This paper describes results from numerical experiments which have been performed to understand the effects
of the ice microphysics, surface friction, and surface heat flux on tropical cyclone (TC) formation. This study uses the
author’s non-hydrostatic model that intends to resolve cumulus convection. However, the horizontal grid size is taken to
be somewhat large; 2 km in an area of 600 km x 600 km. A non-uniform coarse grid is used in the surrounding area with
4,000-km square. Several buoyancy perturbations arranged in the west-east direction, and a weak vortex with the
maximum wind speed of 5 m s–1 are given at the initial time of the numerical time integrations.
It is confirmed from two numerical experiments with and without ice microphysics that the development of a vortex is
slower, and TC formation is delayed, in the presence of ice microphysics. It is also confirmed that a vortex can develop
even without surface friction. It is shown that a strong vortex with the maximum wind speed of 20~25 m s–1 can be
obtained. As expected, however, no eye forms, and further development does not occur. That is, it is confirmed that
surface friction is indispensable to eye formation and a very strong TC having an eye. As for the third concern of this
study, it is shown that a vortex with the maximum wind speed of about 5 m s–1 does not develop in the absence of the
surface heat flux. That is, the surface heat flux plays an important role even in a weak vortex. Important backgrounds and
understandings that are concerned with these results are described, based on studies on TCs in the past 50 years.