Figure 6. Geopotential anomaly at the sea surface relative to the 1000 decibar surface, 
plotted in J.Kg" 1 from ship data collected in a cruise May - July 1961. The 
gradients deduced from the contours gives the surface geostrophic current, which, 
will be along the contour lines. Sea level height in centimeters are given by J.Kg'* 
multiplied by 10. 
Figure 7. Geopotential anomaly at the sea surface relative 
to the 1000 decibar surface, plotted in J.Kg‘1 from 
cruises along line P in the years 1986-1989. Data 
for May 1988 and June/July 1988 are plotted in 
Figure 8. 
These speeds and direction of movement agree with 
predictions for baroclinie Rossby waves and with a 
correlation of time series measurements along line P (White 
and Tabata, 1987). The magnitudes and distribution of the 
anomalies shown here compare well with the available ship 
observations, as discussed below. 
4 COMPARISON WITH SHIP DATA 
Previous studies of the height anomalies in this study area 
have been reported by Tabata (1982) who showed a series 
of maps of geopotential anomaly relative to 1000m deduced 
from hydrographic (bottle cast) station data measured on 
US/Canada ship cruises between 1956 and 1962. Stations 
covered an irregular grid with a spacing of 100 - 200km, 
concentrating on the area under the satellite tracks in 
Figure 1, but extending over the whole area shown. Typical 
cruise duration was about 30 days, but individual features 
tended to be covered in a period of a few days. 
Figure 6 shows contours of geopotential anomaly at the sea 
surface relative to 1000 decibars (about 1000m), 
interpolated from stations occupied in the June 1961 cruise. 
They equivalent heights have similar magnitude and 
distribution to those in Figure 4, though the satellite data 
show more structure. The shape of the sea surface deduced 
from the ship measurements is clearly affected by the 
positions of ship stations. The largest height anomaly 
corresponds to the most active eddy reported by Tabata 
(1982). This appeared in about 50% of observations at a 
position off Sitka, Alaska, centered near 57°N, 138°W, and 
with a diameter of 200-300km. The other major height 
anomalies in the ship data (Figure 6) appear 300Km west- 
southwest of the "Sitka" eddy, further northwest close to the 
Alaska coast, and about 100km west-southwest of Cape St 
James (the southern point of the Queen Charlotte Islands),. 
Similar features also appear in Figure 4. 
The satellite data show that an eddy is often present at the 
"Sitka" location, but that after forming at this location the 
eddies tend to move off to the west with a velocity of 1- 
2cm/s. An eddy is present in the generation area for cycles