THE ACCIDENTAL CLIMATOLOGIST

Twenty years before he won a Nobel prize, not-yet-famous author and philosopher  Albert Camus became an accidental meteorologist.

Seeking work, the young writer found a position at the Algiers Geophysical Institute, cataloging reports from French colonial weather stations ranging from Djibouti to Dakar.

 

Laura Marris writes in The Point  that:

"The data had been piling up...  Camus had plotted curves for 27 years of barometric pressures from 121 weather stations. He also made calculations, averaging monthly meteorological data. This work must have given him a granular picture of the weather, one that was so dry and clinical it was at odds with his experience of the natural world."

The experience drove Camus to pen a precocious note about an epistemological issue that still impacts modeling and GCM intercomparison: as surely as models are not things, neither are sporadic data sets :

 “Like in all sciences of description (statistics—which collects facts—) the biggest problem in meteorology is a practical problem: that of replacing missing observations.

It vexed Camus that :

Temperature varies from one minute to the next.  This experiment shifts too much to be stabilized into mathematical concepts. Observation here represents an arbitrary slice of reality.”