10-- 
5-- 
0-- 
SHIPBORNE DATA 
GEOSAT CYCLES 
Figure 8. Cruise data on height anomalies compared with the sequence of satellite data 
extending from October 1987 to October 1988. The cruise data (heavy lines) for 
May and June/July 1988 are plotted over the corresponding satellite data. 
Measurements were taken at locations indicated by the vertical arrow-heads. The 
satellite data (faint lines) have been interpolated from measurements along 
satellite tracks that cross line P at the positions of the vertical lines. 
11 to 16, before moving westwards, as shown in Figure 4. An 
eddy (not shown in Figure 5) later formed near the same 
location in cycle 29 and moved northwestwards for the 
remainder of the data sequence. 
During the time of the Geosat ERM, the only place where 
hydrographic data were collected repeatedly in the Gulf of 
Alaska was along line P. This extends from the coast at the 
southern end on Vancouver Island out to the site of Ocean 
Station Papa (50°N, 145°W). A few other observations were 
made along line R, from the southern tip of the Queen 
Charlotte Islands out to station Papa. Both lines are in the 
southern area of the Gulf of Alaska where satellite data 
(Figure 2) and earlier ship data (Figure 6) indicate that 
eddy activity is relatively low. 
Figure 7 shows the time series of geopotential anomalies 
interpolated from data collected on 12 cruises along line P 
in the years 1986-1989. The largest observed anomaly, with 
a height of 10cm, was observed at a single (but duplicated) 
station on the June/July 1988 cruise and was not observed 
on the May or November 1988 cruises. 
Figure 8 shows data for these two cruises compared with the 
sequence of satellite data interpolated along line P. The 
heavy lines show ship measurements from the two cruises 
with the mean slope removed. The May data show no 
anomaly higher than about 3cm, the June data show the 
10cm anomaly in duplicated measurements at station PI8. 
The dotted lines show the sequence of Geosat height 
measurements from October 1987 (top) to October 1988 
(bottom). The mean dates for some of the satellite cycles 
are indicated by the arrows, and the ship data is plotted over 
the corresponding satellite data. 
On the June cruise the eddy gave a 10cm amplitude at one 
station and amplitudes less than 3cm at adjacent stations, 
40nm (72km). Geosat tracks cross line P at intervals of 
about 60nm (105km) and give amplitudes of about 5cm on 
each side of the peak. 
This observation confirms that the Geosat tracks are spaced 
too widely for optimum detection of height anomalies. The 
low anomaly amplitude observed in both ship and satellite 
data lead to an estimated 5cm detection limit for these 
eddies with the present satellite data. The precision of the 
ship data is estimated (from repeat observations) as about 
lcm. 
5 GEOSAT OBSERVATIONS OF SURFACE 
WAVEHEIGHT. 
An essential step in deriving a precise range from Geosat 
altimeter data is the fitting of the observed shape of the 
leading edge of the return signal, averaged over many 
pulses, to the expected shape for different values of satellite 
attitude, wind speed and significant wave height (Flayne and 
Hancock, 1990). The resulting estimates of waveheight are 
immune from many sources of error, since they depend only 
on the shape of the pulse and are independent of the 
received signal strength, which varies with atmospheric and 
ionospheric attenuation, and with surface reflectance, and of 
propagation delays. For an instrument such as that launched 
on Geosat, the narrow pulse width easily resolves 
waveheight changes of 0.5m in Hi /3 (significant wave 
height or 4 standard deviations of the measured height 
distribution). 
102