Large-scale and climate models rely on subgrid-scale modeling to
represent (or ``parameterize'') moist convection. Consequently, key
aspects of clouds in the problem of tropical dynamics, such as the
interaction of convection with radiative or surface processes, cannot
be directly investigated. A cloud-resolving modeling approach, on the
other hand, considers the convective response to the prescribed large-scale conditions, but it allows for no feedback from cloud-scale dynamics into the large-scale flow. One of the conclusions drawn from
our cloud-resolving studies of GATE and TOGA COARE cloud systems is that a two-dimensional (2D) framework does offer a meaningful representation of the tropical convection, including interaction of clouds with radiation and effects of convection on surface heat fluxes (Grabowski et al., JAS 1998; Wu et al. JAS 1998). This result leads us to believe that it is possible to develop a three-dimensional (3D) large-scale model (with horizontal resolution of the order of 100 km) in which convective processes are represented in every model column by a 2D cloud resolving model. We refer to this approach as a ``cloud resolving convection parameterization''. Although computationally quite expensive, this approach is ideal for parallel computations and currently seems the only way to include elements of cloud dynamics into 3D large-scale and climate models. This approach has been successfully tested by comparing 2D cloud-resolving simulations of the Walker-like circulation with the simulations applying cloud resolving convection parameterization. We are testing this approach using GATE data and plan to apply it to the problem of equatorial beta-plane dynamics in the near future. Details of the approach and results of these tests will be presented at the meeting