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.

Frederick Gale “Bud” Shuman

Born: July 13, 1919, South Bend, Indiana Died: July 29, 2005 (aged 86), Fort Washington, Maryland Cause of death: Congestive heart failure

Overview

Frederick Gale Shuman, universally known as “Bud,” was the chief modeler of the Joint Numerical Weather Prediction Unit (JNWPU) from its founding in 1954, and Director of the National Meteorological Center (NMC) from 1964 to 1981. He was the first employee hired by JNWPU and the longest-serving mathematical backbone of U.S. operational NWP, bridging the era of the IBM 701 through the computerized modeling systems of the 1970s. His 1957 paper on smoothing and filtering operators – the “Shuman smoother” – remains a fundamental reference in finite-difference NWP.

Education

Year Degree / Institution Notes
1941 B.S., Ball State University, Muncie, Indiana Became a junior weather observer at Indianapolis Airport while still at Ball State
1942–1945 U.S. Army Air Corps meteorology training; MIT M.S. in meteorology Commissioned 2nd lieutenant; served in North Africa and Italy as weather officer
1951 Doctor of Science in Meteorology, MIT  

Career

Early Weather Bureau Career (1941–)

Shuman began his Weather Bureau career as a junior observer at Indianapolis Airport while still completing his Ball State degree. After his WWII service he returned to the Weather Bureau and pursued his MIT doctorate, receiving it in 1951.

Institute for Advanced Study, Princeton (1952–1954)

Between 1952 and 1954 Shuman was a researcher at the Institute for Advanced Study in Princeton, New Jersey, working on mathematical models for weather prediction. At IAS he used one of the world’s earliest operating computers – the JOHNNIAC (a late IAS-class machine, sometimes identified in sources as the IAS machine proper) – to develop and test numerical forecast methods. He attended lectures by J. Robert Oppenheimer and Albert Einstein during this period.

This IAS posting put Shuman directly in contact with the intellectual lineage of Charney, von Neumann, and the Princeton NWP group, and made him the natural choice to staff the operational unit that would industrialize their methods.

Joint Numerical Weather Prediction Unit (1954–)

In 1954 Shuman became the first employee of the newly created Joint Numerical Weather Prediction Unit at Federal Office Building 4, Suitland, Maryland. He was the JNWPU’s chief modeler, responsible for the mathematical implementation of the NWP models on the IBM 701.

The first operational U.S. computer forecast was issued on May 6, 1955. It was, in Shuman’s own later assessment, disappointing. Writing in 1989 he stated plainly:

“Within three years three agencies of the United States Government jointly created a numerical weather prediction service, but it was quickly discovered that current models had very serious defects. After considerable research, the first operationally effective model was achieved in 1958.”

The problems were multiple: the three-level quasi-geostrophic model was too crude vertically; quasi-geostrophic assumptions broke down near the equator, creating a “tropical belt” problem that contaminated the rest of the solution; objective analysis did not yet exist (hand-drawn input fields introduced subjective errors); and the IBM 701’s Williams tube memory failed roughly every 30 minutes, forcing frequent forecast restarts.

The 701 was replaced by the IBM 704 at JNWPU in 1957. With improved hardware and Shuman’s improved barotropic model, the unit began producing forecasts that consistently beat human competition by 1960.

National Meteorological Center Director (1964–1981)

On January 1, 1958, the JNWPU merged with the older National Weather Analysis Center to form the National Meteorological Center (NMC). Cressman continued as director until 1964. Shuman then succeeded him as NMC director, a post he held until 1981 – seventeen years. During his directorship, NMC expanded from IBM 704-era computing to CDC 6600 and IBM 360-class machines, and from barotropic models to full primitive-equation general circulation models.

He continued research at NMC until 1986, five years after his retirement as director.

Key Contributions

The Shuman Smoother (1957)

Shuman’s most cited technical contribution is the paper “Numerical Methods in Weather Prediction: II. Smoothing and Filtering”, published in Monthly Weather Review, volume 85, November 1957, pages 357–361. The paper derived a class of finite-difference operators for smoothing and filtering numerical forecast fields to remove spurious short-wave noise without distorting the meteorologically significant scales. The design method – specifying the response function in wavenumber space and then finding the corresponding finite-difference weights – became standard practice in operational NWP.

The Shuman smoother (specifically the weighted running-mean smoother derived in the paper) is still cited as a foundational reference in texts on numerical methods for atmospheric modeling.

Operational NWP Transition

Shuman wrote the definitive insider history of the NMC/JNWPU period: “History of Numerical Weather Prediction at the National Meteorological Center”, published in Weather and Forecasting, volume 4, number 3, pages 286–296, 1989. This retrospective is the primary source for the view from inside the unit during the critical years of 1955–1960.

Awards and Honors

  • Department of Commerce Silver Medal (1957)
  • Department of Commerce Gold Medal (1967)

Connections to Others

  • George Cressman – Cressman was JNWPU/NMC director (1954–1964); Shuman was his chief modeler and successor
  • Jule Charney – The three-level quasi-geostrophic model that failed on the IBM 701 in 1955 was Charney’s model; Shuman implemented and diagnosed it
  • Nathaniel Rochester / IBM 701 – The 701’s reliability failures (Williams tube mean time between failures ~30 minutes) were a major operational constraint that Shuman had to work around daily
  • J. Robert Oppenheimer – Shuman attended Oppenheimer’s lectures at IAS in 1952–1954, two years before Oppenheimer’s security clearance was revoked
  • Norman Phillips / Bert Bolin / Aksel Wiin-Nielsen – Phillips helped establish the theoretical basis for the JNWPU models from the IAS end; Bolin and Wiin-Nielsen moved from Rossby’s Stockholm group directly to JNWPU

Key Publications

  • Shuman, F.G. (1957). “Numerical Methods in Weather Prediction: II. Smoothing and Filtering.” Monthly Weather Review, 85(11), 357–361.
  • Shuman, F.G. (1989). “History of Numerical Weather Prediction at the National Meteorological Center.” Weather and Forecasting, 4(3), 286–296.

Sources