The present status of the use of the SASS (Scatterometer Satellite System) on SEASAT-A is reviewed based on SKYLAB and aircraft programs. Procedures, which are under development, to obtain vector wind fields from the SASS data in a form that will permit their operational use in computer based numerical weather forecasting are described.
Global computer based weather prediction techniques can be improved in a number of different ways. Three of them are:
improved models of the ocean and atmosphere;
higher resolution; and
more accurately specified initial value conditions. Improved models are under continuous development and testing.1 Greater resolution is limited by computer speed and capacity, but it will eventually be achieved.1 In our opinion, improved accuracies for the initial conditions for numerical weather predictions will provide the greatest potential for the improvement of short range forecasts.2
The data to be obtained from SEASAT-A should go a long way towards a better initial value specification for the atmosphere over the oceans. Tests of the concepts employed for SEASAT-A made from SKYLAB, and by means of several aircraft programs, all show that a combination of a Doppler radar (SASS), a five frequency passive microwave scanning radiometer (SMMR) and a visible and infra-red imaging system (V and IR) on SEASAT-A can be used to determine the vector winds in the first 50 to 100 meters above the ocean surface.
The SMMR, the SASS, and the V and IR all measure quantities whose variations are caused, in part, by properties of the ocean surface (such as small capillary waves, the slopes of longer waves, foam cover and ocean temperature). These properties in turn are caused to vary in part by the winds. The many modes of interaction between the ocean and the atmosphere thus yield results of great importance to meteorology as well as to oceanography. Improved measurements of the winds, in turn, should provide part of what is needed for improved wind forecasts and these improved wind forecasts should, in turn, yield improved ocean wave forecasts.3
Typically, the available meteorological data as gathered by land weather stations, by ships at sea, by balloon borne radiosondes, by aircraft measurements of winds, by the study of low level cloud movements from geostationary spacecraft imagery, and by remote atmospheric soundings from spacecraft are all processed every 12 hours to provide an analysis of the structure of the atmosphere. This initial value specification is then integrated forward in time to produce numerical weather predictions (prognostic charts). The most recent prognostic field is often used at the start of the new analysis cycle, and the various techniques employed pften "remember" features of the past that are 12 hours old, or older, especially in data sparse areas. Also typically, a forecasting center uses from 90 to 95 percent of its computing capacity to calculate the future weather.