Observations of the California Undercurrent Off Washington and Vancouver Island

Observations of the California Undercurrent Off Washington and Vancouver IslandAuthor(s): R. K. Reed and D. HalpernSource: Limnology and Oceanography, Vol. 21, No. 3 (May, 1976), pp. 389-398Published by: American Society of Limnology and OceanographyStable URL: http://www.jstor.org/stable/2835273 .

Accessed: 17/06/2014 05:20

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

American Society of Limnology and Oceanography is collaborating with JSTOR to digitize, preserve andextend access to Limnology and Oceanography.

http://www.jstor.org

This content downloaded from 195.34.79.253 on Tue, 17 Jun 2014 05:20:37 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=limnoc

http://www.jstor.org/stable/2835273?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


Observations of the California Undercurrent off Washington and Vancouver Island

R. K. Reed and D. Halpern Pacific Marine Environmental Laboratory, National Oceanic and Atmospheric Administration, Seattle, Washington 98105

Abstract A detailed oceanographic survey was conducted in the Pacific off the coast of Wash-

ington and Vancouver Island in September 1973, and 12-h time-series measurements were made at eight sites. Data from the time-series stations suggest the presence of semidiurnal internal tidal waves seaward of the continental shelf. The resulting disturbances to the geopotential anomalies, however, had little effect on the derived patterns of geostrophic flow. During this survey, the surface flow relative to 1,000 db was generally southward, but a northward flow was present at intermediate depths along the continental slope. This northward flow appears to be an extension of the California Undercurrent, which remains an identifiable feature to at least 49?N but weakens along its path by mixing with adjacent water. An additional northward flow in the northern part of the area resulted from a cyclonic eddylike feature.

The main feature of the large-scale motion off the Washington and British Colum- bia coasts consists of the separation of the eastward flowing West Wind Drift into northward and southward components. Ac- cording to Uda (1963) and Dodimead et al. (1963), the surface baroclinic currents relative to 1,000 db are variable in direction and generally less than 5 cm s-1. At about 47?N, 128?W, Halpern (1972) made direct current measurements at four depths within the uppe'rmost 50 m for a 32-day in- terval (4 August-5 September 1971) and found mean vector-averaged current speeds of nearly 6 cm s-1 flowing toward the south- east. Countercurrents and eddylike features observed inshore of about 130?W by Fofo- noff and Tabata (1966) and by Ingraham (1967) seemed to be representative of the weak and variable flow. The' complexity of the physical oceanography of the coastal zone off the coasts of Washington and Van- couver Island is increased during periods of large runoff of freshwater from the Co- lumbia Rive'r and from the Strait of Juan de Fuca (Barnes et al. 1972).

At subsurface depths over the continental slope region off California, a current flows northward (California Undercurrent) under the southward moving California Cur- rent (Re'id et al. 1958; Reid and Schwartz- lose 1963; Wooster and Jones 1970). This

current has not been well defined off Ore- gon, presumably because of a lack of adequate observations. Off Washington and Vancouver Island, Dodimead et al. (1963) reported northward flow at 200-500-m depths. Ingraham (1967) and Cannon and Laird (1974) also indicated a weakly de- veloped California Undercurrent.

In this paper we de'scribe the structure of geostrophic, baroclinic currents computed from STD measurements made during Sep- tember 1973 off Washington and Vancouver Island. Emphasis was placed on resolution of the undercurrent. Since the undercurrent flows along the continental slope, a region known for the generation of internal tidal waves (Barbee et al. 1975; Rattray et al. 1969), we made a number of stations consisting of repeated STD casts to estimate the contamination of the geopotential topography produced by internal tidal waves (Defant 1950).

We thank Capt. W. D. Barbee and the officers and crew of the NOAA Ship Ocean- ographer for their efforts and interest during this cruise. J. R. Holbrook provided considerable assistance in the data prepa- ration.

Observations Vertical profiles of temperature and sa-

linity were obtained from the NOAA Ship LIMNOLOGY AND OCEANOGRAPHY 389 MAY 1976, V. 21(3)



390 Reed and Halpern 32, t301 12W52241

509 0o O 00 KM +t o

34 * 32 23-31 53<' 0 oe*..

36 37-45 46 49"5 56 74O>*

48' - 64 HINGTON 48? N 70 N~

96 94 850-93 84 69- 5

SEAMOUNT.4b 98 9350? 508 5 2 10 2 '

151-119 129

13400 0 '@,@*

46?- - s'1s 4 1|- 46? 132 130? 128'W 526? 524?

Fig. 1. Location of STD stations, 7-20 Sep- tember 1973. Open circles indicate the time-series stations. The 100- and 500-fathom (1 fm = 1.829 m) isobaths are also shown.

Oceanographer at 150 stations during 7-20 September 1973 in the region between 47?N and 50?N and between the coast and 130?W (Fig. 1). The measurements were made with a Plessey model 9006 salinity/ temperature/depth (STD) system. Scans of temperature, salinity, and pressure were recorded in digital format at 0.5-s intervals on a Plessey model 8114A digital data log- ger. Within the uppermost 200 m the in- strument was lowered at about 15 m min-', and approximately 5-10 scans were recorded each meter. Between 200 m and 1,500 m the lowering rate was increased to 30 m min-1. Calibration was accomplished by tripping a Nansen bottle at the bottom of every fifth cast; the reversing thermometers used were calibrated by the National Oceanographic Instrumentation Center, and salinity was determined on a laboratory in- ductive salinometer. We believe that the corrected data have random measurement errors of approximately +0. 020C, +0.01%o, and +4 m and no significant systematic errors. Summaries of the STD measurements at each station have been presented by Holbrook (1975).

Variability The baroclinic, geostrophic circulation

pattern will contain errors due to the quasi- synoptic nature of the survey and, as indicated by Defant (1950), to the internal variations of the temperature and salinity fields. Using only one ship we were unable to fulfill the requirement of a synoptic survey; therefore, our results are based on the assumption that the observed flow pattern did not contain significant fluctuations for time scales between a day and a fortnight. From temperature observations made by Halpern (1972), we expected the semidiurnal tidal frequency to be the dominant short-period fluctuation of the density field in this region.

During this cruise, eight time-series stations of 12-h duration were taken (see Fig. 1); casts were taken to nominal depths of 1,500 m every 1.5 h, except at stations 49- 55 where they were made every 2.0 h. The results from these time-series stations suggest that low-frequency oscillations were present within the pyonocline region at all sites. Representative examples of the observed temporal variability are shown for stations 14-22 and stations 85-93 in Figs. 2 and 3. At both sites the isolines of temperature, salinity, and sigma-t exhibit very similar behavior; they rise and fall in a fairly regular manner with a time difference of about 6 h between the high and low positions. Estimates of the maximum amplitudes gave values of 8-9 m, and below depths of 75-100 m there was generally some reduction in amplitude. Although these short data series do not permit any sophisticated analyses, the behavior is similar to what we would expect from semidiurnal, internal tidal waves (see Reid 1956).

It is of interest to examine the effects of these temporal variations of temperature and salinity on the geopotential anomalies.

Fig. 2. Variation in depth of isolines of temperature (?C), salinity (%c), and sigma-t at stations 14-22, 10 September 1973. The variations in geopotential topography (dyn m) for the 10-, 150-, and 500-db surfaces (referred to 1,000 db) are also shown.



Circulation off Washington 391

GMT GMT 04 08 12 16 20 04 08 12 16 20 l 1 - I I I I I l

14 I5 16 17 18 19 20 21 22 14 15 16 17 18 19 20 21 22 O l l l l l l l I I I I 1 1 1 1

32.2 0/0

20 - 120C

IOOC 9 2.4%o

400 1 ~ 4 -180C 32.6%o

60 -.0

a-0

10 80o 3300/o

10 T (?C) S (%/o)

GMT GMT 04 08 12 16 20 04 08 12 16 20

14 15 16 47 18 19 20 21 22 14 15 16 17 18 19 20 21 22 0) 1 1 l l 1 1.25 l l l l l l l l l

I0 I l.2; ; j I I I

24.5

20 1.23

80 _ _ 0.89 ~~~~~~~~150/1000 db _

40 -12

1 20045.5

I60-091

500/1000 db

120 04



392 Reed and Halpern

GMT GMT 12 16 20 00 04 12 16 20 00 04 I I I 1 1 1 1 1 1 1

0 85 86 87 88 89 90 91 92 93 85 86 87 88 89 90 91 92 93 O I I I l l l l l I I I I I

150C 20 32.20 /0

40- 120C 110C ~~~~~~~~3 2 4 0oo

EZL F o 160 -0 0 a-

0 I 8W .3

3 2.6 0/100 75C

80 -

32.8 0/0

100 ̂

T (OC)7C

.100 _ _ 0.45 - 500/ 100~~~~~~~~~330 0/00

1201 i .3A _

GMT GMT 12 16 20 00 04 12 16 20 00 04

85 86 878889 9091 92 93 8586 87 8889 9091 92 93 0 1 .33i I i T

20-131oiood 24.0

24.5 40 -12

E

'60 25.092E

25.5 150/1000b 80 -09

1000.5 500/1000

120 ~~~~26.0




32' 130? 128'W 26' 14' 32' 30' 128? W 1266 124' 30' 5C0 501 50C

500 10.0000 db~ ~ 500 30>, W | 150/lOOOdb **1-

1.32 0.92~~~~~~~~~~~~09 *00 1300 128 "'of'5,.320 0 0.90 *

4 * O * . N N 0904N0 0 4~~~21i'P 0 r&4.

48' - 2 48' 48' 48'AHI

N N N 0.9 N* * * ) (O

SEAMOUNT;,,. * * o 0; SEAMOUNT ;;; 0.94 0.9 \0

10/1000 db / 150/1000db 094 46' 1 30'7'~~~~~~-s 4.L~1.3

46-1 1 1 1 46' 46' 6? 130' 130'. 128'W 126' 104' 130' 130' 128' W 10'1 4 6

320 130' 128?W 126' 124 132' 130? 128?W 126? 124' Sc, ~~~~~~~~~~5bo. 150' Sc

5ntopo 000

to~~~~~~~~~50 1,000 00,72 etme Y3.Oe ice niaeth iesre ttos

VANCUVE VACUE

0 0~~~~~ COB0 0 58

*.0 ".0 0 0 0 0 0 0 70~~~~701

300/10004 db 500/1000 db

465 46' 46' 46'

132' 130' 128' W 126' 124' 13'130' 1281W 126' 124'

Fig. 4. Geopotential topography (in dyn m) of the 10-, 150-, 300-, and 500-db surfaces (referred to 1,000 db), 7-20 September 1973. Open circles indicate the time-series stations.

Figures 2 and 3 indicate that maximum and minimum values of geopotential anomaly generally occur near the times of lowest and highest positions (respectively) of property isolines in the pycnocline. The 10/ 1,000-db values have amplitudes of almost 0.02 dyn m, and the amplitudes decrease toward the deeper levels (referred to 1,000 db).

Temporal variations in geopotential anomaly undoubtedly contribute spurious patterns of geostrophic flow in this region. The standard deviations from the mean geopotential anomalies at the various sites were computed for a number of levels. All the deviations were appreciably greater

than that due to random measurement error, which should be less than 0.001 dyn m according to the method of Wooster and Taft (1958). If we assume a normal distribution, two standard deviations (2cr) represent the reliability, within 95% confi- dence limits, of a geopotential anomaly obtained from a single cast. To obtain estimates of flow, we must use the anomaly difference between station pairs; hence the reliability of an anomaly difference equals 2cr(2)(2)-/2, which can be compared with the relief of geopotential anomaly. The mean (based on all sites) reliabilities of anomaly differences for various levels were: 10/1,000 db-0.028 dyn m; 150/1,000 db-

Fig. 3. Variation in depth of isolines of temperature (?C), salinity (%0), and sigma-t at stations 85- 93, 15-16 September 1973. The variations in geopotential topography (dyn m) for the 10-, 150-, and 500-db surfaces (referred to 1,000 db) are also shown.




0.014 dyn m; 300/1,000 db-0.011 dyn m; and 500/1,000 db-0.009 dyn m. From the maps of geopotential topography (Fig. 4), estimates of the anomaly relief are: 10/ 1,000 db-0.16 dyn m; 150/1,000 db-0.09 dyn m; 300/1,000 db-0.08 dyn m; and 500/1,000 db-0.06 dyn m. Furthermore, even the relief of small-scale features such as the subsurface northward flow near the continental slope was generally at least twice as great as the reliability estimates. Also, a number of features were defined by data taken within 1 or 2 h of each other, where the effects of semidiurnal fluctuations should not be as large as the reliability estimates computed from 12-h series. Con- sequently, it appears that the geopotential topography was not seriously affected by the observed temporal variations, except perhaps well offshore where anomaly gradients are quite weak. Hence we will use the geostrophic relation to describe the circulation in this region.

Geopotential topography Maps of the geopotential topography at

various levels were prepared from these data based on a reference level of 1,000 db. To show flow along the continental slope, we have computed anomalies for stations with maximum depth between 1,000 and 300 m relative to the 1,000-db surface by' a method originally proposed by Jacobsen and Jensen and described by Fomin (1964). The difference in geopotential anomaly, referred to the deepest level common to two stations, was corrected by

A=h/2(81-82). (1)

Here h is the depth difference between the deepest level common to both stations and the 1,000-db surface, and 8, and 82 are the specific volume anomalies at the deepest common level for the two stations. The corrected geopotential anomaly difference was then applied to the anomaly at the outer station, referred to 1,000 db, to obtain an anomaly at the inner station relative to 1,000 db. Because A exceeded 0.01 dyn m at only two of the 17 station pairs where it was computed, the contribution to the geo-

potential anomaly gradients from this source is generally rathe'r small.

The geopotential topography at four levels, referred to 1,000 db, is shown in Fig. 4. Anomalies for those stations inside the 500- fathom (914 m) isobath have been ad- justed to 1,000 db by the method just described. The 10/1,000-db map reveals a weak, southeastward flow of roughly 5 cm s-l. Nearer the coast the flow is also southward; typical speeds are about 15 cm s-1, but in places speeds are appreciably higher. A number of eddylike features were present, as is common off California (Wylie 1966). The southward flow, especially near the coast, contrasts with Ingraham's (1967) findings of northward surface flow in both spring and fall 1963. Reed and Halpern (1973) reported southward inshore surface flow in September 1972, but a month later it was northward. Fofonoff and Tabata (1966) concluded that meridional flow in this region has an annual cycle: southward in summer and northward in winter. Too few comprehensive surveys have yielded data to provide verification of this conclu- sion, however.

The 150/1,000-db topography shows southward flow offshore but northward flow inshore. The northward inshore flow is complicated, however; between 48?N and 49?N it appears to result mainly from a large eddy, and the northward flow in the southern part of the area (presumably the California Undercurrent) appears to end near Juan de Fuca Canyon, where the 100- fathom (183-m) isobath cuts across the continental shelf and extends inside the Strait of Juan de Fuca. This suggests that there are two origins for the northward inshore flow: a southern one for that south of 48?N and a northern one for the flow north of 480N, at least for its more offshore portion. [The broad northward flow north of 48?N reported by Reed and Halpern (1973) may be comparable to this large offshore eddy rather than the California Un- dercurrent.] The 300/1,000-db topography (Fig. 4) reveals the same features as at 150 db, but here there is some evidence that the northward flow from the south (Cali-




fornia Undercurrent) continues along the continental slope all the way to 49?N. The large eddy is still present; thus the northward flow north of 48?N probably has two origins. The 500/1,000-db topography is almost completely identical to the map just discussed, except that the flow is somewhat weaker.

Estimates of the northward flow (Cali- fornia Undercurrent) have been made by estimating the volume transport of the 150/1,000-db layer between cle'arly defined maximum and minimum anomaly values on east-west sections. We do not wish to imply that these transports are absolute, however, because data from this cruise do suggest that weak, but discernible, motion exists at 1,000 db. The station-pairs used on the sections with adequate data are, from south to north, 135-139, 80-79, 56-58, 4-5, and 10-12 (see Fig. 1 for locations). Data from the two southernmost stations (zonal width = 50 km) indicated a northward transport of 1.6 X 106 m3 s-', but at 47.50N (stations 80-79; zonal width = 18 km) the flow was only 0.7 x 106 m3 s-I. North of 48?N the transport estimates are extremely uncertain since it is difficult to separate the boundaries of the flow of southern origin (California Undercurrent) from that of the northern eddy. The values obtained from the remaining station pairs from south to north are 0.2, 0.7, and 0.2 x 106 m3 s-1, respectively. None of these last estimates should be taken to have a high degree of accuracy; perhaps the only conclusions that can be drawn are that the transport is small and that it probably weakens from south to north in this region. Water mass analysis

Several investigators (e.g. Dodimead et al. 1963; Cannon and Laird 1974) have ex- amined the distribution of properties on constant density or salinity surfaces; they found that the warmest (most saline) water occurs near the continental slope and that temperature and salinity decrease toward the north, which was taken as evidence for the existence of a California Un- dercurrent. The same patterns are shown by our data. Off Baja California, Wooster

SA L IN iTY ( ?/o )

33 34 35 12 -,/ I"w

w STA. 80 l

a:STA 96 15 0 SM Pacific < / Equtoril

// j ,5om /

4 _6/ //504. /

w~~~~~

/ ;7 2~~~00 m_.

U / / /Pacific 2~S / / ubrcti 1500m

2i / / ST.8 /

Fig. S. Temperature-salinity diagram showing curves for Pacific Equatorial and Pacific Subarctic water and two of the station curves used for deter- mining the percentage of Pacific Equatorial water. The dashed lines indicate sigma-t density.

and Jones ( 1970) showed that the undercurrent could be detected by extrema in the vertical profiles of temperature and salinity. No such systematic features were revealed by the STD data from this cruise though.

We felt, however, that some' type of nu- merical water mass analysis might be of value in exarining the C'alifornia Under- current. Since this is a transition region with no ne'arby distinct water types, the use of traditional mixing diagrams or box models would not be likely to prove useful. If we assume that mixing occurs along sigma-t surfaces, however, we may deter- minin the percent composition of a water mass in the vertical at a site. Tibby (1941) analyzed the waters off California by as- suming that they were a empixture of the Pa- cific Equatorial and Pacific Subarctic water masses, and Ingratam (1967) used the same criteria for a section near 47tN off Washington. This appears to be a safe assumption because the waters entering the region off Washington from the west seem to come from well north of the North Pa- cific wentral water mass (Dodime'ad et al. 1963; Fofonoff and Tabata 1966).

We therefore plotted the Pacific Equa-




130?W 129? 128? 127? 126, 1251W

96 95 94 85 84 83 82 8180 79 6, ~~~~I I I I I

35 2 =303

- -

500 - -73< X20

E -20 _25" 30

1000 L

25

1500 1

Fig. 6. Vertical section of the percentage of Pacific Equatorial water along 47.5'N, 15-16 September 1973.

torial and Pacific Subarctic water masses (as given by Pickard 1963) on a temperature-salinity (T-S) diagram and plotted the T-S curves for individual stations (Fig. 5). The percentage composition of equatorial water was then determined for the various stations for selected depths, and vertical sections were constructed (Figs. 6 and 7). The section at 47.5?N, which is approximately normal to the coast, indicates a general increase of equatorial water toward the coast; the east-west differences at the various levels are all about 10%. In addition to east-west differences, there is a consistent pattern along the vertical: the percentage of equatorial water is high at 200-300 m, decreases to a minimum at 500 m, and increases to a second maximum at 1,000 m. These features generally agree with the zonal section shown by Ingraham (1967), although the deep maximum seems somewhat deeper in his data, and we con- clude that the water of more southerly origin was nearer the continental slope.

Figure 7 is a vertical section that essen- tially parallels the coastline. Stations were used along the continental slope at water depths of about 1,000-1,500 m, where the flow of the California Undercurrent appears to be strongest. The section has been extended about 100 km south of the Colum- bia River by using station 391 (24 August 1973) of the Coastal Upwelling Ecosystems Analysis program (Holbrook and Halpern 1974). The vertical structure along this section is similar to that in Fig. 6. Along

45?N 46? 47? 48' 49' 50'N

391 137 129 144 80 70 58 4 12 I 1 135 11

35_ 3 >35 -35-

500 - <30

02 '=- 353 1000 - 'A 1C440

1 500-

Fig. 7. Vertical section of the percentage of Pacific Equatorial water along the presumed northerly path of flow. Station 391 was taken on 24 August 1973, and the others were taken between 7 and 20 September 1973.

the various levels, there generally is a decrease to the north in the percentage of equatorial water, but the largest differences are mainly between the two southernmost stations. To obtain a quantitative estimate of the north-south differences, we determined the equivalent thickness for equatorial water at stations 391 and 12 by inte- grating the area under a depth-percentage curve (from 150-1,000 m) and expressing the value in depth units. At station 391 the equivalent thickness was 11% greater than at station 12, suggesting that the water is altered by mixing as it flows north.

How can we interpret the percentage distributions shown in the two sections, especially their vertical structure? We must be cautious in attributing all the observed differences to advection because the derived distributions are highly sensitive to the boundary conditions assumed. That is, idealized T-S curves have been taken to represent the Pacific Subarctic and Pacific Equatorial water masses, but the T-S curves of the actual source waters may be somewhat different than these, which could cause the absolute magnitude of the percentages, as well as their differences in the vertical, to be in error. Also, strong vertical motion or mixing, as perhaps within an eddy, could cause local alteration of the T-S characteristics. Finally, in the deeper water large percentage differences could result from rather small differences in val-




ues and in the imprecision in measuring them. Notwithstanding these consider- ations, the main features seen in the vertical se'ctions are consistent enough, and the differences large enough, that we believe them to represent features relevant to circulation in this region. As we have noted, Ingraham (1967) presented a zonal distribution similar to our section at 47.5?N, and he concluded that southern water flowed northward from 200 to 400 m and from 1,000 to 1,300 m and that in between these layers subarctic water flowed from the west. We invoke a somewhat different explana- tion here, however, because we do not find any very marked differences in the geostrophic flow from 150 to 1,000 db.

We consider first the vertical section along the path of flow (Fig. 7). The rela- tively high percentages of equatorial water from 150 to 300 m appear to represent the most intense flow of the California Under- current, which we and others have concluded is at this level. The fact that lower values occur below 300 m does not imply that the flow has changed direction because the 500/1,000-db geopotential topography from these data clearly indicates that this is not the case. We suggest that the' mech- anism is that the observed weakening of the flow allows more time for lateral mixing with adjacent subarctic water along the path. At 1,000 m the increase in percentage of equatorial water does not appear to be related to any marked change in flow direction according to the 1,000/1,500-db topography (not shown) for this cruise. We sus- pect that the increase may result from a possible decrease in lateral mixing as a result of the much smaller horizontal gradients of temperature and salinity at 1,000 m compared to 500 m (Barkley 1968). If the decrease below 1,000 m is indeed a real feature, a plausible, but very tentative, ex- planation that comes to mind is that this represents a southward return flow of the deep water below (see Reid 1973 for discussion of this feature).

How may we account for the distributions in the zonal section (Fig. 6)? It is obvious that the percentages throughout

the water column are highest near the coast, which is in keeping with the observed northward flow of the California Under- current. Offshore, the vertical distribution of the smaller values here may be explained in the same manner as for Fig. 7 if we rec- ognize that the source waters here are the eastward-flowing West Wind Drift, which mixes with the colder, fresher water to the north (see Fig. 4). The one new feature clearly revealed in Fig. 6 is the very low values offshore at 150 m, compare'd to 200 m; this probably is the result of a pro- nounced southerly component of flow near the western part of this section at 150 db (see Fig. 4), but at this level surface effects might be significant. Although the vertical sections at first looked complex, they appear to be the result of plausible processes, even though the mechanisms cannot be rigorously defended. This analysis doe's provide additional support for the circulation inferred by the geostrophic relation.

Discussion The data from this cruise have revealed

some details of the undercurrent in this area not previously established. First, it had a vertical extent probably in excess of 500 m and a volume transport of perhaps 1 x 106 m3 S-1. Se'cond, the width of the undercurrent was highly variable, being confined quite close to the continental slope in some places and extending well offshore in others (Fig. 4). Third, off Vancouver Island the northward flow was augmented by a large cyclonic eddy. Our results and those of Ingraham (1967) show the e'xis- tence of the undercurrent during spring, summer, and fall, and Dodimead et al. (1963) indicate that it is present in winter. Thus we sugge'st that the undercurrent is a permanent feature off Washington and Vancouver Island.

Numerous investigators (e.g. Reid et al. 1958) have presented e'vidence for the existence of the undercurrent off California, and Wooster and Jones (1970) made a detailed investigation of it off Baja California (near 30?N). These results and ours sug-




gest that the flow exists as an identifiable feature for at least 2,200 km; the southern extent of the undercurrent has not been defined. Wooster and Gilmartin (1961) concluded that off South America a poleward flow was present under the equatorward- flowing Peru Current from 50S to at least 400S. In addition, Hart and Currie (1960) presented evidence for an undercurrent off southwest Africa, and Hughes and Barton (1970) found a poleward-flowing undercurrent from 15?N to 30?N along the continental slope off northwest Africa. These various results suggest that poleward un- dercurrents have considerable latitudinal extent and are general features of eastern boundary currents. Theoretical studies of these poleward flows are especially needed to shed light on fe'atures that merit field investigation.

References BARBEE, W. D., J. G. DwORSKI, J. D. IRsH, L. H.

LARSEN, AND M. RATTRAY, JR. 1975. Mea- surement of internal waves of tidal frequency near a continental boundary. J. Geophys. Res. 80: 1965-1974.

BARKLEY, R. A. 1968. Oceanographic atlas of the Pacific Ocean. Univ. Hawaii.

BARNES, C. A., A. C. DUXBURY, AND B. MORSE.

1972. Circulation and selected properties of the Columbia River effluent at sea, p. 41-80. In A. T. Pruter and D. L. Alvorson [eds.], The Columbia River estuary and adjacent ocean waters. Univ. Wash.

CANNON, G. A., AND N. P. LmRD. 1974. Flow along the continental slope off Washington. Trans. Am. Geophys. Union 56: 1135.

DEFANT, A. 1950. Reality and illusion in oceanographic surveys. J. Mar. Res. 9: 111-119.

DODIMEAD, A. J., F. FAVORITE, AND T. HIRANo. 1963. Review of oceanography of the subarctic Pacific region. Bull. Int. N. Pac. Fish. Comm. 13. 195 p

FOFONOFF, N. P., AND S. TABATA. 1966. Vari- ability of oceanographic conditions between ocean station P and Swiftsure Bank off the Pacific coast of Canada. J. Fish. Res. Bd. Can. 23: 825-868.

FOMIN, L. M. 1964. The dynamic method in oceanography. Elsevier.

HALPERN, D. 1972. Wind recorder, current meter and thermistor chain measurements in the Northeast Pacific, August/September 1971. NOAA Tech. Rep. ERL 24-POL 12. 37 p.

HART, T. J., AND R. I. CuRRiE. 1960. The Ben- guela Current. Discovery Rep. 31: 123-298.

HOLBROOK, J. 1975. STD measurements off the Washington and Vancouver Island coasts, September 1973. NOAA Tech. Memo. In press.

- AND D. HALPERN. 1974. STD measurements off the Oregon coast, July/August 1973. Coastal Upwelling Ecosystems Analy- sis Data Rep. 12. 397 p.

HUGHES, P., ANt) E. D. BARTON. 1974. Stratifi- cation and water mass structure in the upwelling area off northwest Africa in April/May 1969. Deep-Sea Res. 21: 611-628.

INGRAHAM, W. J. 1967. The geostrophic circulation and distribution of water properties off the coasts of Vancouver Island and Washing- ton, spring and fall 1963. Fish. Bull. 66: 223-250.

PICKARD, G. L. 1963. Descriptive physical oceanography. Pergamon.

RATTRAY, M., J. G. DWORSKI, AND P. E. KOVALA. 1969. Generation of long internal waves at the continental slope. Deep-Sea Res. 16 (suppl.): 179-195.

REED, R. K., AND D. HALPERN. 1973. STD observations in the Northeast Pacific, Septem- ber-October 1972. NOAA Tech. Rep. ERL 271-POL 19. 58 p.

RIEm, J. L. 1956. Observations of internal tides in October 1959. Trans. Am. Geophys. Union 37: 278-286.

. 1973. Northwest Pacific Ocean waters in winter. Johns Hopkins.

, G. I. RODEN, AND J. G. WYLLIE. 1958. Studies of the California Current system. Calif. Coop. Oceanic Fish. Invest. Rep., July 1, 1956-January 1, 1958, p. 28-56.

, AND R. A. SCHWARTZLOSE. 1963. Di- rect measurements of the Davidson Current off central California. J. Geophys. Res. 67: 2491-2497.

TIBBY, R. B. 1941. The water masses off the west coast of North America. J. Mar. Res. 4: 112-121.

UDA, M. 1963. Oceanography of the subarctic Pacific Ocean. J. Fish. Res. Bd. Can. 20: 119-179.

WOOSTER, W. S., AND M. GILMARTIN. 1961. The Peru-Chile Undercurrent. J. Mar. Res. 19: 97-122.

, AND J. H. JoNEs. 1970. California Un- dercurrent off northern Baja California. J. Mar. Res. 28: 235-250.

, AND B. A. TAFT. 1958. On the reliability of field measurements of temperature and salinity in the ocean. J. Mar. Res. 17: 552- 566.

WYLLIE, J. 1966. Geostrophic flow of the Cali- fornia Current at the surface and at 200 meters. Calif. Coop. Oceanic Fish. Invest. Atlas 4.

Submitted: 3 July 1975 Accepted: 10 December 1975



Documents

Observations of the California Undercurrent Off Washington and Vancouver Island