Deep Convective Mass Transports: An Estimate from a Geostationary Satellite

Abstract

A technique is described for inferring vertical mass circulations within and around cloud clusters in the tropics. Following Sikdar and Suomi (1972), we model deep convection in terms of three layers?two active and one passive. An inflow layer extends from the sea surface to the top of the trade inversion; an outflow layer extends through the depth of the cumulonimbus anvil. High-density satellite cloud tracer winds define flow for the two active layers. (The intermediate layer is not observed; in that sense it is considered to be passive.) Using the divergence field computed from the satellite winds, vertical velocities are calculated through the top of the inflow layer, and the bottom of the outflow layer; vertical mass transports follow immediately. This simple model is applied to two mature disturbances from the Global Atmospheric, Research Program Atlantic Tropical Experiment (GATE). Vertical velocities and mass transports were estimated for three times, at 3 h intervals. Vertical velocities on the scale of the clusters averaged 2?18 cm s?1, with highest values in the outflow layer. For both GATE clusters, maximum ascent appeared to occur several hours earlier at low levels than at high levels. Vertical velocity was smaller on the larger cluster circulation scale, but still positive leading to rather large upward transport of mass outside the clusters. Cluster-scale mass transports ranged from 15 to almost 35 mb h?1. An exception was inflow level transport for the more mature of the GATE clusters, which became slightly negative, implying a large middle tropospheric transport of mass into the cluster. Centers of divergence and upward mass transport were well matched to convective activity in the satellite pictures, even reflecting changes in position and intensity.