The general character of the unsmoothed 

 graph is typical of the other indices also. 

 Figure 10 shows time series graphs drawn 

 from smoothed data for the six indices. It 

 is apparent that fluctuations with periods 

 on the order of a year, although less in 

 amplitude than shorter period fluctuations, 

 are not negligible. The tendencies toward 

 predominantly positive anomalies for the 

 California index and predominantly negative 

 anomalies for the Oyashio index subsequent 

 to 1947 (figs. lOd and lOe) reflect trends 

 noted in the discussion of decade mesins. 



CORRELATION OF WIND INDICES 

 WITH OCEANOGRAPHIC DATA 



In attempting to correlate wind indices 

 with oceanographic parameters such as tem- 

 perature and mass transport, it must be 

 borne in mind that the indices are rather 

 crude representations of the wind field. 

 Also, wind is only one of several factors 

 involved in the interaction of sea and 

 atmosphere and its role is not solely con- 

 fined to the action of surface stress, but 

 also it influences heat exchange at the 

 surface through evaporation and conduction. 

 Moreover, only part of the mechanical energy 

 transferred from the wind to the sea retains 

 its kinetic form as wind currents, since 

 variable amounts are converted into poten- 

 tial energy through vertical mixing and 

 expended in wave formation and turbulence. 



Fluctuations in wind stress set up 

 dynamic imbalances among the forces arising 

 from mass distribution, the earth's rota- 

 tion, gravity and inertia, which drive the 

 ocean circulation. The response of the 

 currents to the changing forces acting upon 

 them is complex and, due to the greater 

 inertia of the water, tends to lag behind. 

 Thus the wind indices can reflect only the 

 variations in wind energy available and it 

 should not be assumed at this stage that 

 the ocean currents are proportionately 

 accelerated or decelerated. 



Nevertheless, utilizing such indices 

 we may endeavor to gain greater insight into 

 the mechanisms linking meteorological and 

 oceanographic fluctuations and the effects 

 of these fluctuations on fish populations. 

 For example, the stronger northerly winds 

 over the California current, during the last 

 decade, (illustrated by the larger values 

 of the California index in fig. 8a, page 11) 



have been tentatively related by Reid, 

 Roden and Wyllie (1958) to subnormal ocean 

 temperatures during the same period through 

 advection and upwelling. It has been hy- 

 pothesized that the latter is closely asso- 

 ciated with the failure of the sardine fish- 

 ery off central California which occurred 

 also during this period. Consequently, keen 

 interest was aroused by a reversal of these 

 conditions during 1957 and 1958 which were 

 characterized by relatively weak average 

 winds, warm water temperatures and evidence 

 of sardine spawning observed north of Point 

 Conception for the first time in several 

 years (CCOFI Progress Report, 1958). 



Another example is given in the follow- 

 ing description of a correlation test between 

 wind indices and sea surface temperatures. 

 This test was stimulated by a recent paper 

 by K. S. Ketchen (1956) who examined sea 

 surface temperature at Triple Island, B.C., 

 as a factor influencing survival of young 

 lemon sole in Hecate Strait. Comparing 

 year-class strength with mean temperatures 

 for various 2- and S-month intervals, during 

 the time of year when the pelagic stage of 

 the species occurred, he obtained correla- 

 tion coefficients from -0.70 to -0.90. The 

 explanation postulated for these strong 

 negative correlation coefficients was that 

 growth rate during the pelagic stage is 

 inversely related to sea temperature. When 

 the temperature is cold, the larvae are 

 carried for a longer period by the northward 

 current through Hecate Strait which results 

 in greater numbers being deposited on the 

 rearing grounds. Ketchen suggested that the 

 observed relationship between water tempera- 

 ture and brood strength might be dependent 

 on the wind-induced current. 



The ocean currents in the vicinity of 

 Triple Island are generally northward, and 

 the mean isotherms of sea surface tempera- 

 ture during winter and spring are roughly 

 perpendicular to the coastline. Considering 

 these facts, it might be expected that the 

 sea surface temperature in this area should 

 be related through advection to the north- 

 ward component of geostrophic air flow. 



Pressure differences at location 8 

 were chosen from the network in figure 1 

 (page 3) as the most representative index 

 of north-south geostrophic wind upstream 

 from Triple Island (for location see fig. 4, 

 page 6). Monthly mean sea surface tempera- 

 tures at Triple Isleuid have been compiled 



14 



