Terrence M. Joyce and Jane
Dunworth-Baker
Woods Hole Oceanographic Institution
Woods Hole, Massachusetts
Abstract. Hydrographic changes in the Northwest Pacific Ocean are examined using data for two time periods: 1945: 1975 and 1976:1998. The largest changes in T/S (2°C,0.2 psu) are within the inter-frontal zone between the Kuroshio Extension (KE) and the subarctic (or Oyashio) front near the Shatsky Rise, and are consistent with a southward shift of the Kuroshio Bifurcation Front (KBF). Other major changes seen are a freshening of approximately 0.04 psu of the newly formed North Pacific Intermediate Water (NPIW); a shoaling over time of the halocline in the center of the Western Subarctic Gyre (WSAG); and a southward shift of the subarctic front between the dateline and 150°W. Because long-term T/S changes near the Shatsky Rise are well correlated, we have examined subsurface thermal data at 100 and 200-m depth in the region and found that shifts in front locations show a high degree of correlation with the Pacific Decadal Oscillation index on an inter-annual basis, and lag that index by about one year at 200-m depth.
1. Introduction
Studies of interannual to decadal
timescale variability in the NW Pacific have generally utilized sea surface
temperature or subsurface temperature. Our
interest is to take advantage of a new summary of the hydrography in the N.
Pacific (Macdonald et al., 2001),
updated here with recent data from the World Ocean Circulation Experiment (WOCE),
which allows hydrographic data to be combined and analyzed by grouping and averaging
on potential density surfaces. Macdonald et al. presented a hydrographic climatology
and our intent is to explore variability. However, the availability of hydrographic
data for a study of temporal variability is much less than for subsurface temperature
data (from MBTs and XBTs), and we have therefore chosen to examine only two
time periods and the changes between them. The time periods are 1945:1975 and
1976:1998, spanning what has been come to be called a major regime shift (Trenberth,
1990; Mantua et al., 1997) in the
atmosphere, ocean and biology of the North Pacific. While we concentrate on
temperature and salinity changes, other hydrographic data are available and
have been used to examine changes over time, such as dissolved oxygen and silica.
2. Hydrographic
Changes
Hydrographic station data for
each of the two periods (Figure 1)
show a high density of observations near Japan decreasing to the east. In the
latter period (containing WOCE data), the data coverage is clearly biased along
'lines' with little or no data in between. Clearly, the estimation of spatial
changes must address this sampling problem and one may consider two choices:
comparison of changes along synthetic hydrographic 'sections' or spatially interpol-ating
onto a rather coarse grid. We chose the latter and used a nearest neighbor method
with a 5° resolution in latitude and longitude, then interpolating onto a finer
grid for comparison and display. Thus, we will be examining rather broad scale
changes between the two time periods. We have chosen to show T/S fields and
differences at two depths: 100 and 200 m (Figure 1).
Dynamic height fields and differences are shown for the surface relative to
a 1000-dbar pressure surface.
We begin with dynamic
height, which shows the mean (1945:1998) baroclinic surface height structure
of the KE and the WSAG with a minimum height SE of Kamchatka and a maximum height
south of Japan. The temporal change in dynamic height is plotted along with
mean frontal locations of the KE (using the 12°C isotherm at 300 m following
Mizuno and White, 1983) and the subarctic front or Oyashio boundary (using the
33.6 psu salinity at 100 m following Kawai (1972) and there is clear evidence
that major changes in dynamic height follow the fronts. A southward shift of
the axis of the KE is indicated by negative dynamic height anomalies of 5-10
cm along its axis. These dynamic height changes are
largest to the west of the dateline and are supported by T/S changes. One can
see evidence for a southward shift of the Oyashio front between the two periods
in the region to the east of the dateline to at least 150°W, the eastern boundary
of our domain; this is seen by a negative salinity (and temperature) change
along the northern of the two fronts plotted in Figure
1.
The largest T/S changes occur
in the zone between the two fronts centered roughly at 160°E. Kawai cites a
paper dating back to Uda in 1935 where it was first shown that various fronts
split apart, with a major bifurcation of the Kuroshio Extension (Kuroshio Bifurcation
Front: KBF) occurring near the Shatsky Rise at 155°E. A northern branch of this
front then merges with the Oyashio to become a subarctic front to the east.
This region of bifurcation in the KE was examined in detail by Mizuno and White
(1983) in their analysis of five years of XBT data between 1976 and 1980. They
interpret their data as reflecting temporal shifts in the thermal structure
of the KE and link location of bifurcations of the KE with the topography of
the Shatsky Rise following Levine and White (1983). Another study by Belkin
and Mikhailichenko (1986) specifically targeted this longitude in a high-resolution
hydrographic survey interpreted in terms of synoptic frontal locations. We find
that long-term salinity changes are consistent with the interpretation of dominant
changes in time at both 100-m and 200-m depth being due to meridional shifts
of fronts. Since T/S changes are well correlated this will allow us in the next
section to interpret temperature changes alone in terms of meridional shifting
of fronts rather than diabatic forcing near 160°E.
In Figure 1, the
local maximum in mean salinity at 200 m, SE of Kamchatka, is within the WSAG
and reflects a shallower halocline near the gyre center; it is co-located with
the minimum in mean dynamic height. In the same region, there is a salinity
increase with time at both depths due to the shoaling of the halocline, a reduction
of 5 cm in surface dynamic height over time with perhaps some suggestion of
a westward intensification, but no noticeable temperature change. The salinity
increase at 100 m extends into the eastern subpolar gyre as well. In summertime,
a distinct temperature minimum layer develops in this region (see Talley et
al, 1990 for a hydrographic section in the region) centered near a depth of
100 m, making temperature a poor 'tracer' for vertical motion. Strong vertical
gradients in dissolved oxygen and silica exist in this region in the upper ocean
and, indeed, silica increases over time while oxygen decreases (neither is shown
here) consistent with this interpretation of salinity change. Changes in halocline
depth over time are consistent with a spinning up of the subpolar gyre(s), and
will be discussed later.
We also have investigated the changes in NPIW
between the two periods (Figure 2)
and have found it to be freshening by approximately 0.04 psu. This analysis
was done on a potential density surface of 26.8, which is the climatological
level at which this water mass is formed by subsurface mixing processes (Talley,
1993) between source waters in the Sea of Okhotsk and the Kuroshio. The core
thickness of this layer can be traced by its planetary potential vorticity (Yasuda,
1997) from the Sea of Okhotsk (Figure 2, upper panel)
as it moves into the interfrontal zone where it first appears as a sub-surface
salinity minimum. Between the two time periods, this layer has freshened (Figure
2, lower panel) over much of its pathway, with the greatest freshening within
the inter-frontal zone, not in the Sea of Okhotsk. If this freshening were due
to an increase in transport of the subarctic gyre, bringing more subpolar water
into the KE, one would expect the dissolved oxygen of the layer to increase
in time, but in fact it decreases (not shown). Thus, NPIW changes are not simply
linked to wind-driven transport changes, and could also depend upon long-term
changes in the net freshwater flux in the subpolar North Pacific, as suggested
by Wong et al., (1999) with lower-salinity
water in the surface layer reducing ventilation by vertical diffusion, as reflected
in changes in the oxygen concentration at depths near a potential density of
26.8. Indeed, the oxygen in the Sea of Okhotsk is lower at this density in the
second period.
3. Subsurface Temperature
Changes
Because the long-term changes
in subsurface temperature at depths of 100 and 200 m are so well correlated
with salinity, it is sensible to ask if the more numerous temperature data may
now be interpreted in terms of what we have seen from the hydrography. All available
MBT and XBT data have been extracted from the WOD98-05 (Conkright et al., 1998). These have been augmented
by more recent data from the Global Temperature and Salinity Pilot Project (GTSPP:
available at http://www.nodc.noaa.gov/GTSPP/gtspp-bc.html) up through and including
1999. They have been gridded annually and with a resolution of 1° of latitude
and 2° of longitude. Interannual anomalies have been calculated and for our
purposes here, averaged for a box region between 154: 166°E and 36:42°N, suggested
by the 'bullseye' temperature difference in Figure 1.
In fact, this region was one with a local maximum in interannual variability
for temperature at 100 and 200-m depths for the time period 1954:1999. We have
seen (not shown) that compositing the data before and after 1976 produced a
similar, temperature difference as found with the hydrography. The advantage
of the thermal data is greater temporal resolution than with the hydrography.
We have done this (Figure 3) for
both depths analyzed above and plotted a filtered PDO index following Mantua
et al., smoothed in time with a running
3-point .25, .50, .25 filter. One can see clearly
that subsurface thermal anomalies are highly correlated with one another and
anti-correlated with the PDO (r = -0.8), lagging the latter by about
a year (not shown), with a suggestion of a longer lag with increasing depth.
Since 100 m lies within the winter mixed layer, temperature changes at this depth may be due to surface forcing as well
as meridional shifts of water masses. A depth of 200 m lies below mean winter
mixed layer depths in the region (Huang and Qiu, 1994) and is less susceptible
to diabatic effects. The 'regime shift' is one of predominantly low-to-high
PDO index, but with interannual variations that are reflected in both the index
and the subsurface temperature field. The period over which Mizuno and White
did their analysis (1976-1980) is one of small change
relative to the whole record. The meridional amplitude or zonal extent of the
branching of the KBF as reflected in Figure 3
is highly time-dependent and linked to ocean dynamics associated with the PDO.
4. Discussion
Recently Deser et al. (1999) examined 50-m and 400-m temperature
variability for the period of 1968:1991 spanning the regime shift in the PDO
or the North Pacific Index (NPI), after Trenberth and Hurrell (1994). The NPI
is based on SLP variations off the Aleutians, and is not a SST-based index like
the PDO, but the two indices are highly anti-correlated. Deser et al found that leading EOFs were correlated
with the intensity of the Aleutian low, with 50 m nearly in phase and the
400-m temperature lagging by about 4 years, bracketing the one year lag at 200
m of our simple temperature box centered at 39°N, 160°E. The spatial EOF pattern
at the two depths was substantially different, however, with the 50-m EOF centered
on the dateline and the 400-m off the east coast of Japan: again bracketing
our Shatsky Rise location. Deser et al. further examined wind stress variability and showed that there
were changes in the Sverdrup circulation associated with the NPI changes, with
low PDO (high NPI) periods ones with reduced wind-driven transport and the opposite
during periods of high PDO (low NPI). Miller et al (1998) composited data from
1970:1976 and 1980:1986 and explored changes in thermocline depth (400 m) using
data and a coarse-grid wind-driven model. They found that the model produced
westward-intensified variability between these periods in agreement with the
data and the second period was one of increased transports in both the subarctic
and subtropical gyres. This spin-up of the subarctic region is consistent with
the above changes in salinity, dynamic height, oxygen and silica in the center
of the WSAG.
Diabatic effects were the focus of a study
by Yasuda and Hanawa (1997), who had a particular interest on mode waters of
the North Pacific, and they found that the strong SST signal near the dateline
was produced by a combination of changes in air-sea heat flux and Ekman layer
transport of heat, with enhanced production of central mode water during the
years following the regime shift. While they did not comment, their data showed
the maximum temperature change at depths of 100- 300 m was located over
the Shatsky Rise in the region of the KBF (their Figure 2 at 39°N), not in the
region of mode water formation
The analysis of section data at 137°E by Qiu
and Joyce (1992) indicated an increase in Kuroshio geostrophic transport of
about 10 Sv for the period after the regime shift compared to the period before;
but this time series is too short to make any definitive statement about time
lags relative to the PDO. Kuroshio transport must be closely related to dynamic
height changes associated with thermocline adjustments, and therefore the 400-m
temperature changes analyzed by Deser et al and Miller et al. Yet changes at this depth lag those
at 200 m. If the latter are independent of diabatic effects and thus associated
with frontal positions as we have argued, they precede changes at 400 m in time.
It appears that wind-driven transport of the North Pacific is probably the single
most important driver of the long-term hydrographic changes observed in this
study, and that meridional shifts in fronts are important in creating changes
in the location of water masses, hence hydrographic variability in the Northwest
Pacific. The phasing of the changes suggest, however, that shifts in the KBF
and perhaps other fronts may not result from a change in an upstream transport
condition near Japan but from adjustment to wind-driven changes coming from
the east.
The evolution of NPIW towards colder, fresher
characteristics does not seem simply related to wind-driven transport changes
between the two periods. An increase in the subpolar component of NPIW would
increase the contribution of fresher end-member which mixes with the warmer,
saltier Kuroshio waters in the KE (e.g., See Joyce et al., 2001). However, oxygen content changes are not consistent
with this. Neither is a simple change in properties of the low PV source water
in the Sea of Okhotsk, as salinity changes there are small.
Acknowledgments. This work was part of a JAMSTEC
project at WHOI aimed at identifying sites for long-term monitoring in the NW
Pacific, under the leadership of Sus Honjo. We also acknowledge an exchange
of ideas with K. Hanawa, T. Suga, B. Qiu, and R. Krishfield and two anonymous reviewers during the course
of this work.
References
Belkin, I.M., and Yu.G. Mikhailichenko,
1986. Thermohaline structure of the Northwest Pacific Frontal Zone near 160°E.
Oceanology, English translation, 26, 47-49.
Conkright, M.E, S. Levitus,
T.P. Boyer, T. O'Brien, C. Stephens, D. Johnson, L. Stathoplos, O. Baranova,
J. Antonov, R. Gelfeld, J. Burney, J. Rochester, C. Forgy, 1998. Temperature,
Vol 5, World Ocean Database 1998 CD-Rom
Scientific Dataset, version 1.0. National
Oceanographic Data Center Internal Report 14.
Deser, C.M., A. Alexander, and
M.S. Timlin, 1999. Evidence for a wind-driven intensification of the Kuroshio
Extension from the 1970s to the 1980s. J. Climate, 12, 1697-1706.
Huang, R.X., and B. Qiu, 1994.
Three-dimensional structure of the wind-driven circulation in the subtropical
North Pacific. J. Phys Oceanogr., 24, 1608-1622.
Joyce, T.M., I. Yasuda, Y. Hiroe,
K. Komatsu, and F. Bahr, 2001. Mixing in the meadering Kuroshio Extension and
the formation of North Pacific Intermediate Water, J. Geophys. Res.,
106, 4397-4404.
Kawai, Hideo, 1972. Hydrography
of the Kuroshio Extension, In Kuroshio, H. Stommel, K. Yoshida, Eds.,
U. Tokyo Press, 517 pp.
Levine, E.R., and W.B. White,
1983. Bathymetric influences upon the character of North Pacific fronts, J. Geophys.
Res., 88, 9617-9625.
Macdonald, A.M., T. Suga, and
R,G. Curry, 2001. An isopycnally averaged North Pacific Climatology. J. Atmos.
and Oceanic Tech., 18, 394-420.
Mantua, N.J., S.R. Hare, Y.
Zhang, J.M. Wallace, and R.C. Francis, 1997. A Pacific interdecadal climate
oscillation with impacts on salmon production. Bull. Amer. Meteor. Soc.,
78, 1069-1079.
Miller, A.J., D.R. Cayan, and
W.B. White, 1998. A westward-intensified decadal change in the North Pacific
thermocline and gyre-scale circulation. J. Climate, 11, 3112-3127.
Mizuno, K., and W.B. White,
1983. Annual and interannual variability in the Kuroshio current system. J. Phys.
Oceanogr., 13, 1847-1867.
Talley, L.D., T.M. Joyce, and
R.A. DeSzoeke, 1991. Transpacific sections at 47°N and 152°W: distribution of
properties. Deep-Sea Res., 38, Suppl., S63-S82.
Talley, L.D., 1993. Distribution
and formation of North Pacific Intermediate
Water. J. Phys. Oceanogr., 23,
518-537.
Trenberth, K.E., 1990. Recent
observed interdecadal climate changes in the Northern hemisphere. Bull. Amer.
Meteor. Soc., 71, 988-930.
Trenberth, K.E., and J.W. Hurrell,
1994. Decadal atmosphere-ocean variations in th4e Pacific.
Climate Dyn., 9, 303-319.
Qiu, B., and T.M. Joyce, 1992.
Interannual variability in the mid- and low-latitude western North Pacific.
J. Phys. Oceanogr., 222, 1062-1079.
Wong, A.P.S., N.L. Bindoff,
and J.A.Church, 1999. Large-scale freshening of intermediate waters in the Pacific
and Indian Oceans, Nature, 400, 440-443.
Yasuda, T., and K. Hanawa, 1997.
Decadal changes in the Mode Waters in the midlatitude North Pacific, J. Phys.
Oceanogr., 27, 858-870.
Yasuda, I., 1997. The origin
of the North Pacific Intermediate Water. J. Geophys. Res., 102,
893-909.