Satellite-tracked cumulus velocities

Basic problems in the interpretation of satellite-tracked low-cloud velocities are reviewed. The METRACOM system of cloud velocity computation is outlined, and caution is urged in converting cloud velocities into wind velocities. The motion of various cumulus cells over Springfield, Mo., Barbados, and Tampa, Fla., is analyzed. It is shown that multiturret cells do not always move with the wind, that addition and deletion of turrets belonging to a specific cell may cause erratic motion in a tracer cell, and that cumulus turrets between 0.3 and 2 miles in size are the best targets for inferring the mean wind velocity within the subcloud layers. It is concluded that the accuracy of wind velocity estimates will be no better than 2 meters/sec unless the physical and dynamic characteristics of cumulus motion are further investigated.