STOCHASTIC AND PROBABILISTIC METHODS FOR OCEAN, ATMOSPHERE, AND CLIMATE DYNAMICS BY ADAM H. MONAHAN, BOUALEM KHOUIDER, NORMAN MCFARLANE, JOHN SCINOCCA, AND KNUT VON SALZEN STOCHASTIC AND PROBABILISTIC METHODS FOR OCEAN, ATMOSPHERE, AND CLIMATE DYNAMICS WHAT: Graduate students, postdoctoral fellows, and senior researchers from universities and operational climate modeling centers met to discuss stochastic and probabilistic approaches to the dynamics of the atmosphere, the ocean, and the climate system WHEN: 14–18 July and 21–23 July 2008 WHERE : Victoria, British Columbia, Canada W eather and climate are distinct aspects of the variability of the earth’s fluid envelope. The idea that “climate is what you expect, but weather is what you get’’ is fundamentally probabilistic: the evolution of weather from day to day is seen as the realization of a chaotic (or stochastic) system, while climate is the probability distribution of all possible states of the weather. Understanding the connections between weather and climate is compli- cated by the facts that i) processes in the fluid flow result in interactions across space and time scales (so that weather and climate are dynamically coupled) and ii) there is no unique or unambiguous scale separation between fast and slow parts of the system. One person’s signal is another person’s noise. Probability and stochastic processes provide natu- ral tools for studying the dynamics of the weather– climate connection. In the summer of 2008, the University of Victoria in Victoria, British Columbia, Canada, hosted a summer school and workshop on the subject of Stochastic and Probabilistic Methods for Ocean, Atmosphere, and Climate Dynamics, bringing together graduate students, postdoctoral fellows, and senior researchers to address questions arising from such a “probabilistic perspective on climate.” These questions ranged from the charac- terization of statistical structures in observations through fundamental issues in the development of stochastic and probabilistic tools to the application of these tools in practical problems. The summer school presented a sequence of short (3–4 h) introduc- tory courses: probabilistic methods for atmosphere, ocean, and climate science (by Adam Monahan, University of Victoria), information theory and sta- tistical predictability (by Richard Kleeman, New York University), data assimilation (by Saroja Polavarapu, Meteorological Service of Canada), geophysical fluid statistical mechanics (by Xiaoming Wang, The Florida State University), parameterization in large- scale atmospheric modeling [by Norm McFarlane, Canadian Centre for Climate Modelling and Analysis (CCCma)], modern methods for transport equations (by Boualem Khouider, University of Victoria), and Monte Carlo methods (by Alexandros Sopasakis, University of North Carolina, Charlotte). The lecture notes from the summer school are available for download (at www.pims.math.ca/scientific/summer- 211 FEBRUARY 2011 AMERICAN METEOROLOGICAL SOCIETY | Downloaded from http://journals.ametsoc.org/doi/pdf/10.1175/2010BAMS2992.1 by guest on 21 July 2020