model output, in Dickson, 1999), and Arctic-wide,
the whole pattern of atmospheric pressure and ice
drift appears to have shifted counterclockwise in a
similar sense. A record extent of open water was
recorded in the eastern Arctic in the early 1990s
(Serreze et al., 1995; Maslanik et al., 1996; McPhee
et al., 1998).
The third new feature to emerge during the
1990s was the evidence to link this long-term variability of the Arctic Ocean with seas further south.
A combination of current measurements, upwardlooking sonar and satellite imagery show that the
amplifying NAO was accompanied by an increased
annual efflux of ice through the western Fram
Strait to a record volume-flux of 4687 km
3 yr
91 in
1994–95 (Vinje et al., 1998; see also Kwok and
Rothrock, 1999) and the southward recirculation
of extreme warmth from the eastern Fram Strait
(see above) was also observed to affect the Denmark
Strait Overflow, some 2500 km further south
(Dickson et al., 1999).
Thus although not a part of the WOCE Field
Experiment, a sufficient observing effort was in
place across the Arctic during the 1990s to capture
an unprecedented, extreme and system-wide suite
of changes. If our longest proxies for the dominant
forcing prove valid, these would have included
effects of the most positive winter NAO index for
over 500 years (Stockton and Glueck, 1999; see
Section 7.3.2.2). Of direct importance to the aims
of WOCE, these changes document a potential for
global as well as regional effects on climate, acting
locally through the effect of a changing ocean
stratification and ice cover on surface heat flux,
and remotely through the variable export of fresh
waters and heat to the headwaters of the global
thermohaline circulation. The latter seems particularly apt in view of the global importance that
coupled models currently assign to relatively
minor changes in freshwater distribution at high
latitudes (Rahmstorf and Ganopolski, 1999).
7.3.8 Interdecadal variability of
Kuroshio transport
Passing into the Pacific sector, we find that the climatic variability over the North and Equatorial
Pacific over the past few decades has hardly been
less extreme than that of the North Atlantic. In
mid-latitudes, the NP index, reflecting the areaweighted mean SLP averaged over most of the
extratropical North Pacific Ocean (Trenberth and
Hurrell, 1994) underwent a striking change
amounting to a regime shift after about 1976. For
most of the years since then, including the WOCE
period, the index has been negative, reflecting an
eastward-shifted and intensified Aleutian low (Section 7.3.2.3). At lower latitudes (Section 7.3.2.4),
the Southern Oscillation has recently exhibited an
unusual amplitude of decadal and longer-term
variability (much like the recent behaviour of the
NAO), with a pronounced change toward more
negative values since the mid-1970s, and prolonged warm conditions into the start of the
WOCE period in 1990–95.
The large scale of these climatic events suggest
the possibility of a response in the circulation of
the North Pacific. Though there is increasing evidence to link decade-scale fluctuations of ocean
and atmosphere in this sector (e.g. Mantua et al.,
1997; Nakamura et al., 1997), this has tended to
concern the ocean surface rather than the interior.
South of Japan, however, Qiu and Joyce (1992)
have used the twice yearly hydrographic surveys
along 137°E conducted by the Japan Meteorological Agency (JMA) from 1967 to 1988 to compute
the long-term variations in the geostrophic transport of the Kuroshio, defined as the total eastward
flow north of 30°N referenced to a depth of
1.25 km. Apart from resolving the fact that there is
no apparent difference in Kuroshio transport
between its straight-path and meander-path years
(the WOCE period was essentially ‘straight-path’),
these authors describe a decadal increase of
10.8 Sv in their Kuroshio transport between 1970
and 1980 and between 1982 and 1988. Miller
et al. (1998) use a general circulation model to
suggest that decadal transport variations of the
Kuroshio current system over the same period
were a response to large-scale wind forcing.
For the latitudes of the Kuroshio Extension
(32–37°N) east of Japan, Deser et al. (1999) appear
to confirm this increased transport in two ways.
First the Sverdrup transport, inferred from the
wind stress curl field, shows a decadal increase of
11.5 Sv between the 1970s and 1980s. Second,
combining the change in ocean temperature across
the Kuroshio Extension between the 1970s and
1980s with climatological salinity profiles from
Levitus and Boyer (1994b), they find that the
depth-averaged geostrophic transport referenced to
1 km increased by 11.6 Sv between 32°N and
7.3 The World during WOCE
573
Dickson et al.
the whole pattern of atmospheric pressure and ice
drift appears to have shifted counterclockwise in a
similar sense. A record extent of open water was
recorded in the eastern Arctic in the early 1990s
(Serreze et al., 1995; Maslanik et al., 1996; McPhee
et al., 1998).
The third new feature to emerge during the
1990s was the evidence to link this long-term variability of the Arctic Ocean with seas further south.
A combination of current measurements, upwardlooking sonar and satellite imagery show that the
amplifying NAO was accompanied by an increased
annual efflux of ice through the western Fram
Strait to a record volume-flux of 4687 km
3 yr
91 in
1994–95 (Vinje et al., 1998; see also Kwok and
Rothrock, 1999) and the southward recirculation
of extreme warmth from the eastern Fram Strait
(see above) was also observed to affect the Denmark
Strait Overflow, some 2500 km further south
(Dickson et al., 1999).
Thus although not a part of the WOCE Field
Experiment, a sufficient observing effort was in
place across the Arctic during the 1990s to capture
an unprecedented, extreme and system-wide suite
of changes. If our longest proxies for the dominant
forcing prove valid, these would have included
effects of the most positive winter NAO index for
over 500 years (Stockton and Glueck, 1999; see
Section 7.3.2.2). Of direct importance to the aims
of WOCE, these changes document a potential for
global as well as regional effects on climate, acting
locally through the effect of a changing ocean
stratification and ice cover on surface heat flux,
and remotely through the variable export of fresh
waters and heat to the headwaters of the global
thermohaline circulation. The latter seems particularly apt in view of the global importance that
coupled models currently assign to relatively
minor changes in freshwater distribution at high
latitudes (Rahmstorf and Ganopolski, 1999).
7.3.8 Interdecadal variability of
Kuroshio transport
Passing into the Pacific sector, we find that the climatic variability over the North and Equatorial
Pacific over the past few decades has hardly been
less extreme than that of the North Atlantic. In
mid-latitudes, the NP index, reflecting the areaweighted mean SLP averaged over most of the
extratropical North Pacific Ocean (Trenberth and
Hurrell, 1994) underwent a striking change
amounting to a regime shift after about 1976. For
most of the years since then, including the WOCE
period, the index has been negative, reflecting an
eastward-shifted and intensified Aleutian low (Section 7.3.2.3). At lower latitudes (Section 7.3.2.4),
the Southern Oscillation has recently exhibited an
unusual amplitude of decadal and longer-term
variability (much like the recent behaviour of the
NAO), with a pronounced change toward more
negative values since the mid-1970s, and prolonged warm conditions into the start of the
WOCE period in 1990–95.
The large scale of these climatic events suggest
the possibility of a response in the circulation of
the North Pacific. Though there is increasing evidence to link decade-scale fluctuations of ocean
and atmosphere in this sector (e.g. Mantua et al.,
1997; Nakamura et al., 1997), this has tended to
concern the ocean surface rather than the interior.
South of Japan, however, Qiu and Joyce (1992)
have used the twice yearly hydrographic surveys
along 137°E conducted by the Japan Meteorological Agency (JMA) from 1967 to 1988 to compute
the long-term variations in the geostrophic transport of the Kuroshio, defined as the total eastward
flow north of 30°N referenced to a depth of
1.25 km. Apart from resolving the fact that there is
no apparent difference in Kuroshio transport
between its straight-path and meander-path years
(the WOCE period was essentially ‘straight-path’),
these authors describe a decadal increase of
10.8 Sv in their Kuroshio transport between 1970
and 1980 and between 1982 and 1988. Miller
et al. (1998) use a general circulation model to
suggest that decadal transport variations of the
Kuroshio current system over the same period
were a response to large-scale wind forcing.
For the latitudes of the Kuroshio Extension
(32–37°N) east of Japan, Deser et al. (1999) appear
to confirm this increased transport in two ways.
First the Sverdrup transport, inferred from the
wind stress curl field, shows a decadal increase of
11.5 Sv between the 1970s and 1980s. Second,
combining the change in ocean temperature across
the Kuroshio Extension between the 1970s and
1980s with climatological salinity profiles from
Levitus and Boyer (1994b), they find that the
depth-averaged geostrophic transport referenced to
1 km increased by 11.6 Sv between 32°N and
7.3 The World during WOCE
573
Dickson et al.