This paper presents a comparison of Geosat radar altimeter-derived wind speeds and significant wave heights with those measured by buoys in the National Data Buoy Center (NDBC) network. Measurements from a subset of 43 buoys moored in coastal regions and deep ocean, including the North Pacific, North Atlantic, and Gulf of Mexico, were examined. Data from other buoys within the NDBC network were not included because of their proximity to land. Seven months of altimeter comparisons were obtained during the Geosat geodetic mission for the periods May-August and October-December 1985. Only Geosat data within 150 km of buoy locations were accessed; this resulted in 1166 wind speed and significant wave height pairs. An error analysis was performed to better understand the differences between altimeter and buoy-derived results and to establish consistency between the two sets of measurements. The following error sources were identified and quantified in terms of their impact on wind speed and significant wave height uncertainties: (1) inaccuracies due to buoy instrumentation, (2) errors resulting from temporal and spatial separations, (3) errors arising from averaging of the buoy data in time and averaging of altimeter data over its footprint, and (4) altimeter instrumentation errors. Four algorithms relating altimeter radar cross section to ocean surface wind speed were investigated. The Brown algorithm (Brown et al., 1981), and the smoothed form of the algorithm of Brown et al. (Goldhirsh and Dobson , 1985) were found to give the smallest rms errors. These were 1.7 m/s and less for a maximum altimeter track-buoy separation of 50 km for most cases considered. This uncertainty agreed with the error obtained from the error analysis. Significant wave heights measured by the altimeter were found to be in agreement with buoy-derived values to within 0.5 m rms. This value was also found to be in agreement with that derived from the error analysis. The measurement goals for Geosat were 1.8 m/s rms for wind speeds from ! m/s to 18 m/s and 0.5 m rms for significant wave height or 10%, whichever is greater (Frain et al., 1985). These goals were met for the range of wind speeds compared. frames needed to produce yield predictions for fishery management and for increased fish catches.Radar remote sensing requires an understanding of the relationships between the signal the instrument is measuring and the physical characteristics of the medium from which the Copyright 1987 by the American Geophysical Union.Paper number 7C0530. 0148-0227/87/007 C-0530505.00 signal backscatters. In many cases these relationships can be established by quasi-empirical means with the aid of in situ measurements. The determination of ocean wind speed and significant wave heights from Geosat radar altimeter measurements represents such a case.Two parameters which can be derived from the altimeter measurements are ocean surface wind speed and significant wave height (SWH). Wind speed is related to the backscattered power, and the SWH is determined from th...

Simultaneous measurements of refractive index and radar reflectivity have been made in the clear atmosphere. The theoretical relationship between the refractive index spectrum and reflectivity de· scribed by Tatarski was used to compute reflectivity from the refractive index data. Excellent agree· ment is shown between the computed and measured reflectivity. These results indicate, (1) reflections from the clear atmosphere can be described quantitatively in terms of the refractivity spectrum, and (2) the -j refractivity spectrum well represents the atmosphere for these data out to limiting micro· scales smaller than 5 em.

Wind blowing across the ocean surface generates small, centimeter scale roughness. The radar cross section (RCS) measured by spaceborne, nadir‐looking altimeters operating at about 13.5 GHz is responsive to this roughness. Present altimeter algorithms use RCS to infer wind speed. We compare Geosat altimeter estimates of wind speed and nearly coincident estimates from NOAA data buoys to determine whether altimeter algorithms can be improved by using more of the information available from altimetry. We find that a marginal improvement in wind speed retrievals can be obtained by including additional RCS information and significant wave height in the retrieval algorithm. Perhaps most important, results also suggest that wind‐wave growth is suppressed in the presence of preexisting ocean swell.

Fine‐scale measurements of the ocean surface have been made from stereo photographs. By taking these photographs from a height of about 20 ft above the sea, vertical resolutions as small as 0.3 inches were measured. The wave‐number spectra obtained from the photographs verify the Phillips k−3 prediction for the behavior of ocean‐wave spectra at high wave numbers.

No abstract