154 MALKUS [chap. 4 



The storage in this 90-m deep oceanic layer was computed from the relation 



S = cp^\—dv, (29) 



where c and ptv are the specific heat of water and its density, both taken as 

 unity, t is time and v denotes volume. The average temperature, T, for the 

 layer 0-90 m was determined from the depth profile for each month by the 

 method of equal areas. Then dTjci was approximated by taking two-month 

 overlapping temperature differences centered at each month; finally each 

 result was multiplied by the volume of water (area of ellipse times 90 m depth 

 of layer). The resulting seasonal march of S is shown in line 9, Table XI, 

 expressed per unit area to compare with previous figures. The storage is zero 

 in late September and late February. The oceanic heat content decreases rapidly 

 in fall and early winter, while the increase in spring is more gradual. The 

 average storage rate from peak summer to peak winter is about — 2.5 kg cal cm"^ 

 per month ( — 85 cal cm""^ day~i) ; from winter to summer the average is 

 + 1.8 kg cal cm~2 per month ( + 61 cal cm~2 day^i). Clearly 8 is in general not 

 a negligible term in equation (1) and, according to these results, may amount 

 in some months to 25-50% of the dominant terms [Qe and R). 



Colon's estimates of 8 compare favorably with most figures available in the 

 literature. Gabites (1950) obtained storage values for oceanic areas by latitude 

 belts, assuming that the surface temperature variation is maintained for the first 

 25 m with a steady decrease from there to zero at 125 m. His values for the belt 

 10°-20°N are almost identical with those in Table XI, as are those of Fritz 

 (1958) computed from bathythermograph records. Much larger values, however, 

 were obtained by Pattullo (1957) from the identical data as used by Colon; 

 the reasons for the discrepancy lie in different analyses. Pattullo treated the 

 sample temperatures as totally representative, which as mentioned showed a 

 larger annual range than did climatic charts, and she also assumed a much 

 greater depth of the seasonal cycle. Her rates of storage are about double those 

 presented here. This magnitude uncertainty is probably inherent today in 

 storage calculations. Nevertheless, the inconsistency of the higher values with 

 transfer formula results and joint budget requirements leads us, for the present 

 anyway, to place somewhat greater confidence in Colon's determinations. 



Estimation of the oceanic heat-flux divergence, Qvo, in the Caribbean is 

 simplified by the existence of one main, well-defined current which crosses the 

 ellipse approximately from east to west. The following equation, therefore, could 

 be used to calculate Qro '■ 



Qro = C r 31 hy TydAy-C j Mhe TedAc 



(30) 



where Mhy and Ty are the horizontal water-mass flow and temperature across 

 the Yucatan Channel ; the Mi,e and T^ denote the same ]mrameters on the east 

 side of the Caribbean. The A's, denote the lateral areas aross the outflow and 

 inflow boundaries respectively. The nature of the available data was such that 



