In an operational model, however, a factor for 

 the cloudiness could be put in. 



The visual quality of the fit in figures 4.3 and 

 4.4 is backed by statistical evidence. The average 

 deviation of the prediction from the data in the 

 sea surface temperature was slight. The predic- 

 tion exceeded the data by 0.02°C on the average, 

 but the hypothesis that the deviation between the 

 two was zero was statistically significant in 

 excess of the 75% level. 



The average deviation of the prediction over 

 the data is mixed layer depth was 1.04 meters, 

 and here the results were significantly in excess 

 of the 90% level. 



2. Wind Driven curre)itf< i)i the Dii.ved layer 



The data used to test the current portion of the 

 model was from Mooring 280 set by Woods Hole 

 Oceanographic Institution in 1968. The mooring 

 was set at 39-10 N, 70-03 W. The instruments 

 provided wind data at 2 meters above the sea and 

 current data at a depth of 12 meters for 48 days. 



In using this data set, several problems were 

 encountered. First, a sea current existed at the 

 mooring throughout the record, tending approx- 

 imately 045° - 065° magnetic, the velocity being 

 about 20 cm sec"'. Although there were strong 

 inertia] oscillations throughout the record, they 

 were difficult to separate from the overall record 

 because of the strength of the current and the 

 noise in the data set. For these reasons, I was 

 unable to separate clearly the inertial oscillations 

 from the overall record using basic Fourier 

 Analysis. I did not have at my disposal the com- 

 plex filtering schemes which would have been 

 necessary to separate out the oscillations. (Mil- 

 lard, 1978) 



A more fundamental problem exists with the 

 testing of a model of this sort with moored cur- 

 rent data. My equations are Eulerian equations, 

 because the assumption of the Rossby number 

 being less than one eliminates the non-linear 

 terms. In truth, however, the model is is Lagran- 

 gian. As the water particle moves, it is acceler- 

 ated by various forces, and the water column in 

 question is changed by wind energy and insola- 

 tion. These charges are relative to the current 

 position of the water, not on the original position 

 of that water. The Eulerian equations can be ap- 

 plied in the manner I have because of the 

 assumption of zonal wind stresses. The data set 

 from a moored array is an Eulerian one, showing 

 the changes in the water moving past a fixed 



point. What occurs at this fixed point is not neces- 

 sarily dependent on the conditions that are expe- 

 rienced at the point of measurement. In other 

 words, the advection terms in the equations of 

 motion become important. The Eulerian data can 

 give a good first approximation only of the water 

 movement, since any strong constant current in 

 the area renders my assumption of the Rossby 

 number being less than one invalid. Moored 

 arrays would be useful and valid, however, in an 

 area where there is no constant current, with 

 horizontally homogeneous water and zonal winds 

 for some distance away from the mooring, such as 

 if the mooring were placed in the center of a gyre. 



Therefore, the fit obtained from this data is 

 tenuous at best. 



The only method I could use to extract some 

 reasonable inertial transport from the data was 

 to use a vector solution for the net sea current in 

 the area and subtract that net current from the 

 data set, thereby leaving the inertial component. 

 The tidal oscillations in the area are almost zero. 



To solve for this net current, I hindcasted what 

 the frictionless inertial period transport would 

 be, given the wind field in a certain data window 

 under study. I then took the overall data and 

 solved for the sea current. This frictionless solu- 

 tion may be used because we know that the ocean 

 is close to being "frictionless." I used the follow- 

 ing vector solution: 



NET 

 SEA 

 CURRENT 



INERTIAL 

 TRANSPORT 



DATA 

 SET 



Figure 4.5 Lagrangian view of solution for Net 

 Sea Current. 



The "net sea current" as obtained above was 

 subtracted from the data set to give the inertial 

 transport. This "net sea current" as solved for 

 above was not necessarily consistent over the 



13 



