Abstract
Chao's numerical and theoretical work on multiple quasi equilibria of the intertropical convergence zone (ITCZ) and the origin of monsoon onset is extended to solve two additional puzzles. One is the highly nonlinear dependence on latitude of the “force” acting on the ITCZ due to the earth's rotation, which makes the multiple quasi equilibria of the ITCZ and monsoon onset possible. The other is the dramatic difference in such dependence when different cumulus parameterization schemes are used in a model. Such a difference can lead to a switch between a single ITCZ at the equator and a double ITCZ, when a different cumulus parameterization scheme is used. Sometimes one of the double ITCZ can diminish and only the other remains strong, but still this can mean different latitudinal locations for the single ITCZ. A single idea based on two off-equator attractors for the ITCZ symmetric with respect to the equator, due to the earth's rotation, and the dependence of the strength and size of these attractors on the cumulus parameterization scheme solves both puzzles. The origin of these rotational attractors, explained in Part I, is further discussed. Each attractor exerts on the ITCZ a force of simple shape in latitude; but the sum gives a shape highly varying in latitude. Also the strength and the domain of influence of each attractor vary when change is made in the cumulus parameterization. This gives rise to the high sensitivity of the force shape to cumulus parameterization. Numerical results, of experiments using Goddard's GEOS GCM, supporting this idea are presented. It is also found that the model results are sensitive to changes outside of the cumulus parameterization. The significance of this study to El Niño forecast and to tropical forecast in general is discussed.