Varying planetary heat sink led to global-warming slowdown and

acceleration
Xianyao Chen and Ka-Kit Tung
Science 345, 897 (2014);
DOI: 10.1126/science.1254937

This copy is for your personal, non-commercial use only.

If you wish to distribute this article to others, you can order high-quality copies for your
colleagues, clients, or customers by clicking here.
Permission to republish or repurpose articles or portions of articles can be obtained by
following the guidelines here.

The following resources related to this article are available online at

www.sciencemag.org (this information is current as of August 21, 2014 ):

Downloaded from www.sciencemag.org on August 21, 2014

Updated information and services, including high-resolution figures, can be found in the online
version of this article at:
http://www.sciencemag.org/content/345/6199/897.full.html
Supporting Online Material can be found at:
http://www.sciencemag.org/content/suppl/2014/08/20/345.6199.897.DC1.html
A list of selected additional articles on the Science Web sites related to this article can be
found at:
http://www.sciencemag.org/content/345/6199/897.full.html#related
This article cites 47 articles, 7 of which can be accessed free:
http://www.sciencemag.org/content/345/6199/897.full.html#ref-list-1
This article appears in the following subject collections:
Atmospheric Science
http://www.sciencemag.org/cgi/collection/atmos
Oceanography
http://www.sciencemag.org/cgi/collection/oceans

Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the
American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright
2014 by the American Association for the Advancement of Science; all rights reserved. The title Science is a
registered trademark of AAAS.
 R ES E A RC H | R ES E A R C H A R T I C L E S

specific mixes of these fishes can remove sea- 23. S. N. Arnold, R. S. Steneck, PLOS ONE 6, e28681 (2011).
weeds, enhance CCA, enhance corals (6, 9, 12, 33), 24. J. Atema, M. J. Kingsford, G. Gerlach, Mar. Ecol. Prog. Ser. 241,
151–160 (2002).
and replace chemical cues of degradation with 25. D. B. Rasher, E. P. Stout, S. Engel, J. Kubanek, M. E. Hay,
chemical stimulants for recruitment. Proc. Natl. Acad. Sci. U.S.A. 108, 17726–17731 (2011).
26. T. D. Andras et al., J. Chem. Ecol. 38, 1203–1214 (2012).
27. D. L. Dixson, M. S. Pratchett, P. L. Munday, Anim. Behav. 84,
RE FERENCES AND NOTES
45–51 (2012).
1. L. Alvarez-Filip, N. K. Dulvy, J. A. Gill, I. M. Côté, A. R. Watkinson, 28. J. M. Leis, U. Siebeck, D. L. Dixson, Integr. Comp. Biol. 51,
Proc. Biol. Sci. 276, 3019–3025 (2009). 826–843 (2011).
2. N. A. J. Graham, K. L. Nash, Coral Reefs 32, 315–326 (2013). 29. D. Lecchini, V. P. Waqalevu, E. Parmentier, C. A. Radford,
3. T. A. Gardner, I. M. Côté, J. A. Gill, A. Grant, A. R. Watkinson, B. Banaigs, Mar. Ecol. Prog. Ser. 475, 303–307 (2013).
Science 301, 958–960 (2003). 30. G. Gerlach, J. Atema, M. J. Kingsford, K. P. Black, V. Miller-Sims,
4. J. F. Bruno, E. R. Selig, PLOS ONE 2, e711 (2007). Proc. Natl. Acad. Sci. U.S.A. 104, 858–863 (2007).
5. T. P. Hughes, N. A. J. Graham, J. B. C. Jackson, P. J. Mumby, 31. A. H. Baird, M. S. Pratchett, A. S. Hoey, Y. Herdiana,
R. S. Steneck, Trends Ecol. Evol. 25, 633–642 (2010). S. J. Campbell, Coral Reefs 32, 803–812 (2013).
6. D. E. Burkepile, M. E. Hay, Proc. Natl. Acad. Sci. U.S.A. 105, 32. M. S. Pratchett, D. McCowan, J. A. Maynard, S. F. Heron,
16201–16206 (2008). PLOS ONE 8, e70443 (2013).
7. P. J. Mumby, R. S. Steneck, Trends Ecol. Evol. 23, 555–563 33. S. M. Lewis, Ecol. Monogr. 56, 183–200 (1986).
(2008). 34. P. F. Sale et al., Trends Ecol. Evol. 20, 74–80 (2005).
8. A. J. Cheal et al., Coral Reefs 29, 1005–1015 (2010). 35. H. B. Harrison et al., Curr. Biol. 22, 1023–1028
9. D. B. Rasher, A. S. Hoey, M. E. Hay, Ecology 94, 1347–1358 (2012).
(2013). 36. S. R. Palumbi, Annu. Rev. Environ. Resour. 29, 31–68
10. S. K. Wilson et al., Conserv. Biol. 26, 995–1004 (2012). (2004).
11. M. Nyström et al., Ecosyst. 15, 695–710 (2012).
12. T. P. Hughes et al., Curr. Biol. 17, 360–365 (2007). AC KNOWLED GME NTS
13. A. J. Cheal, M. Emslie, M. A. MacNeil, I. Miller, H. Sweatman, Support was provided by NSF grant OCE-0929119, NIH grant
Ecol. Appl. 23, 174–188 (2013). U01-TW007401, and the Teasley Endowment. We thank V. Bonito
14. H. Cesar, L. Burke, L. Pet-Soede, The Economics of World Wide for assistance and the Fijian government and the Korolevu-i-wai
Coral Reef Degradation (Cesar Environmental Economics district elders for research permissions. Data are tabulated in
Consulting, Netherlands, 2003). the supplementary materials.
15. I. B. Kuffner, V. J. Paul, Coral Reefs 23, 455–458 (2004).
16. H. Watanabe, T. Fujisawa, T. W. Holstein, Dev. Growth Differ. 51,
167–183 (2009). SUPPLEMENTARY MATERIALS
17. Y. Benayahu, Y. Loya, Bull. Mar. Sci. 31, 514–522 (1981).
www.sciencemag.org/content/345/6199/892/suppl/DC1
18. T. R. McClanahan, R. Arthur, Ecol. Appl. 11, 559–569 (2001).
Materials and Methods
19. D. F. Gleason, B. S. Danilowicz, C. J. Nolan, Coral Reefs 28,
Fig. S1
549–554 (2009).
Tables S1 to S4
20. Y. Golbuu, R. H. Richmond, Mar. Biol. 152, 639–644 (2007).
References (37–42)
21. L. Harrington, K. Fabricius, G. De’ath, A. Negri, Ecology 85,
3428–3437 (2004). 21 April 2014; accepted 23 June 2014
Fig. 5. Density and species richness of fish 22. J. Tebben et al., PLOS ONE 6, e19082 (2011). 10.1126/science.1255057
recruits, and density of predators, in MPAs and
non-MPAs. (A) Density (TSE) and (B) species
richness of recruits. (C) Density of predators. n =
30 transect per site. P < 0.001 for all contrasts. CLIMATE

the recovery of more-degraded habitats (34). How-

ever, export to degraded reefs will be constrained
if recruits avoid chemical cues from degraded
Varying planetary heat sink led to
reefs. In some locations, export occurs (35), but
in other areas, there is minimal connectivity,
even between protected and exploited popula-
global-warming slowdown
tions separated by small distances (36). These
divergent outcomes could occur if intact reefs
and acceleration
export larvae to similar communities but not to Xianyao Chen1,2 and Ka-Kit Tung2*
degraded, seaweed-dominated communities. The
protected and fished areas we studied differ in A vacillating global heat sink at intermediate ocean depths is associated with different
coral cover, seaweed cover, herbivory rates, and climate regimes of surface warming under anthropogenic forcing: The latter part of the
biomass of herbivorous fishes (9). Thus, orga- 20th century saw rapid global warming as more heat stayed near the surface. In the 21st
nisms settling only 100 m apart experience dra- century, surface warming slowed as more heat moved into deeper oceans. In situ and
matically different environments. Our findings reanalyzed data are used to trace the pathways of ocean heat uptake. In addition to the
suggest that once degradation passes some crit- shallow La Niña–like patterns in the Pacific that were the previous focus, we found that
ical but as yet undetermined threshold, recruit the slowdown is mainly caused by heat transported to deeper layers in the Atlantic
behavior may constrain the value of healthy reefs and the Southern oceans, initiated by a recurrent salinity anomaly in the subpolar North
as larval sources for populations in degraded Atlantic. Cooling periods associated with the latter deeper heat-sequestration
habitats. mechanism historically lasted 20 to 35 years.

I
To produce the desired connections between
healthy reefs as a source of larvae and degraded ncreasing anthropogenic greenhouse-gas emis- Because the heat capacity of the atmosphere and
reefs as targeted settlement sites where recruit- sions perturb Earth’s radiative equilibrium, the cryosphere is small, about 90% of the total
ment can promote reef resilience, managers will leading to a persistent imbalance at the top heat content is in the form of ocean heat content
have to suppress the chemical barrier produced of the atmosphere (TOA) despite some long- (OHC) (2, 3). Although the magnitude of the TOA
by seaweeds and enhance the chemical “call” of wave radiative adjustment. Energy balance 1
Key Laboratory of Physical Oceanography, Ocean University
corals and CCA. A promising strategy could be to requires that this TOA imbalance for the planet of China, Qingdao, China. 2Department of Applied Mathematics,
reduce the harvest of critical species of herbivo- equal the time rate of increase of the total heat University of Washington, Seattle, WA, USA.
rous fishes beyond MPA borders. Feeding by content in the atmosphere-ocean system (1). *Corresponding author. E-mail: [email protected]

SCIENCE sciencemag.org 22 AUGUST 2014 • VOL 345 ISSUE 6199 897

R ES E A RC H | R E S EA R C H A R T I C LE S

radiative imbalance is uncertain, variously esti- The current period of global-warming slow- (10, 12, 13), were used in this study (supplemen-
mated (1, 2, 4) at 0.5 to 1.0 W m−2, its sign is down coincidently occurred during the era of tary materials). Data below 2000 m are very sparse,
believed to be positive on multiyear time scales; Argo profiling floats (7), which were gradually but are probably not needed for our purpose of
therefore, this forced total OHC should be in- deployed since the end of the 20th century and revealing recent signals from above.
creasing monotonically over longer periods even achieved near-global coverage of the world’s
through the current period of slowed warming oceans by 2005. Before that milestone, other Observed SST and OHC
(5, 6). In fact, that expectation is verified by ob- data types, such as the expendable bathythermo- The global-mean temperature in the first decade
servation (red curve in Fig. 1A, which is an ap- graphs, were available but contained depth bias. of the 21st century has very little trend (6, 14),
proximation to the total OHC). What the extent After its recent correction (8–10), the largest re- in contrast to the latter part of the 20th
of radiative forcing does not determine, though, maining uncertainty, related to the infilling of century, which saw rapid global warming (2).
is OHC’s vertical distribution, which is controlled the undersampled areas, especially in the South- Figure 1A shows that, in the 21st century, the
by the ocean’s internal variability, with each ocean ern Ocean, was estimated to be less than 0.05 × global-mean SST warming rate has slowed to
basin having its own characteristic time scale. 1023 J since 1993 by Lyman and Johnson (11) almost zero for more than a decade—called a
And it is the vertical distribution that is impor- for the global mean. Monthly objectively ana- slowdown or hiatus—with short-term El Niño–
tant in determining the sea-surface temperature lyzed subsurface temperature and salinity at Southern Oscillation (ENSO)–like variation super-
(SST), given how voluminous the ocean column is. 24 levels in the upper 1500 m, called the Ishii data imposed on it. Thus, two time scales are seen: a

Fig. 1. Integrated OHC. Integrated from the surface to different indicated depths in the global ocean (A), the Atlantic (B), the Pacific (C), the Southern Ocean (D),
and the Indian Ocean (E). Shown is the 12-month running mean deviation from the climatological mean (1970 to 2012) for each layer, so attention should not be
focused on the absolute distance between the curves but should be on their relative changes in time. Color lines show the OHC in the left scale, in units of 1023 J.
The black line shows the mean SSTup to 2013. (Insets) The division of the globe into the Pacific, the Atlantic, the Indian Ocean, and the Southern Ocean. Although
shown in the figure, data in the earlier decades were not as reliable (see Data and Materials and Methods); the discussion in the text is focused on the better-
observed regions and periods.

898 22 AUGUST 2014 • VOL 345 ISSUE 6199 sciencemag.org SCIENCE

 R ES E A RC H | R ES E A R C H A R T I C L E S

Fig. 2. In situ data coverage in 5°-by-5° grid as a

percentage of full coverage. Shown as background
blue color shades at each depth as indicated by the
axis on the left. Integrated OHC from the surface to
each depth (lines with red to yellow colors) for global
(A), Atlantic (B), and Pacific (C) oceans are super-
imposed, with magnitude scale on the right axis. No
climatology is removed; all values refer to difference
from their 2000 values.

multidecadal nonlinear trend with periods of

accelerated warming separated by periods of
slowdown and a shorter 2- to 7-year ENSO time
scale. We argue that they are caused by differ-
ent mechanisms.
During the current hiatus, radiative forcing at
the TOA by the increasing greenhouse-gas con-
centration in the atmosphere produces addition-
al warming in deeper and deeper ocean layers.
This deepening warming is seen in Fig. 1A, calcu-
lated by using the in situ Ishii data (10), as an
increasing fanning out of the OHC curves (in-
tegrated from the surface to different depths)
after 1999. Globally, an additional 0.69× 1023 J
has been sequestered since 1999 in the 300- to
1500-m layer by 2012 (Fig. 1A), which, if absent,
would have made the upper 300 m warm as
fast as the upper 1500 m since 1999. Because
the latter has an uninterrupted positive trend,
there would have been no slowdown of the warm-
ing of the surface or the upper layers. Therefore,
the enhanced ocean heat sink is the main cause
for the current slowing in surface warming.
One can also see from Fig. 1A that there has not
been a major reduction in TOA radiative im-
balance, as inferred from an absence of a no-
ticeable downward bent in slope in the red curve
on multiyear time scales. In the prior two dec-
ades, 1980 to 2000, we have the contrasting
situation of accelerating surface warming while
little additional heat (over the climatological
mean) went into the deeper layers. Most of the
Fig. 3. ORAS4 reanalysis. (A to E) Same as Fig. 1 except with use of ORAS4 data. added heat stayed near the surface, as seen by

SCIENCE sciencemag.org 22 AUGUST 2014 • VOL 345 ISSUE 6199 899

R ES E A RC H | R E S EA R C H A R T I C LE S

the distance between all OHCs integrated from the The 1997/98 El Niño released a large amount with the associated shallow subduction of heat
surface to different depths changing little in time. of heat, about 0.42 × 1023 J, from the upper 300 m below the surface in the Pacific, has been pro-
Figure 1, B to E, shows that during the 21st of the Pacific Ocean to the atmosphere to be posed as the reason for the current hiatus (16).
century the increased OHC penetration below radiated as outgoing long-wave radiation. The However, our result shows that the change in heat
300 m occurs mainly in two ocean basins: the At- deeper ocean was not involved, as seen by the storage under the Pacific is not nearly enough to
lantic and the Southern oceans, each accounting OHC curve of the upper 1500 m tracking close- compensate for the surface hiatus—the entire up-
for slightly less than half of the global energy ly the upper 300-m content in Fig. 1C. Therefore, per 300 m is itself in a net La Niña phase. Figure
storage change since 1999 at those depths. The the time scale was short, and the upper 300 m 4, C and D, shows that the dominant OHC var-
Pacific shows very little change in the spreading recovered 3 to 4 years later. The 1982 El Niño also iability below 300 m occurs mainly in the Atlantic
of OHC with depth between the two periods. released a large amount of heat from the upper basin and the Antarctic Circumpolar Current (ACC)
This recent contrast between the Pacific and 300 m of the Pacific but did not “trigger” a period region. This combined warming pattern in the
the Atlantic is deemed reliable because these of slowdown in global surface warming. lower layer became positive toward the end of the
are the two basins with adequate data coverage Over the whole globe, the dominant spatial 20th century, after two decades with very little heat
[over 80% on 5°-by-5° grids (Fig. 2)]. A separate mode of variability in OHC in the upper 300 m uptake anomaly below 300 m (see the red PC1).
analysis of Ocean Reanalysis System 4 (ORAS4) [as shown by the first empirical orthogonal The Atlantic initiated the heat sequestration toward
(15) data, shown in Fig. 3, verifies the in situ function (EOF), which explains the most vari- the end of the 20th century below 700 m. Indian
result here (10). The reanalysis uses a model to ance], occurs mainly in the tropical Pacific and Ocean’s deeper layers warmed last and with much
assimilate all forms of observational data and has the structure of ENSO variability (Fig. 4, A smaller amplitude. Figure S1 shows these results
dynamically interpolates across periods of sparse and B). Figure 4, E and F, shows the time series in finer vertical resolution, whereas fig. S2 re-
coverage. Although barely noticeable, ORAS4 [called the principal components (PC)] associ- produces these results by using ORAS4 data.
does have slightly more heat sequestration in ated with these EOFs. The spatial pattern of the In the Southern Ocean, the largest warming
the 300 to 700 m in the Pacific since 2001 com- upper 300 m resembles that associated with La below 300 m occurs in the ACC region (17, 18),
pared with the Ishii data during the current hiatus Niña in the current period, because the blue PC1 in the Atlantic sector, diminishing in amplitude
period. Nevertheless, neither data set supports changes to negative around year 2000. The lead- as the current flows eastward (Fig. 4C), first to the
the model result of Meehl et al. (5) that the heat ing pattern is very similar to the La Niña pattern Indian Ocean and then to the Pacific. During
uptake in this layer in the Pacific dominates over in the SST (fig. S1A), and their PC1s are almost the current 14 years, the 300- to 1500-m layer of the
other ocean basins during hiatus periods. the same (fig. S1I). The prevalence of La Niña, North and South Atlantic together contributed

Fig. 4. EOF of the global oceans. The first (left column) and second (right column) EOF modes of the OHC in the layers 0 to 300 m (A and B) and 300
to 1500 m (C and D), respectively, in the same unit of 1019 J. (E and F) The PC time series associated with each EOF mode (blue for the upper layer and
red for the lower layer). The PCs have been normalized by the standard deviation, which is multiplied to the EOF modes. The percentage of variance
explained by each EOF is indicated in each EOF panel. Because the second EOF is small, attention should be focused on the first EOF.

900 22 AUGUST 2014 • VOL 345 ISSUE 6199 sciencemag.org SCIENCE

 R ES E A RC H | R ES E A R C H A R T I C L E S

0.16 T a standard deviation of 0.03 W m−2 (per (westward) trade wind (23). The warm water was decades. Curry et al. (27) reported a freshening
Earth’s area) to the global mean of 0.30 T 0.04 W only slightly displaced downward and can quickly of the subpolar North Atlantic waters from the
m−2, slightly more than the rest of the global slosh up and eastward when the west trade-wind 1950s to 1990s and cited global warming as one
oceans combined. [The global value for the entire anomaly relaxes, as suggested by England et al. of the reasons. However, over the longer period
0 to 1500 m in the Ishii data is 0.49 T 0.05 W m−2.] (23). The Pacific mean is further broken down reported here, there was very little trend in the
Over a meridional section (Fig. 5, A and B) of into the tropical versus the subtropical Pacific in salinity, only a vacillating cycle.
OHC across the Atlantic basin, there was a dra- fig. S5. There is some indication of heat subduc- There is no accepted single theory, but theo-
matic and broad shift in how heat penetrated all tion by the subtropical cell, as previously pro- retical explanations of these vacillating regime
the way to 1500 m in mid- to high-latitude North posed (23, 24), but the magnitude is small and shifts mostly involve variations of the AMOC
Atlantic in 1999 to 2012 relative to the prior decades. does not affect the Pacific mean shown in Fig. 5C. (20, 25). Because of excessive evaporation, trop-
The detrended version is shown in fig. S3; the spa- ical surface water is more saline. A faster AMOC
tial pattern in the North Atlantic is consistent with The salinity mechanism tends to transport more saline water to the North
a change of the thermohaline circulation (19, 20). A mechanism that can account for the speed Atlantic subpolar region, where it loses some
There is a “dipole” meridional pattern near the with which heat penetrates to such great depths heat to the cold atmosphere and sinks. The
200-m depth under a “monopole” SST warming, is deep convection caused by vertical density dif- heat from the transported tropical water tends
similar to the pattern generated by regressing the ferences. Salinity changes at subpolar North At- to melt more ice (20), which makes the surface
data onto either the Atlantic Meridional Overturn- lantic are known to affect deep-water formation water in the subpolar regions less dense. These
ing (AMO) index (21, 22) or the detrended tropical to initiate such an ocean circulation shift (25). two effects oppose each other. Eventually the
North Atlantic SST, known from model “water hos- The salinity there (Fig. 6A) shifted to a positive fresh water from ice melts wins out. The less-
ing” experiments to be a manifestation of variations anomaly that penetrated vertically to 1500 m dense water then slows the AMOC after a few-
in the Atlantic Meridional Overturning Circulation very rapidly in the 21st century, reaching histor- years lag (28). A slower AMOC then transports
(AMOC) (22). The spatial pattern of this possibly ically high values since measurements began (26). less tropical saline water northward, and the
recurring internal variability can be seen in the This is in contrast to the negative anomaly dur- opposing phase of the cycle commences. Other
longer records shown in fig. S4, although poorer, ing the prior three decades, when surface warm- versions of the mechanism are also available
earlier data are needed to see the previous episode. ing was rapid. Because the ocean data were less (29, 30). Because the record for the AMOC strength
In contrast to the Atlantic, the OHC change sparse in the North Atlantic, we extend the plot is short (only since 2004), the above scenario can-
in the Pacific Ocean occurs within a shallow back to 1950 and reveal that the salinity anomaly not be verified observationally, but it is consistent
layer (above 300 m) and mainly in the east- was also positive before 1970, during another with the mechanism described by Danabasoglu
west direction (see Fig. 5, C and D, which are episode of surface hiatus. These salinity shifts et al. (31) as indeed occurring in the Community
the meridional mean of the Pacific Ocean), correspond well in timing to the OHC shifts (Fig. Climate System Model 4 (CCSM4) (32).
similar to that during ENSO events. During the 6B), which are also coincident with surface tran- Although there is very little trend in the OHC
current period, a thin layer of colder surface water sitions from global-warming slowdown to rapid in the subpolar North Atlantic where the salinity-
overlies a very warm pool of water in the tropical warming and then to the current slowdown, with induced vacillation cycle dominates, there is a
western Pacific, driven by an anomalously strong intervals between shifts lasting about three linear OHC trend equatorward of 45°N and °S in

Fig. 5. Contrasting the change in heat content in the Pacific versus the Atlantic. (A and B) OHC (for each 5-m layer) in the Atlantic basin zonally
averaged over the basin as a function of latitude. (C and D) OHC integrated meridionally (35°S to 65°N) over the Pacific basin as a function of longitude.
(A and C) Averaged over the recent 14 years. (B and D) Averaged over the previous 14 years. Not detrended.

SCIENCE sciencemag.org 22 AUGUST 2014 • VOL 345 ISSUE 6199 901

R ES E A RC H | R E S EA R C H A R T I C LE S

the Atlantic basin (including the Southern Ocean) The present work follows the original proposal of tions of the AMOC, then we can infer that the time
(fig. S6), which is likely anthropogenically forced. Meehl et al. (5, 24) regarding global deep-ocean scale of the latter was also 30 years between re-
This secular increase in OHC is not associated heat sequestration. However, our observational gime shifts historically. Tung and Zhou (41) showed
with a corresponding trend in salinity. It reflects result does not support their Pacific-centric view. that these multidecadal cycles exist in the Central
mainly the increase in greenhouse heating from The duration of the cooling periods in the CCSM4 England Temperature for the past 353 years, with
above and does not necessarily reflect the speed model they used is typically 10 years, with one an average period of 70 years (but could be as
of the heat transport by the AMOC. In the Pacific, rare 15-year hiatus in 375 years and none over short as 40 years), consistent with an earlier
there is again very little heat-uptake trend. 15 years. The current hiatus already lasted over study using multiproxy data by Delworth and
15 years using their definition of hiatus as pe- Mann (44). They and Zhang et al. (42) argued
Discussion riods with zero trend. Comparing that model with that this oscillation is due to an internal variabil-
Many different explanations for the recent global- observation, we found that model’s Atlantic has ity and not to anthropogenic aerosol forcing (37).
warming slowdown have been proposed, but they too little variability with too high frequency (fig. Because the atmosphere and the upper ocean
fall mainly into two categories. The first involves a S7 versus Fig. 6). This artifact appears to be at- in the equatorial Pacific are intimately coupled,
reduction in TOA radiative forcing: by a decrease tributable to a new overflow parameterization the strong trade winds go hand in hand with the
in stratospheric water vapor (33), an increase in scheme in CCSM4 in the Denmark Strait and east-west SST gradient generated by the La Niña
background stratospheric volcanic aerosols (34), Faroe Bank Channel (31). pattern, known as the Bjerknes feedback. Kosaka
by 17 small volcano eruptions since 1999 (35), in- From another perspective, Wu et al. (43) and Xie (16) proposed that the cold eastern Pa-
creasing coal-burning in China (36), the indirect showed that the instrumental global-mean sur- cific was key to understanding the current warm-
effect of time-varying anthropogenic aerosols (37), face temperature record since 1850 contains two ing hiatus. England et al. (23) attributed the
a low solar minimum (38), or a combination of and a half cycles of a multidecadal variation. Each cold eastern equatorial Pacific to the stronger
these (39). Response to solar cycle changes was cycle lasted on average 65 years, consisting of an trade winds for the past 20 years, and the trade
found to be small (40, 41). The aerosol cooling accelerated warming plus a slowdown period. wind intensification to the multidecadal nega-
should have a signature in subsurface ocean (42), They showed that the primary location of the tive phase of the Interdecadal Pacific Oscillation
and yet it is not seen, perhaps suggesting that multidecadal variability is in the North Atlantic (IPO). The IPO, defined using the SST in the Pa-
the proposed radiative effects may be too small. and only secondarily in the Pacific. If this multi- cific Basin including the ENSO region, describes
The second involves ocean heat sequestration: decadal variability is indeed related to the varia- but does not explain the cold eastern Pacific status.

Fig. 6. Climate shifts in

salinity and OHC. North
Atlantic subpolar (45° to
65°N) mean salinity (A) and
5-m layer OHC (B),
12-month running mean, as
a function of years; not
detrended, but the
climatology for the period
1950 to 2012 was removed.

902 22 AUGUST 2014 • VOL 345 ISSUE 6199 sciencemag.org SCIENCE

 R ES E A RC H

How could the Pacific SST cool when the heat sink 10. M. Ishii, M. Kimoto, J. Oceanogr. 65, 287–299 (2009). 38. J. Hansen, M. Sato, P. Kharecha, K. von Schuckmann,
was located in other ocean basins? Why didn't 11. J. M. Lyman, G. C. Johnson, J. Clim. 21, 5629–5641 (2008). Atmos. Chem. Phys. 11, 13421–13449 (2011).
12. M. Ishii, A. Shouji, S. Sugimoto, T. Matsuoto, Int. J. Climatol. 39. G. A. Schmidt, D. T. Shindell, K. Tsigaridis, Nat. Geosci. 7,
the Atlantic SST simply cool as heat was being 25, 865–879 (2005). 158–160 (2014).
subducted in its basin? The Atlantic SST and its 13. M. Ishii, M. Kimoto, K. Sakamoto, S. I. Iwasaki, J. Oceanogr. 62, 40. J. Zhou, K. K. Tung, J. Clim. 23, 3234–3248 (2010).
upper layers did start to cool (Fig. 1B) after its 155–170 (2006). 41. K. K. Tung, J. Zhou, Proc. Natl. Acad. Sci. U.S.A. 110,
subpolar salinity peaked and then started to de- 14. C. A. Katsman, G. J. van Oldenborg, Geophys. Res. Lett. 38, 2058–2063 (2013).
L14610 (2011). 42. R. Zhang et al., J. Atmos. Sci. 70, 1135–1144 (2013).
crease after 2006 (Fig. 6A). Before 2006, our 15. M. A. Balmaseda, K. Mogensen, A. T. Weaver, Q. J. R. 43. Z. Wu, N. E. Huang, J. M. Wallace, B. Smoliak, X. Chen,
warm salt subduction mechanism does not allow Meteorol. Soc. 139, 1132–1161 (2013). Clim. Dyn. 37, 759–773 (2011).
the Atlantic to cool when its subpolar salinity 16. Y. Kosaka, S.-P. Xie, Nature 501, 403–407 (2013). 44. T. L. Delworth, M. E. Mann, Clim. Dyn. 16, 661–676 (2000).
was increasing, because poleward transport of 17. S. T. Gille, Science 295, 1275–1277 (2002). 45. R. Zhang, T. L. Delworth, Geophys. Res. Lett. 34, L23708
18. J. R. Toggweiler, J. Russell, Nature 451, 286–288 (2008). (2007).
warm salty water and increasing subpolar sub- 19. R. Zhang, Geophys. Res. Lett. 35, L20705 (2008). 46. R. Zhang, T. L. Delworth, I. M. Held, Geophys. Res. Lett. 34,
duction are parts of the same mechanism of en- 20. J. Jungclaus, H. Haak, M. Latif, U. Mikolajewicz, J. Clim. 18, L02709 (2007).
hanced AMOC upper-ocean transport. During 4013–4031 (2005). 47. W. T. Welch, K. K. Tung, J. Atmos. Sci. 55, 1285–1302 (1998).
this first part of the hiatus period, the heat deficit 21. C. Wang, L. Zhang, J. Clim. 26, 6137–6162 (2013).
22. R. Zhang, Geophys. Res. Lett. 34, L12713 (2007). AC KNOWLED GME NTS
must be transferred to other ocean basins, mostly 23. M. H. England et al., Nat. Clim. Change 4, 222–227 (2014). The research of K.-K.T. is supported by NSF under AGS-1262231.
to the Pacific because it is the only other major 24. G. A. Meehl, A. Hu, J. M. Arblaster, J. T. Fasullo, X.C. is supported by Natural Science Foundation of China under
ocean basin in the Northern Hemisphere, likely K. E. Trenberth, J. Clim. 26, 7298–7310 (2013). 41330960 and 41176029. He thanks First Institute of Oceanography
through the atmosphere. Zhang and Delworth 25. I. V. Polyakov, V. A. Alekseev, U. Bhatt, E. I. Polyakova, of China for its support during his visit to University of
X. Zhang, Clim. Dyn. 34, 439–457 (2010).
(45) and Zhang et al. (46) showed by using models 26. H. Hátún, A. B. Sandø, H. Drange, B. Hansen, H. Valdimarsson,
Washington. The authors are grateful to many colleagues and
the anonymous reviewers who commented on the manuscript
that, as the northward heat transport by the AMOC Science 309, 1841–1844 (2005). and helped improve it. The data sets used in this study are publicly
is increased, the atmospheric heat transport de- 27. R. Curry, B. Dickson, I. Yashayaev, Nature 426, 826–829 (2003). available in http://rda.ucar.edu/datasets/ds285.3/ and https://
creases in compensation (and vice versa), pro- 28. M. Dima, G. Lohmann, J. Clim. 20, 2706–2719 (2007). climatedataguide.ucar.edu/climate-data/oras4-ecmwf-ocean-
29. M. G. Wyatt, J. A. Curry, Clim. Dyn. 42, 2763–2782 (2014). reanalysis-and-derived-ocean-heat-content.
viding a multidecadal component to the Pacific 30. M. G. Wyatt, S. Kravtsov, A. A. Tsonis, Clim. Dyn. 38, 929–949 (2012).
Decadal Oscillation (PDO). The concept behind 31. G. Danabasoglu et al., J. Clim. 25, 5153–5172 (2012). SUPPLEMENTARY MATERIALS
the flexible atmospheric heat transport is known 32. P. R. Gent et al., J. Clim. 24, 4973–4991 (2011).
www.sciencemag.org/content/345/6199/897/suppl/DC1
as nonlinear baroclinic adjustment (47). Thus, 33. S. Solomon et al., Science 327, 1219–1223 (2010).
Materials and Methods
34. S. Solomon et al., Science 333, 866–870 (2011).
almost-“synchronized” hemisphere-wide atmo- 35. B. D. Santer et al., Nat. Geosci. 7, 185–189 (2014).
Data
spheric changes are possible (30). On climate Figs. S1 to S7
36. R. K. Kaufmann, H. Kauppi, M. L. Mann, J. H. Stock, Proc. Natl.
Reference (48)
time scales and from an energy perspective, the Acad. Sci. U.S.A. 108, 11790–11793 (2011).
amount of radiative energy available to heat the 37. B. B. B. Booth, N. J. Dunstone, P. R. Halloran, T. Andrews, 17 April 2014; accepted 11 July 2014
N. Bellouin, Nature 484, 228–232 (2012). 10.1126/science.1254937
global SST, including the Pacific, is what remains
after accounting for the energy sink; it does not
matter if the latter is located outside the Pacific.
◥
Conclusion
The fact that the global-mean temperature, along
REPORTS
with that of every major ocean basin, has not
increased for the past 15 years, as they should QUANTUM OPTICS
in the presence of continuing radiative forcing,
requires a planetary sink for the excess heat. Al-
though the tropical Pacific is the source of large
interannual fluctuations caused by the exchange
All-optical routing of single photons
of heat in its shallow tropical layer (3), the current
slowdown is in addition associated with larger by a one-atom switch controlled
decadal changes in the deeper layers of the At-
lantic and the Southern oceans. The next El Niño,
when it occurs in a year or so, may temporarily
by a single photon
interrupt the hiatus, but, because the planetary Itay Shomroni,* Serge Rosenblum,* Yulia Lovsky, Orel Bechler,
heat sinks in the Atlantic and the Southern Gabriel Guendelman, Barak Dayan†
Oceans remain intact, the hiatus should continue
The prospect of quantum networks, in which quantum information is carried by single photons
on a decadal time scale. When the internal varia-
in photonic circuits, has long been the driving force behind the effort to achieve all-optical
bility that is responsible for the current hiatus
routing of single photons. We realized a single-photon–activated switch capable of routing a
switches sign, as it inevitably will, another episode
photon from any of its two inputs to any of its two outputs. Our device is based on a single atom
of accelerated global warming should ensue.
coupled to a fiber-coupled, chip-based microresonator. A single reflected control photon toggles
RE FE RENCES AND N OT ES the switch from high reflection (R ~ 65%) to high transmission (T ~ 90%), with an average
1. K. E. Trenberth, J. T. Fasullo, M. A. Balmaseda, J. Clim. 27, of ~1.5 control photons per switching event (~3, including linear losses). No additional control
3129–3144 (2014). fields are required. The control and target photons are both in-fiber and practically identical,
2. Intergovernmental Panel on Climate Change, Summary for making this scheme compatible with scalable architectures for quantum information processing.

P
Policymakers, Climate Change 2007: Synthesis Report
(Cambridge Univ. Press, Cambridge, 2007). hotons are a key player in the growing behavior at the level of single photons—namely,
3. K. E. Trenberth, J. T. Fasullo, Earth’s Future 1, 19–32 (2013).
4. K. E. Trenberth, J. T. Fasullo, Science 328, 316–317 (2010).
field of quantum information science. photon-photon interactions—is considered a major
5. G. A. Meehl, J. M. Arblaster, J. T. Fasullo, A. Hu, K. E. Trenberth, The fact that they do not interact with challenge also in the realization of quantum net-
Nat. Clim. Change 1, 360–364 (2011). each other has made them ideal for the works, in which quantum information processing
6. D. R. Easterling, M. F. Wehner, Geophys. Res. Lett. 36, L08706 communication of quantum information
(2009). Department of Chemical Physics, Weizmann Institute of
7. D. Roemmich et al., Oceanography 22, 46–55 (2009).
yet has prevented so far the realization of deter- Science, Rehovot 76100, Israel.
8. J. M. Lyman, G. C. Johnson, J. Clim. 27, 1945–1957 (2014). ministic all-optical quantum gates based on sin- *These authors contributed equally to this work. †Corresponding
9. J. M. Lyman et al., Nature 465, 334–337 (2010). gle photons. The difficulty to achieve nonlinear author. E-mail: [email protected]

SCIENCE sciencemag.org 22 AUGUST 2014 • VOL 345 ISSUE 6199 903