The Computational Physics of the Global Atmosphere

Abstract

It was in the late 1950’s that Edward Teller encouraged me to investigate the role that newly available computing power at Livermore might play in achieving a more complete numerical simulation of the global atmosphere. This led to the construction of the first atmospheric model with an explicitly computed cycle of water vapor evaporation, transport, and condensation. With this and similar efforts elsewhere confidence was developed in the feasibility of such dynamical and physical modeling. But it also became clear that fundamental problems in the prediction of the large scale eddies characterizing the flow of the atmosphere required a deeper understanding of two-dimensional turbulent flows in general. These problems have occupied me for the past quarter century, and with this report I wish to acknowledge Edward’s initial support and guidance.