Local fluctuations may obscure the average circulation completely, 

 especially when the observations are nnade over a short period of time. 

 In one convenient method of indirect current measurement the difference 

 of the vertical distribution of mass (specific volume) between two sta- 

 tions is used to calculate the component of the current vector normal 

 to the line between the stations. There are some essential assunaptions 

 which must be satisfied. The region under consideration must be sev- 

 eral degrees of latitude removed from the Equator, a steady state must 

 exist (no accelerations), and dissipative forces such as friction must be 

 negligibly small. Another problem in these computations is the proper 

 choice of a reference level, the surface along which no component of 

 gravity acts. It is usually assumed that at a depth where the specific 

 volume change of the water between stations is small connpared to that 

 in the surface layers, the isobaric surface is "level", i.e. changes are 

 so small that they will not materially change the computed currents in 

 the surface layers. To be precise, the computed surface currents are 

 relative to those at the reference level. ^' 



When the method of computing "relative currents, " based on the 

 "distribution of mass, " is applied to the Hawaiian Islands region, it is 

 assumed that the restrictions mentioned above hold. However, because 

 of the boundary effect of the islands, the steady state and the friction 

 assumptions may not be exactly satisfied, particularly in places of high 

 current speeds. Distortions in the calculated circulation pattern may 

 therefore exist, but they are not expected to materially change its prin- 

 cipal features. 



Figure 5 shows the geopotential topography^' in dynaunic meters, 

 of the sea surface referred to 600 m.Z' The outstanding feature is the 

 counterclockwise, rotary current motion or vortex centered at about 

 19 45'N. , 157 20'W. with speeds on the order of 1 to 2 knots. There 



— 



— A more detailed, non-mathematical discussion can be found 

 in Sverdrup et al. (1942:389-395). 



3/ 



— Also referred to as geopotential surface, dynamic topography, 



dynamic heights, or dynamic height anomalies. 



4/ 



— Strictly speaking the reference level in dynamic computations 



is an isobaric surface. However, under the assumptions used, this sur- 

 face is so close to an isobathic surface that, for simplicity, meters will 

 be used to designate the reference depth. 



