Abstract
A technique is proposed for computing horizontal velocity divergence and the vertical component of vorticity from radiosonde wind data. Utilizing a quadratic Taylor's series representation of the horizontal wind field, one can consider nonlinear variations in the wind directly in estimates of the first-order derivatives of the wind components. These nonlinear variations are found to be significant in a number of cases. Vertical motions computed by the kinematic method and horizontal divergence are modified by an adjustment scheme. Comparison of these results with those derived from computations from a linear Taylor's series representation of the wind suggests that the quadratic model is superior to the linear. Synoptic analyses of vorticity, divergence, and vertical motions over the United States at 0000 and 1200 GMT on Apr. 13, 1964, reveal good agreement with the circulation patterns and associated weather.
On leave of absence to The Advanced Study Program at the National Center for Atmospheric Research, Boulder, Colo.