OCEANOGRAPHIC CLIMATE OF HAWAIIAN ISLANDS REGION 403 



short, since the shape of the characteristic heating ^.j^^^^^^^ fl_, ^^^j^ therefore be too high and re- 



curves reflects the oceanograpliic clnnale, one can pCpS 



say that, during 1955, Oahu was located within an suit in excessive heat advection. Neglecting the 



oceanographic cliuiate normally to be found to the effects of diffusion in the siniphfied heat budget 



soutli, antl, during 1957, Oahu was located within equation would also tend to increase the computed 



an oceanographic chmate normally to be found to advection. 



the north. The assumptions, then, would affect the magni- 



During 1956, the dip in the heating curve which tude of advection, but not necessarily the advec- 



signals the end of the warm advection period tion periods. This view is supported by the 



occurred during March-April, approximately one consistency of the advection periods throughout 



month earher than in the mean curve. The the survey region, despite data of less than 



significance of two peaks during the May to July desirable quality. 



period rather than one as in the mean Koko Head In the discussion of the characteristic heating 



curve is not apparent. The dip in the heating curves obtained from the Oahu monitoring station 



curve during August of 1956 to below and, subse- (fig. 26), deviations of the rate change of tempera- 



quently, to above the mean curve can be inter- ture from the mean were explained in terms of 



preted as more pronounced cold advection. advection. This implies that year-to-year changes 



From March to October 1958, the shape of the in the heat exchange and diffusion are insufficient 



heating curve resembled the mean curve except to account for these deviations, 



that the March-April peak, the April-May dip. In the vicmity of the Hawaiian Islands, exam- 



and the June peak, occurred approximately one ination of the vertical temperature gradient below 



month later. the mixed surface layer suggests that the stability 



During the autumn and winter months from remains relativety constant throughout the ye&T. 



October 1955 to March 1958, the dual dips in the Significant changes ui the vertical diffusion are 



heating curve are as apparent as in figure 22C. therefore unlikely, even though diffusion may not 



They reflect the cold advection peaks of No- be negligible in heat budget considerations. The 



vember-December and Januarj'-February, and calculated heat losses from the sea surface (evapo- 



in each case deviate from the mean Koko Head ration, back radiation, and conduction of sensible 



curve. heat) seasonally vary bj' about 5 percent. On a 



Thus, certain changes in the surface temperature year-to-year basis, then, the variation of these heat 

 can be interpreted in terms of physical processes. losses would probably be less than 5 percent. 

 Figure 26 also shows that deviations from the Remaming is the incident radiation at the sea 

 mean characteristic heating curve can be large, surface which, due to cloud cover, can vary con- 

 exhibiting features which can be found in the siderably. Maxuuum deviations in the charac- 

 characteristic curves of adjacent areas. teristic heating curve due to this cause would occur 



Before closing the discussion on heat advection, during Ma}', June, and July, when insolation is at 



the vahdity of the results will be examined. a maxmimn and the depth of niLxed layer is at a 



Although shapes of curves and times of advection minimum. To estimate reasonable deviations, 



peaks and dips in the rate of change of tempera- assume that the year-to-year variation in the mean 



ture were stressed, it is useful to review the monthlj- cloud cover is less than the seasonal range 



magnitude of advection encountered. In the of one tenth. On this basis, deviations of less than 



previous section displacement velocities of up to 0.2° C. mos.~' in June and less than 0.1° C mos.~' 



one knot were calculated, which seem high for in December can be expected. 



the mean conditions discussed here. However, Deviations from the mean characteristic heating 



on the basis of the assumption used in the deriva- curves in figure 26, which were interpreted in tenns 



lion of the simplified heat budget, higli advection ^^ advective changes, were as follows: 

 values are to be expected. First, it is probable 



that the actual depth through which heat is dis- jAp''" ^^^^ + Jjp° C '„"or-''' 



tributed, or the "effective" mixed layer, is greater ivw-mb^TlOSS-January 1956."//- +0.S° C. mos/' 



than the measured mixed layer so that the com- March-April 1956 -0.3° C. mos.-' 



puted heat exchange in terms of temperature Septumbor 1956 +0.2° C. mos.-' 



