ADP, EDP and Computers (continued)

After Gerry O'Mahony returned to the Bureau, following his year with Prof Moran studying statistics at the ANU, Len Dwyer sent him and J. V. Maher to attend the meeting of the WMO Technical Commission for Climatology (CC1) in London at which Gerry had his somewhat spectacular public discussion with Forsdyke on identification of punch-cards. Len arranged that Gerry would return to Australia via Europe and the US with time to survey the state of progress with the use of computers to make numerical weather predictions. Gerry has informed me that he saw nothing spectacular and that in the US progress was slow "because of the lack of available software".

On page 11 of his history of the Meteorology Department of the University of Melbourne (June, 1993) Uwe Radok describes how the acquisition of the prototype digital computer CSIRAC by the Physics Department in 1956 made him aware of the possibility of using it for a program of computer modelling research in the Meteorological Department. He was becoming accustomed to the use of punched paper tape to code the CSIRAC memory when two brand-new physics graduates, Peter Schwerdtfeger and Dick Jenssen, came to see whether they could do postgraduate work with him.

Uwe tells how Dick Jenssen quickly mastered CSIRAC and other machines of the day and within less than the prescribed two years qualified for his M.Sc. degree with numerical predictions of 500 mb circulations. These predictions took almost five hours to make a 24 hour forecast, not counting time correcting machine errors which on average occurred at half hour intervals. Uwe and Dick Jenssen contributed a paper on numerical prediction to the Bureau's rain seminar in 1960 after Uwe was able to adapt the model to provide vertical velocities.

Bureau of Meteorology