André Robert Atmospheric Science

Climatologiste: The first Canadian scientist to successfully perform a simulation of the atmosphere’s general circulation using a computer model.

The Story

André Robert was a Canadian born in the US of French-Canadian parents. He was, for the most part, educated in Grand-Mère, Québec. After a B.Sc. in Mathematics from Laval University, he started a career as a weather forecaster with the Canadian Meteorological Service. But in 1959 Robert transferred to the research section where he was engaged in the development of atmospheric models for short and medium-range forecasts. His chief accomplishments were to develop and implement efficient numerical techniques to solve the interacting time-dependent partial differential equations that govern the evolution of the atmosphere. Robert also had a strong influence on the international climate modeling community. He was a founding member and one of the first chairs of the Working Group on Numerical Experimentation (WGNE) a very influential group of researchers established in the early 1970s that oversaw the development of modern day climate science and meteorology.

Between 1963 and 1970 Robert developed the first successful spectral method for global forecasting. It was part of his Ph.D. at McGill University. Afterwards he combined it with the semi-implicit method and also developed a time filter that was instrumental in the improvement of climate models. In 1973 he became director of the Canadian Meteorological Centre. Then in 1980 he went back to pure research at RPN (Recherche en prévision numérique, MSC) where he successfully coupled the semi-implicit method with Lagrangian advection so as to allow a further gain in efficiency in integrating the meteorological equations. After his retirement from Environment Canada in 1987 Robert developed the MC2 grid-point model at UQAM (Université du Québec à Montréal) a multi-scale compressible model that was able to simulate flows on many scales.

Robert inspired many colleagues, young and old, and left an amazing legacy. His clear and very compact and efficient explanations were a highlight of all his presentations and memorable to all that experienced them. His methods were adopted widely partly because he was frequently invited to lecture in the USA and Europe at places like the Geophysical Fluid Dynamics Laboratory in Princeton, NJ, the National Meteorological Center in Washington DC, now known as the National Centers for Environmental Prediction. He was also invited to the National Center For Atmospheric Research in Boulder, Colorado, the United Kingdom Meteorological Office, the École normale supérieure in Paris, France, and the World Meteorological Organization in Geneva, Switzerland.

In contrast to many scientists in this field who aim primarily at improving accuracy, Robert always devised efficient numerical methods to attain a given degree of precision with the least amount of computation. The methods that he developed are now used in models at the world’s largest weather prediction and climate research centres.

Source: Personal communication from John Digby Reid, Ian Rutherford, Richard Asselin, Canadian Meteorological and Oceanographic Society, and from Jean Côté, Environment Canada (retired) and UQAM.

Further Reading: André Robert (1929-1993): His Pioneering Contributions to Numerical Modelling, by Andrew Staniforth (pp 25-54), in Numerical Methods in Atmospheric and Oceanic Modelling: the André J. Robert Memorial Volume:  edited by Charles A. Lin, René Laprise and Harold Ritchie,  581 pages,  Published 1997, ISBN 0-9698414-4-2, Publisher Canadian Meteorological and Oceanographic Society and NRC Research Press.

The Person

April 28, 1929
New York, NY
Date of Death
November 18, 1993
Place of Death
  • BSc, Laval University, 1952
  • MSc, University of Toronto, 1953
  • PhD, McGill University, 1965
Last Updated
May 17, 2012
Profile viewed 27512 times

Other scientists who may be of interest: