On Cellular Cloud Patterns. Part 2: Laboratory Model

Abstract

In a mathematical model (presented in Part 1), an unstable lapse rate ? and a uniform vertical velocity w0 were imposed upon a horizontal layer of fluid between two porous boundaries. The stability of the nonlinear solutions was investigated, and it was shown that rolls would be stable if w0 were sufficiently small and ? sufficiently large, while hexagons would be stable if w0 were sufficiently large and ? sufficiently small. Furthermore, the direction of flow in the hexagons (related to the question of open or closed cellular cloud patterns) depends upon the sign of w0. The validity of these results is tested in laboratory experiments in which measurements were made of the Rayleigh number, heat flux, vertical velocity, plan form and cell size. Hexagons and rolls are indeed found to be the realized flow in the respective ranges of parameters in which they were theoretically predicted to be stable. Their occurrence is mapped on a regime diagram. The dependence of the direction of flow in hexagons was also found to depend upon w0 in agreement with the theory.