halocline, as noted above, is a measure of the 

 minimum mixed-layer temperatm-e of the previous 

 winter. By definition, minimum temperature 

 occurs at the beginning of the heating season. 

 Seaward of upwelling influence, distribution of 

 July temperatm-e at 100 m. thus approximates the 

 initial field of temperature. Temperature variation 

 at this depth did not generally exceed 1° C. over 

 the plume and offshore provinces, either spatially 

 or from year to year (Owen, 1967b). Compared 

 with spatial and temporal temperature changes in 

 the near-surface layer, this variation was small 

 enough to permit the assumption that none of the 

 variation of July mixed-layer temperature was 

 due to initial temperature differences. 



Two sources of variation remam : thermal diffu- 

 sion and mixing depth. Both may be expected to 

 be substantially affected by salinity gradients at 

 plume-sea interfaces, so that their effects on tem- 

 perature distribution may be treated collectively 

 as the "plume effect." It is this influence to which 

 the recm-rent ridge of higher temperatures was 

 ascribed in the previous section. 



On this basis, the validity of the proposed plume 

 effect can be examined by comparing the pattern 

 of plimie disposition with the pattern of near- 

 surface temperature change from the respective 

 winters. Average values of salinity gradients 

 normal to the plume edges were estimated from 

 figure 4 (horizontal gradient estimates) and figure 

 8 (vertical gradient estimates along lat. 44° N.) 

 and entered with values of maximum within- 

 plume temperature change in table 2. The asso- 

 ciation of larger temperature change within the 

 plume with large plume-edge gradients is clear 

 from this table as well as from the figures them- 

 selves. 



To examine the plume effect further, mixed- 

 layer temperature change, AG, was plotted as a 

 function of surface salinity for each July of the 4 

 years on the basis that surface salinity inversely 

 represents the degree of "presence" of the plume 

 and hence the magnitude of its interface gradients. 

 Sets of paired data from each hydrographic sta- 

 tion for each year were subjected to Spearman's 

 rank-difference correlation test (Tate and Clel- 

 land, 1957) ; testing showed significant inverse 

 relation at levels of p<0.001 for 1961, p<0.001 

 for 1962, 0.1<:p<0.2 for 1963, and p«:0.001 

 for 1964. Differences between these levels of sig- 

 nificance appear to depend on the degree of plume 

 development and salinity gradients. 



Validity of the assimied sea surface heat- 

 exchange field must remain an open question. It 

 is difficult, however, to see how the particular 

 field of heat exchange could arise to produce the 

 closed curves of AG in the plume regime in the 

 absence of oceanographic mechanisms discussed 

 previously. The lack of advective heat change in 

 the plume and offshore provinces is certain to be 

 an approximation: ridgelike dynamic topography 

 indicates possible advective transport of warmer 

 offshore waters of more southern origin in 1961 

 and 1962. That this approximation is sufficiently 

 good for showing the plume effect, however, is 

 indicated by the differential heating in the plume 

 in 1964, in spite of a flow pattern that suggests 

 advective transport of colder water from the 

 northwest. Advective effects thus appear to be 

 masked by local change. 



SUMMARY OF OCEANIC PROCESSES AND 

 VARIATIONS OFF OREGON AND WASHINGTON 



The region studied off the coast of Oregon and 

 Washington is divisible in summer into three 

 oceanographic provinces above and two provinces 

 below the main halocline. This division is based on 

 discontinuities in the distributions of variables 

 that denote discontinuities in the influence of 

 physical mechanisms, so that the oceanographic 

 processes may be considered similar within and 

 dissimilar between provinces. These processes — 

 upwelling in the nearshore province, modification 

 by land rvmoff in the plume province, and net 

 dilution of siu"face layers in the offshore province — 

 produce recurrent distributions of heat, salt, mass 

 and, together with biological processes, oxygen. 



Annual variations in the distribution of vari- 

 ables are attributable to changes in wind field, 

 in fresh-water discharge from the land, and in 

 advection. The variations are not sufficiently 

 large, however, to obscure the basic patterns 

 generated by dominant processes. 



The balance of processes off Oregon and Wash- 

 ington is atypical in one respect. Whereas advec- 

 tion strongly influences recurrent patterns of near- 

 surface distributions in oceanic regions lacking 

 large fresh-water sources, its role in the study area 

 is limited to disposition of land runoff. The low- 

 salinity plume itself establishes the conditions for 

 differences in local processes that generate the 

 recurrent distribution patterns described here. 



OOEANOGRAPHIO CONDITIONS IN NORTHEAST PACIFIC OCEAN 



521 



