A Quantitative Study of Satellite Winds for Mesoscale Meteorology

Abstract

Quantitative characteristics and utility of mesoscale satellite winds are investigated for a low-level high spatial resolution data set obtained from a sequence of 6 min interval synchronous meteorological satellite images of the central region of the United States on 20 May 1977. Attention is focused on the quantitative errors introduced by height assignment in the presence of vertical wind shear and by the objective analysis of such irregularly spaced data to a regular grid point array. It is shown that assignment of the wind vectors to a single level introduces a local variability and systematic horizontal shears due to the vertical wind-shear effect giving a variability comparable to that expected in natural mesoscale phenomena with 100 km length scale. The random component of the local variability can be reduced by appropriate averaging which is possible because of the data density. The error introduced by the objective analysis procedure is estimated by examining the differences between various analysis methods. This sensitivity test is made both for grid spacing and for objective analysis method and includes the use of an analytical function field. Although there is a large variation in the results, it is estimated for the more reasonable cases that the variations in most areas are not greater than those expected from other error sources.