Wind Speed and Stress Over the Ocean: Scatterometer versus Surface Measurements

Abstract

Once routine, global wind measurements of known accuracy become available from satellites, both oceanography and meteorology will be relieved of a great observational burden. The SEASAT experience demonstrated that its scatterometer system, SASS, could do the job, but the following problems still exist:

• i) The derivation of wind speed or stress requires a purely empirical algorithm.

• ii) There is evidence of systematic biases in the SASS winds.

• iii) The SASS algorithm is based on mid-latitude surface measurements during summer.

• iv) Differences in SASS and surface winds are sometimes attributed to spatial versus time averaging.

• v) It seems likely that backscatter depends on more parameters than have been considered in retrieval algorithms.

It is argued that at this stage, the next likely scatterometers, on NROSS (KU-band similar to SEASAT) and certainly ERS-1 (C-band), will both require surface measurements on which to base or improve their retrieval algorithms. Since the ocean surface is still a “black box” as far as the backscatter relationship to the wind is concerned, it is further argued that one algorithm is required to derive wind speed and another to derive wind stress. A measurement program is then outlined, which would improve upon the present state of scatterometer anemometry, by providing spatially averaged wind speed and stress at a variety of oceanic sites, together with the necessary supplementary data.