP) and ultimately the FGGE 1978-79 observing campaign. Princeton President Robert Goheen and ESSA Administrator Robert White spoke.

  • Lorenz’s 1969 papers. Lorenz published six papers in 1969 alone, five on predictability. The most influential, “The predictability of a flow which possesses many scales of motion” (Tellus 21, 1969), introduced the multi-scale predictability argument that has become the basis of much modern theory. He also published “Three approaches to atmospheric predictability” in BAMS 50 (1969). These were Lorenz’s mature statements on what chaos meant for operational forecasting – and they were aimed squarely at the GARP community.

The conventional story: by 1969, Lorenz and Charney had converged on a shared understanding that the deterministic forecast horizon was around two weeks, and GARP was designed around that constraint – you need global observations because errors will grow and you must keep re-initialising.

5.2 Did Lorenz and Charney co-author?

Direct answer: there is no record of a peer-reviewed paper jointly authored by Lorenz and Charney. This is somewhat surprising given their 25-year MIT proximity. They served together on the 1966 Charney NAS predictability panel, but the report was Charney-authored. They both contributed to GARP documents and AMS proceedings volumes (1969 Princeton GARP symposium, various panels), but again no joint first-author paper.

The likely explanation is style: Charney worked in large collaborative projects (ENIAC team, IAS group, GARP); Lorenz worked alone or with computational assistants (Ellen Fetter, Margaret Hamilton on the LGP-30). Their intellectual territories overlapped only at the edges – Charney on baroclinic instability and large-scale dynamics, Lorenz on energetics and predictability of low-order systems.

Open question: verify by checking Charney’s full bibliography in Phillips’s NAS memoir. The proceedings volumes of the 1969 Princeton conference and subsequent GARP scientific meetings may contain Charney+Lorenz joint contributions in chapter form rather than journal form.

5.3 The Charney Report on CO2 (1979)

The Charney Report – formally Carbon Dioxide and Climate: A Scientific Assessment (National Academy Press, 1979) – was produced by an ad hoc study group of eight scientists meeting at Woods Hole 23-27 July 1979. Lorenz was not a member. The group included Charney as chair, Akio Arakawa, James Baker, Bert Bolin, Robert Dickinson, Cecil Leith, Henry Stommel, and Carl Wunsch. Their central conclusion – 3 degrees C equilibrium climate sensitivity, +/- 1.5 degrees C – has held up across four decades of IPCC assessments.

That Lorenz was not invited is consistent with the division of labour: chaos and predictability were his specialty; equilibrium climate sensitivity was a separate question of forced response, not of internal variability or sensitive dependence. Stommel’s presence (representing oceanography, by then back at WHOI) was the MIT representation.

5.4 ECMWF establishment (1975) and the Lorenz visits

The European Centre for Medium-Range Weather Forecasts was established by convention on 11 October 1973, started operations in Reading in late 1975, and went operational for routine 10-day forecasts in 1979. Its scientific charter was an explicit test of Charney’s “two-week predictability” claim: could you in fact get useful skill out to 10 days operationally?

Lorenz’s known ECMWF visits:

  • 1995 ECMWF Seminar on Predictability (4-8 September 1995, Reading), where Lorenz first publicly described the Lorenz 96 model – the 40-variable model that has since become the workhorse of predictability research and data assimilation algorithm testing. His proceedings contribution “Predictability: A problem partly solved” appeared in the seminar volume.
  • Late-1990s and early-2000s ECMWF predictability workshops, where Lorenz was a recurring invited speaker.

There is no direct evidence that Lorenz formally advised ECMWF as an institution, but he was a frequent visitor and his ideas suffuse the EPS development. Tim Palmer’s 2019 essay for the 25th anniversary of the EPS opens with Fig 1 – the Lorenz 1963 attractor – as the conceptual core of why ensemble prediction is needed.

5.5 AMS predictability panels through 1970s-80s

The AMS sponsored a sequence of predictability-themed conferences and Meteorological Monographs volumes. Lorenz was a recurring contributor and frequently a session-chair, though never – consistent with his personality – the lead organiser. The 1980s saw the predictability question shift from “is there a limit?” (settled, yes, ~2 weeks) to “how do we operationalise that limit?” (open, until 1992).

6. How chaos entered operational NWP: 1992 as the pivot year

6.1 The simultaneous EPS launches

1992 is the year ensemble prediction goes operational at two centres almost simultaneously:

Centre System Operational date Lead architects Initial-perturbation method
ECMWF (Reading) EPS late 1992 (often cited 19 Nov or 24 Nov 1992) T.N. Palmer, F. Molteni, R. Buizza Singular vectors
NMC/NCEP (Washington) global ensemble 7 December 1992 E. Kalnay, Z. Toth Bred vectors (breeding of growing modes)

This was not coincidence. Both centres had been working through the late 1980s on initial-perturbation methods for medium-range ensembles. Palmer had been thinking about the problem since the mid-1980s (initially in the Met Office Synoptic Climatology Branch, from 1982); Kalnay had been at NMC since 1987. Computer power had only just reached the point where 10+ daily integrations were affordable. By 1991 it was clear at both centres that the systems were ready.

The shared intellectual debt to Lorenz. Palmer (2019) is explicit:

“Lorenz’s motivation for developing this model [the 1963 system] was that it provided a counter-example to the claim that deterministic long-range forecasting using empirical methods would be possible…. However, Lorenz’s model can also be used to show the Achilles Heel of deterministic prediction within the deterministic limit.”

The Toth-Kalnay 1993 BAMS paper “Ensemble Forecasting at NMC: The Generation of Perturbations” cites Lorenz extensively. Palmer’s papers cite Lorenz on essentially every page where the underlying chaos rationale is invoked.

6.2 Singular vectors vs bred vectors

Palmer’s 2019 retrospective is the clearest first-person account of how the two approaches diverged. The short version:

  • Bred vectors (NMC/NCEP, Toth-Kalnay): start from a random perturbation, evolve it through a short forecast cycle, rescale, repeat. Mathematically equivalent to finding the leading Lyapunov vector by the power method. Metric-independent.
  • Singular vectors (ECMWF, Palmer-Molteni-Buizza): compute the eigenvectors of the operator A^T A where A is the linearised forecast propagator over a target window (typically 48 hours). These are the directions of fastest finite-time growth in a chosen metric (energy norm at ECMWF). Required the adjoint model that ECMWF had developed for 4DVAR data assimilation – a capability NMC did not then have.

Palmer’s argument for singular vectors over bred vectors centred on the fact that observations rotate the error field in wavenumber space (damping large scales more than small), and that singular vectors mimic this transformation while bred vectors do not. The empirical verdict over 25+ years has favoured ECMWF’s EPS as better-calibrated, though singular vectors are computationally more expensive.

6.3 The “Lorenz 75th birthday year” coincidence

Lorenz turned 75 on 23 May 1992. Both EPS launches happened in the second half of 1992. Whether this was deliberate timing is unclear – it appears not to have been; the coincidence is real but operational dates were driven by computer-system upgrade schedules at both centres rather than any homage. Palmer and Kalnay both made the explicit Lorenz-debt acknowledgment in their inaugural papers, and at the 1992 AMS Annual Meeting (Atlanta, January 1992 – thus before the operational launches) there was a session honouring Lorenz that previewed the operational systems.

By the 1996 Molteni-Buizza-Palmer-Petroliagis QJRMS paper “The ECMWF Ensemble Prediction System: Methodology and Validation,” the EPS was firmly established. By the late 1990s every major NWP centre worldwide had an operational ensemble.

6.4 The intellectual closure

The deepest closure is this: Charney’s 1950 ENIAC forecast, the founding event of operational NWP, was a deterministic calculation. Lorenz’s 1963 discovery, made on a desktop LGP-30 a few miles up the Charles River from where Charney was teaching, demonstrated the deterministic forecast was intrinsically limited. The 1992 ensemble launches showed how to live with that limit: replace single deterministic forecasts with probabilistic distributions. The 30-year arc from ENIAC (1950) to chaos (1963) to operational EPS (1992) is the central narrative of late-20th-century NWP, and MIT was the institutional home of two of the three pivots.

7. The Lorenz Center (2011-present)

7.1 Founding

Founded 2011 by Daniel H. Rothman (MIT geophysicist, joined MIT faculty 1986) and Kerry A. Emanuel (MIT atmospheric scientist, joined 1981), under the auspices of the MIT School of Science with the backing of then-Dean of Science Marc Kastner. Located on the top floor of Building 54 (the Green Building).

7.2 Mission

The center is explicitly framed as a climate think tank dedicated to curiosity-driven scientific inquiry, deliberately decoupled from the “pressing practical demands of climate forecasting” that the founders saw consuming much of climate science by the late 2000s. Rothman and Emanuel argued that the field had drifted toward applied/policy work and needed to recover its theoretical core. Naming the center after Lorenz was the natural choice: Lorenz had been emeritus at MIT until his death in 2008, and his entire scientific style – working alone or with a single assistant, pursuing fundamental questions on a small computer, ignoring institutional politics – embodied the curiosity-driven mode the center wants to encourage.

7.3 Activities

  • John Carlson Lectures – an annual public lecture series, given by leading climate scientists, intended for general audiences. Recent lecturers (verify list): Susan Solomon, Tim Palmer, Brian Hoskins, others.
  • Lorenz-Houghton Postdoctoral Fellowship – multi-year fellowship for early-career theorists working on climate dynamics. Named jointly for Lorenz and Houghton.
  • Visiting scholar programme.
  • Workshops on climate dynamics, chaos, predictability.

7.4 Current leadership

As of 2026: Daniel Rothman remains a co-founder/director. Raffaele Ferrari (ocean dynamicist, MIT EAPS) is co-director after stepping down as chair of PAOC (Susan Solomon succeeded him as PAOC chair). Kerry Emanuel is Cecil and Ida Green Professor Post-Tenure of Atmospheric Science and remains active in center affairs through retirement (he formally retired from MIT teaching in 2022 after 41 years).

7.5 The 2018 Chaos and Climate symposium

Held February 2018 at MIT, organised by EAPS to mark the centennials of both Lorenz and Charney (both born 1917). Keynote: Ernest Moniz (MIT physics, former Secretary of Energy). Panel: Sir Brian Hoskins, Inez Fung, Kerry Emanuel, Allison Wing, John Bush, moderated by Robert van der Hilst. Organisers: Raffaele Ferrari, Kerry Emanuel, John Marshall, Paola Rizzoli, Dan Rothman.

Quotable lines from the symposium:

  • Emanuel: “History may well record that Ed Lorenz had hammered the last nail into the coffin of the Cartesian universe.”
  • David Randall on Charney: “Being in the room with Charney was like being in the room with a tiger, a very friendly tiger.”

The contrast between the two Randall-style portraits – Charney the friendly tiger, Lorenz the “soul of an artist” / “gentle, quiet soul” – captures the temperamental complementarity that defined the department for 25 years.

8. Open questions and gaps

  1. Exact MIT administrators of the 1983 EAPS merger. Need to identify the EPS chair (Earth and Planetary Sciences) and Meteorology chair at the moment of merger, and the Dean of Science. The Phillips-Lorenz-Houghton transitions of department-headship in the meteorology department are reasonably clear but the planetary-sciences side is not yet documented in these notes.

  2. Lorenz-Charney joint contributions in proceedings volumes. Direct journal-paper co-authorship looks unlikely, but the 1966 Charney panel report, the 1969 Princeton GARP proceedings, and various AMS Meteorological Monographs may contain co-authored chapters. A proper bibliography check is needed.

  3. Lorenz’s actual ECMWF role. Was he formally an external advisor, or simply a frequent invited visitor? The 1995 Reading seminar is the most-cited event; need to check whether he served on any ECMWF advisory bodies.

  4. The Tim Palmer / MIT connection. Palmer gave the 2012 John Carlson Lectures at MIT but does not appear to have been a postdoc there. Worth verifying through his Oxford and ECMWF biographies.

  5. Why exactly 1956 for Charney’s move. Conventional story: von Neumann’s failing health (he was diagnosed with cancer in summer 1955) plus IAS computer-project wind-down. But Houghton must have been actively recruiting; the institutional MIT side of the negotiation is not well documented in published sources.

  6. The Saltzmann visit. Lorenz visited Barry Saltzmann (often spelled Saltzman) – where, and when exactly? Lorenz’s NAS memoir says simply “Lorenz visited Saltzmann” without dating; the visit is conventionally placed in 1961-1962 and at Travelers Research Center in Hartford, but verification would be useful given Robert White’s TRC presidency at that exact time.

  7. The 1992 AMS Annual Meeting session honouring Lorenz. Need to verify it existed and check speaker list – this would be the natural “preview” of the operational EPS launches later that year.

9. Key sources cited

Primary

  • Lorenz, E. N. (1963). “Deterministic Nonperiodic Flow.” J. Atmos. Sci. 20: 130-141.
  • Lorenz, E. N. (1969). “The predictability of a flow which possesses many scales of motion.” Tellus 21: 289-307.
  • Lorenz, E. N. (1969). “Three approaches to atmospheric predictability.” BAMS 50: 345-349.
  • Lorenz, E. N. (1995). “Predictability: A problem partly solved.” Proceedings of the Seminar on Predictability, Vol. 1, ECMWF, Reading.
  • Charney, J. G. et al. (1966). NAS panel report establishing the 5-day error doubling and 2-week predictability limit.
  • Charney, J. G. (chair) (1979). Carbon Dioxide and Climate: A Scientific Assessment. National Academy of Sciences.
  • Phillips, N. A. (1956). “The general circulation of the atmosphere: a numerical experiment.” QJRMS 82: 123-164.
  • Toth, Z. and Kalnay, E. (1993). “Ensemble Forecasting at NMC: The Generation of Perturbations.” BAMS 74: 2317-2330.
  • Molteni, F., Buizza, R., Palmer, T.N., Petroliagis, T. (1996). “The ECMWF Ensemble Prediction System: Methodology and Validation.” QJRMS 122: 73-119.
  • Palmer, T. N. (2019). “The ECMWF ensemble prediction system: Looking back (more than) 25 years and projecting forward 25 years.” QJRMS 145 (S1): 12-24. arXiv:1803.06940.

Memoirs and biographies

  • Emanuel, K. (2011). “Edward Norton Lorenz, 1917-2008.” Biographical Memoirs NAS.
  • Palmer, T. N. (2009). “Edward Norton Lorenz. 23 May 1917 - 16 April 2008.” Biographical Memoirs of Fellows of the Royal Society 55: 139-169.
  • Phillips, N. A. (1995). “Jule Gregory Charney.” Biographical Memoirs NAS 66: 80-113.

Institutional histories

  • MIT EAPS, “About Us / History.” https://eaps.mit.edu/about/
  • MIT News (2011), “Wind, war and weathermen.” https://news.mit.edu/2011/timeline-forecasting-0607
  • MIT News (2018), “Chaos and climate: Celebrating two pioneers of modern meteorology.” https://news.mit.edu/2018/mit-chaos-and-climate-celebration-two-pioneers-modern-meteorology-0214
  • Lorenz Center, https://www.lorenz.mit.edu/about
  • MIT News obituaries: Lorenz (2008), Phillips (2019), Sanders (2006), Newell (2003), Mollo-Christensen (2009).
  • Wikipedia, “Green Building (MIT).”
  • SAH Archipedia, “Green Center for Earth Sciences (Bldg. 54).”

Archival

  • MIT ArchivesSpace: Houghton papers; Starr papers.
  • WHOI Archives: Stommel papers.

Secondary

  • Shen, B.-W. et al. (2024). “Lorenz’s View on the Predictability Limit of the Atmosphere.” MDPI Encyclopedia 3: 887-905.
  • Lewis, J. M. (2007), “Robert White and the political reshaping of the Weather Bureau” (TRC connection).

Notes compiled by research assistant, 2026-05-15, for forthcoming blog post in the NWP history series. ~5100 words.