It is an emotional experience to be in the company of distinguished delegations from nearly a score of European nations, and to see them stand in silence in honour of a fellow countryman. It happened to me earlier this week in Reading, England, at a meeting of the Council of the European Centre for Medium Range Weather Forecasts.
ECMWF is a co-operative effort by 17 European countries to provide daily weather predictions for the period two to 7 days ahead. Established in 1976, it has a team of top meteorological scientists and a very large computer installation, and it is now the world leader in its field of speciality.
Back in the early 1980s, Patrick Kilian Rohan, then director of the Irish Meteorological Service, became the third president of the ECMWF Council, the centre's ruling body. It was to mark his passing about two months ago, and the excellent work that he had done for the organisation, that former colleagues and many who had never met him did him the poignant honour of a minute's silent tribute.
But I also came across a catchy weather phrase in Reading. "Model biodiversity", an obvious analogy with its biological and evolutionary equivalent, has come to be recognised among meteorologists as an ambition devoutly to be wished.
The models of interest to meteorologists, of course, are not the Naomi Campbells or the Twiggys of this world, but rather those mathematical models of the atmosphere that are used for weather forecasting.
The concept of "model biodiversity" recognises the benefits to the science of meteorology of many different organisations or countries developing their own weather forecast computer models, an exercise that might seem, on first consideration, to be somewhat wasteful; the wheel, so to speak, being reinvented simultaneously at many different centres of would-be excellence around the world.
All the models, of course, incorporate the basic tried and trusted methods that have come to be the norm, but they differ from each other in the ways in which they are, from time to time, incrementally improved.
Different approaches are tried in different places; in a sense, the models are in competition, like the different species in the natural world, trying to adapt themselves to the complexities of the earthly atmosphere. Successful innovations give a model a competitive edge; the "fittest" models survive, and the less successful methodologies become extinct.
And indeed diversity, again as in the natural world, allows models to adapt to the local environment in which they have to operate. A model which works well over the level, temperate waters of the north Atlantic may be less successful in coping with the Alps.