PUBLICATIONS
Geophysical Research Letters
RESEARCH LETTER
10.1002/2015GL063083
Key Points:
• SST anomalies in the Pacific were a
primary cause of the severe winter
of 2014
• Warmth in the tropical west Pacific led
to cold in central North America
• Specified SST model experiments
replicate the observed
teleconnections

Supporting Information:
• Readme and Figures S1–S5
Correspondence to:
D. L. Hartmann,
dhartm@washington.edu

Citation:
Hartmann, D. L. (2015), Pacific sea
surface temperature and the winter
of 2014, Geophys. Res. Lett., 42,
doi:10.1002/2015GL063083.
Received 8 JAN 2015
Accepted 17 FEB 2015
Accepted article online 19 FEB 2015

Pacific sea surface temperature
and the winter of 2014
Dennis L. Hartmann1
1

Department of Atmospheric Sciences, University of Washington, Seattle, Washington, USA

Abstract It is shown from historical data and from modeling experiments that a proximate cause of the
cold winter in North America in 2013–2014 was the pattern of sea surface temperature (SST) in the Pacific
Ocean. Each of the three dominant modes of SST variability in the Pacific is connected to the tropics and has a
strong expression in extratropical SST and weather patterns. Beginning in the middle of 2013, the third mode
of SST variability was two standard deviations positive and has remained so through January 2015. This
pattern is associated with high pressure in the northeast Pacific and low pressure and low surface temperatures
over central North America. A large ensemble of model experiments with observed SSTs confirms that SST
anomalies contributed to the anomalous winter of 2014.

1. Introduction
Much interest has focused on the very cold winter in the central and eastern U.S. and Canada in 2013–2014
(hereafter the winter of 2014). It has been suggested that melting of polar sea ice could be responsible for
causing unusual weather events (Francis and Vavrus [2012] but see Barnes [2013]), perhaps through the
intermediacy of stratospheric warmings [Kim et al., 2014]. This would be remarkable given the small area
of the Arctic Ocean and its presence at the tail end of the atmospheric and oceanic chain moving energy from
the tropics toward the poles. Taking a different perspective, Ding et al. [2014] use models and observations to
show that at least some of the recent Arctic warming has been forced by changes in the tropical Pacific. Here
the coupled variability between the Pacific Ocean and the global atmosphere is used to account in part for the
unusual winter of 2014. Using a combination of analysis of past data and a large ensemble of specified sea surface
temperature (SST) modeling experiments, a strong case can be made that variability of the ocean-atmosphere
system originating in the tropics is the most likely “cause” of the winter of 2014 anomalies.
The ocean provides the long-term memory for the climate system through its higher heat capacity, slowly
varying large-scale modes, and long adjustment time to changes in forcing. The El Niño–Southern Oscillation
(ENSO) phenomenon is a coupled ocean-atmosphere phenomenon that is the dominant mode of interannual
variability globally and has provided the opportunity to make useful forecasts of seasonal weather and
climate conditions a season or more in advance [Neelin et al., 1998; Rasmusson and Wallace, 1983]. ENSO has a
related decadal variability, the Pacific Decadal Oscillation (PDO), which was first defined in terms of North Pacific
sea surface temperature (SST) anomalies [Mantua and Hare, 2002; Zhang et al., 1997], and is now thought to
influence global mean temperature on decadal time scales [Huber and Knutti, 2014; Kosaka and Xie, 2013;
Trenberth and Fasullo, 2013]. Interannual anomalies of SST are known to have a significant influence on climate
over land, and these relationships play an important role in seasonal climate forecasting [Palmer and Anderson,
1994] and in the explanation of drought [Seager and Hoerling, 2014]. While the connection between tropical
SST anomalies and winter weather in the extratropics is well established [Horel and Wallace, 1981], SST
anomalies in middle latitudes are driven by atmospheric anomalies and it is more difficult to show an
influence of extratropical SST anomalies in driving atmospheric anomalies [Kushnir et al., 2002; Lau and Nath,
1994]. The extratropical SST anomalies associated with ENSO are caused primarily by atmospheric anomalies
propagating from the tropics [Alexander et al., 2002].
In this paper we explore anomalies in monthly mean SST in the Pacific and their relation to recent weather
anomalies. We first present a decomposition of Pacific Ocean SST variability into three key modes: ENSO,
which has its strongest variability on time scales of 2 to 6 years, and two decadal modes that are related to
ENSO and have a strong expression in the North Pacific. We next relate these SST patterns to atmospheric
variability and show that one of the decadal modes is associated with atmospheric anomalies that look very

HARTMANN

©2015. American Geophysical Union. All Rights Reserved.

1

 Geophysical Research Letters

10.1002/2015GL063083

much like those during the winter of 2014. This mode of variability showed a positive anomaly of two
standard deviations beginning in the summer of 2013. To demonstrate a causality relationship between
this large anomaly in the SST mode and the winter of 2014, we analyze a large ensemble of specified SST
modeling experiments. Since these specified SST experiments produce anomalies with similar structure and
amplitude to the observed anomalies in 2014, we conclude that SST variations rooted in the tropics played a
key role in producing the cold winter of 2014.

2. SST Variability: ENSO, PDO, and North Pacific Mode
The Pacific Decadal Oscillation (PDO) has classically been defined as the first empirical orthogonal function
(EOF) of the anomalies of monthly mean SST poleward of 20°N in the Pacific Ocean [Mantua et al., 1997]. Bond
et al. [2003] noted that the second EOF of North Pacific SST was anomalously negative in 1999–2002 and
emphasized that the PDO alone is not sufficient to characterize the variability of the North Pacific. Here we
include the ocean region from 30°S to 65°N and 120°E to 105°W, which includes the tropical and North Pacific.
We use the NOAA Extended Reconstructed Sea Surface Temperature (ERSST) data set [Smith et al., 2008] to
compute the EOF decomposition after removing the seasonal cycle, the global mean, and the linear trend
from the monthly data. We have obtained similar results with the Hadley Center Sea Ice and SST Data Set
(HadISST) [Rayner et al., 2003] (not shown). Although the data set starts in 1854, we focus on the period since
1900 when the data are more complete. The first EOF expresses about 30% of the variance, and the next two
modes each express about 8% of the variance (Figure S1 in the supporting information). The second and third
modes are not distinguished from each other by their explained variance [North et al., 1982], and their relative
rankings will change for different sampling periods. For example, the third mode for the period from 1900
becomes the second mode in the period since 1979. The remaining EOFs form a continuum of modes that
are not distinguished from each other by their explained variance. The SST modes shown here are for all
12 months, but the basic structures are the same if winter or summer half years are used instead. The structures
and explained variances are very similar if we use instead only the period since 1950, when the observational
network had improved further. The first three SST modes are shown in Figure 1 as regressions of the SST
anomalies with the principal component time series of each mode. The first mode captures much of the
equatorial variance, but the second and third modes also have significant connections to the tropics.
In the supporting information we show that these three modes are related to each other (Figure S2). The
second mode tends to become strongly positive after a warm ENSO event. This expresses the fact that the
extratropical signature of a warm event, which gives cold SST in midlatitudes via the “atmospheric bridge”
[Lau and Nath, 1996], tends to persist longer than ENSO itself. So its structure is cold in midlatitude western
Pacific and also cold in the eastern equatorial Pacific. The third mode tends to be in its positive phase prior
to ENSO warm events and is closely related to the “footprinting” mechanism [Vimont et al., 2001, 2003].
The structures of the second and third modes in Figure 1 appear to be dominated by amplitude away from
the equator, but they are nonetheless connected to the tropics. The variance of SST is large in the North
Pacific (not shown), so the regressions are large there, but both the correlation patterns associated with these
two modes and their strong temporal relationship to ENSO indicate their tropical roots.
During the satellite era since 1979, the third mode has been very strong, and it enters as the second mode of
global SST during that period (not shown). In the classical analysis of the PDO using the region poleward of
20°N [Mantua and Hare, 2002], the first mode is that identified as the PDO, while the second mode is very
similar to the third mode for the larger domains employed here. Deser and Blackmon [1995] referred to the
second mode SST structure as the North Pacific Mode (NPM). The NPM mode is very robust to the region
used. It appears as the second mode for the Pacific region north of 20°N, one of two modes centered in the
North Pacific for the Pacific Ocean north of 30°S shown here, and also as the second mode of global SST for
the period since 1979.
The extratropical signatures of the second and third modes in Figure 1 may be influenced by the North Pacific
Gyre Oscillation [Ceballos et al., 2009; Chhak et al., 2009; Di Lorenzo et al., 2008] and by fluctuations in the
Kuroshio and Oyashio Currents [Frankignoul et al., 2011; Kwon et al., 2010]. Seager et al. [2001] performed
an ocean heat budget analysis in the Kuroshio-Oyashio Extension region and showed how surface fluxes,
Ekman current-induced ocean heat convergence, and Rossby wave adjustment combine to generate the
SST anomalies. Thus, both atmospheric forcing and the oceanic dynamic response are important for the

HARTMANN

©2015. American Geophysical Union. All Rights Reserved.

2

 Geophysical Research Letters

10.1002/2015GL063083

a)

EOF 1

b)

EOF 2

c)

EOF 3

Figure 1. The regressions of monthly mean anomalies of global SST onto the first three EOFs of SST for the ocean area 30°S–65°N,
120°E–105°W and the period from January 1900 to July 2014. Contour interval is 0.1°C. Positive values are red, negative values are
blue, and the zero contour is white. Robinson projection from 60°S to 60°N.

evolution of the North Pacific SST and surface height [e.g., Miller et al., 1998; Vivier et al., 1999; Qiu, 2000;
Cummins and Freeland, 2007]. Local atmospheric forcing seems to be very important in the North Pacific
at least on seasonal and interannual time scales, and both local and remote influences may be important
[Alexander et al., 2002; Liu and Alexander, 2007]. Smirnov et al. [2014] conclude that internal ocean dynamics
play little role in interannual SST variability in the eastern Pacific in comparison to the western Pacific where
the Kuroshio and Oyashio Currents may play a more important role.
The principal components of the first three EOFs are shown in Figure 2 in units of standard deviations for
the period from 1979 to January 2015. These EOFs are based on the Pacific north of 30°S during the period
from 1900 to 2014. During the most recent period shown in Figure 2, hints of the relationships between the
three EOFs during warming events can be seen. Both the second and third modes show large-amplitude
variations associated with major warming events. Since the 1998 warm ENSO event, the principal components
of the first two EOFs have been mostly negative, which is a reflection of the tendency for cool ENSO and
negative PDO since then. After May 2013, however, the first two EOFs were near neutral amplitude, while the
third EOF (NPM) maintained a positive value near two standard deviations above neutral. Unusually warm
temperatures in the northeast Pacific prevailed during this period.
The very high SST anomaly in the North Pacific that emerged in May 2013 is linked to reduced cooling of
the ocean during the prior winter of 2013. The SST, heat content, surface wind stress, and surface heat flux
HARTMANN

©2015. American Geophysical Union. All Rights Reserved.

3

 Geophysical Research Letters

10.1002/2015GL063083

8

6

anomalies for 2013 are well illustrated
in the State of the Climate 2013 report
[Blunden and Arndt, 2014], particularly
in Figures 3.1 to 3.9 of Newlin and Gregg
[2014]. These show strongly suppressed
wind stresses and positive anomalies of
surface heat fluxes into the North
Pacific Ocean during 2013. It is likely
that these anomalies are rooted in the
tropics, but it is important to confirm
what caused these anomalous fluxes in
2013 and to what extent the heat
stored then influenced the weather of
the following winter.

PC1

Standard Deviations

4

2

PC2

0

−2

PC3
−4

−6

−8

1980

1985

1990

1995

2000

2005

2010

2015

Year

Figure 2. Time series of the principal components of the first three EOFs of
monthly mean Pacific SST poleward of 30°S as shown in Figure 1. Only
values from January 1979 to January 2015 are shown. Units are standard
deviations. Values are offset by four standard deviations for clearer
viewing.

3. Meteorological
Connections to Pacific
SST Anomalies

We have regressed the monthly
principal components of the first three
EOFs of SST on the 500 hPa height and
lowest model level temperature from
both the Twentieth Century Reanalysis that starts in 1871 [Compo et al., 2011] and the National Centers for
Environmental Prediction (NCEP)/National Center for Atmospheric Research (NCAR) Reanalysis that starts in
1948 [Kalnay et al., 1996]. We obtain similar results from both reanalysis data sets and from shorter subsets of
each data set. For the sake of brevity we show the results for the NCEP/NCAR Reanalysis.
Since the focus of this paper is the influence of the North Pacific Mode on the winter of 2014, we will
show only the regression for that mode here, but correlation patterns for all three modes are shown in the
supporting information (Figure S3). Figure 3a shows the anomaly of the 500 hPa field for November 2013
to March 2014, the winter of 2014. It shows a ridge over the North Pacific and a deep trough over the central
North America centered just south of Hudson Bay. The northerly flow between these two centers, which is
roughly aligned with the Rocky Mountains, brought cold Arctic air into central North America and gave
unusually cold conditions there.
Figure 3b shows the regression of the global 500 hPa height field from NCEP/NCAR reanalysis during months
of November through March onto the second principal component of global SST based on the period from
1979. The second mode of global SST for 1979 to 2014 is very similar to the third mode of SST for the period
1900 to 2014 shown in Figure 1. Figure 3b is not very different when based on the SST since 1900 and the
NCEP/NCAR data since 1948, but we show the satellite era data to be consistent with the model simulations
discussed later. The regressions show the amplitude explained by a one-standard deviation variation of
the principal component. A large high pressure occurs over the northern Pacific centered near the Aleutian
Islands with a downstream low centered near Hudson Bay. These are connected to a low along 30°N over the
Pacific Ocean.

Figure 4 shows the same anomalies and regressions as in Figure 3, except for the temperature anomalies at
the lowest model level. During the winter of 2014 extreme cold anomalies persisted across North America
while warm anomalies persisted in the northeast Pacific and in Alaska. These anomalies are also reflected
in the regression of the North Pacific Mode of SST onto the low-level temperatures in Figure 4b. The
regression shows a maximum cold anomaly of about 1°C over North America for one standard deviation
departure of the NPM; so for two standard deviations of the NPM SST, the regression predicts a surface
temperature anomaly of about 2°C over North America, which is close to the 2.6°C anomaly in Figure 4a.
The statistical significance of the features in the regression plots shown here is assessed by determining
whether the correlations associated with them are different from zero at the 95% level. The degrees of
freedom have been determined using the approximation recommended by Bretherton et al. [1999]. Although

HARTMANN

©2015. American Geophysical Union. All Rights Reserved.

4

 Geophysical Research Letters

10.1002/2015GL063083

a) Observed 500 hPa Height Anomaly Nov-March 2013-14

50

0

−50

b) Regression of NCEP/NCAR Reanalysis onto EOF2 of global SST 1979-2014
20
10
0
−10
−20

c) Regression of ESRL-GFSv2 Ensemble onto EOF2 of global SST 1979-2014
20
10
0
−10
−20

Figure 3. (a) The observed anomaly in 500 hPa height for November 2013 to March 2014 (contour interval is 10 m, positive
is red, negative is blue, and the zero contour is white). (b) Regression of the 500 hPa height anomalies for the months of
November to March for the years from 1979 to 2014 onto the principal component time series for the second EOF of global
SST for the period from 1979 to 2014 (contour interval is 3 m). (c) Same as Figure 3b, except that the height data are
from the 50-member ensemble from the Earth System Research Laboratory-Global Forecast System version 2 (ESRL-GFSv2)
using the observed SST.

the SST indices are highly autocorrelated, the extratropical height anomalies are not, so that for the 67 year
record from 1948 and the 5 months of November to March, the data used to perform the height regressions
possess about 250 degrees of freedom. For the record since 1979, the degrees of freedom still number
130. With this large number of degrees of freedom, relatively small correlations become significant, and the
question shifts to whether the fraction of variance explained is interesting or not. Thus, the absolute values of
the correlation coefficients are of more interest than their proven statistical significance. The correlation
coefficient map corresponding to Figures 3b and 4b have correlations between 0.3 and 0.4 for the centers
over North America. For comparison, the correlations associated with the ENSO mode exceed 0.6 in the

HARTMANN

©2015. American Geophysical Union. All Rights Reserved.

5

 Geophysical Research Letters

10.1002/2015GL063083

a) NCEP/NCAR 995 hPa Temperature Anomaly Nov-March 2013-14

4

2

0

−2

−4

b) Regression NCEP/NCAR Reanalysis onto EOF2 of global SST 1979-2014
1
0.5
0
−0.5
−1

c) Regression of ESRL-GFSv2 Ensemble onto EOF2 of global SST 1979-2014

0.5

0

−0.5

Figure 4. The same as Figure 3, except that the variable plotted is the temperature at the lowest model level. Contour
interval is (a) 0.5 and (b and c) 0.1.

extratropics for SST EOFs computed from both the global and Pacific regions. We therefore conclude that
the winter of 2014 over North America was favored by the prevailing NPM SST anomaly pattern over the
Pacific Ocean.
Wang et al. [2013] explore the influence of SST modes on North American weather. They estimate that the
three leading modes they consider explain about 50% of the total effect of SST on North American climate,
but the modes they considered in the Pacific correspond roughly to the ENSO mode and the PDO mode
(or Kuroshio extension mode), and they do not consider the NPM described here. Since our analysis shows
that the second and third modes explain the same amount of variance, any linear combination of the two
is as valid a representation as either mode alone. Since the third mode reaches large amplitude during the
period of interest, while the first and second modes are near neutral, it is meaningful to consider only the
third mode in detail here.

HARTMANN

©2015. American Geophysical Union. All Rights Reserved.

6

 Geophysical Research Letters

10.1002/2015GL063083

The pattern of SST and height variability we find with the NPM is similar to the structure that Seager et al.
[2014] associate with the California drought, particularly during the November 2013 to April 2014 period.
They used model experiments with specified SST to show that these anomalies are keyed to warm SST
anomalies in the tropical Pacific west of the dateline. The ridge-trough pattern over North America suppresses
precipitation in California while it also induces cold anomalies in the middle of the continent. We emphasize
the cooling effect here. It should be noted that the prior two winters of 2011–2012 and 2012–2013 were not
strong positive NPM years (Figure 1). The ridge was farther west offshore, and the deep trough over North
America was much less developed. The drought during those years was more likely associated with the strong
La Niña conditions during the 2011–2012 winter, with continued warm SST in the western equatorial Pacific
the following year, and the strongly positive NPM mode in 2013–2014 completed 3 years of intense drought
in California.
Similarly, Wang et al. [2014] associated the California drought of 2013–2014 with the anomaly height pattern
for November-December-January, which is very similar in structure to Figure 3. They focused particularly on
the dipole pattern between the ridge in the Gulf of Alaska and the trough centered just south of Hudson Bay.
They found that this pattern was correlated with the Niño4 SST index, 1 year prior. In general outline this
is consistent with our conclusion that the winter of 2014 pattern is associated with a precursor mode of ENSO.
Wang et al. [2014] go on to conclude that this pattern is becoming more prevalent in consequence of
human-induced warming of the climate. Seager et al. [2014] conclude that the California drought is largely a
natural variability and has precedents in earlier times. We do not investigate any connections to climate
change in this paper.

4. Specified SST Experiments
So far, it has been demonstrated that the structure that dominated the winter weather in North America in
2014 is similar in shape and slightly smaller than the regression onto the NPM SST pattern that dominated
in that year. To make a stronger case that the SST anomaly caused some important part of the weather
anomaly, rather than merely being associated with it, we turn to Atmospheric Model Intercomparison Project
(AMIP)-style experiment data provided by the NOAA-ESRL Physical Sciences Division, Boulder, Colorado, from
their website at http://www.esrl.noaa.gov/psd/repository/alias/facts/. The experiments chosen apply the
observed radiative forcing and specify the observed SSTs under an atmospheric model. Large ensembles of
experiments starting in 1979 were conducted with several different models of the atmosphere. Three model
ensembles are considered here; a 50-member ensemble with the ESRL-GFSv2 model, a 30-member ensemble
with the fifth-generation European Centre/Hamburg (ECHAM5) model, and a 20-member ensemble with
the Community Atmosphere Model 4 (CAM4). We use the ensemble mean anomalies for each model to both
regress them onto the SST indices. To be consistent with our data analysis, we remove the linear trend.
Figure 3c shows the regression of the 500 hPa height anomalies for the ESRL-GFSv2 50-member ensemble
mean onto the NPM mode. The ECHAM5 and CAM4 models produce similar results (Figure S5). The predicted
anomaly has key features in common with the observed anomaly in Figure 3a but with smaller amplitudes. The
depth of the low at about 35°N in the central Pacific is well simulated, but the ridge on the West Coast and
especially the downstream low over North America are much attenuated compared to the observed regression
pattern in Figure 3b.
Figure 4c shows the corresponding regression for the simulation of the temperature at the lowest model
level. The story is very similar to that for the 500 hPa height. The regression of the simulated temperatures
onto the NPM SST mode captures many of the key features of the anomalies of 2013–2014, especially the
cold conditions over North America. The regression pattern produced by the AMIP simulations has a much
weaker cool anomaly over North America than for the regression on observations. It is unclear why the
models underestimate the observed connections between SST and North American weather. The model
resolution in these experiments is 0.75 to 1°, and it is possible that the wave pattern is sensitive to the
resolution of the Rocky Mountains, which are smoothed somewhat in these simulations.
In future work we could look to see if a significant number of the ensemble members do get the observed
downstream amplitude in 2014, but that is an additional data processing chore. It is important to remember
that fixed SST experiments do not capture the full interaction between the atmosphere and ocean [Barsugli and
Battisti, 1998], and simple interpretations of ensembles often underestimate the confidence of seasonal

HARTMANN

©2015. American Geophysical Union. All Rights Reserved.

7

 Geophysical Research Letters

10.1002/2015GL063083

predictions [Eade et al., 2014]. For the 1979 to 2014 period there is little difference between the regression on
the global SST pattern and the NPM pattern for the Pacific north of 30°S. Although the winter of 2014 was a
2 sigma event in terms of the NPM of SST, the regressions with NCEP/NCAR Reanalysis include 35 (or 67 for
the period since 1948) winters and are not very sensitive to the inclusion or not of the winter of 2014.

5. Discussion and Conclusion
It appears that the sudden increase in the NPM SST anomaly pattern in the North Pacific during the middle of
2013 can be associated with the subsequent cold winter in the center of North America. Although the pattern of
SST anomaly is very strong in the North Pacific, it is virtually certain that the forcing for these anomalies originates
with warm SST in the tropical west Pacific. Preliminary experiments indicate that the atmospheric pattern during
the winter of 2014 is not efficiently forced by the extratropical SST anomalies but can be generated with the
observed warm SST anomalies in the tropical west Pacific (P. Ceppi, personal communication, 2015). This result
is consistent with a long history of observational and modeling work indicating that SST anomalies in the
extratropics are strongly driven by atmospheric circulation anomalies [Cayan, 1992; Davis, 1976], while SST
anomalies in the tropics can strongly force the atmospheric circulation [Lau and Nath, 1994].
The NPM mode of SST variability remains strongly positive as of this writing in January 2015, and its future
evolution will be interesting to watch. The reasons why the NPM anomalies arose so strongly in the middle of
2013 remain to be explored. The stage was set for the very positive NPM extratropical SST anomaly of the
2014 winter by a reduction in heat extraction from the North Pacific in the prior winter of 2013. The annual
mean 500 hPa anomaly for 2013 shows a large high pressure over the North Pacific centered at about 50°N,
150°W (not shown). Generally, warm anomalies of the SST in the tropical western Pacific have persisted
since the winter of 2013 (http://www.esrl.noaa.gov/psd/map/clim/sst.shtml).

Acknowledgments
The SST data were obtained from the
NOAA ERSST data set at http://www.esrl.
noaa.gov/psd/data/gridded/data.noaa.
ersst.html. Twentieth Century Reanalysis
and NCEP/NCAR Reanalysis products
were obtained from the NOAA websites
http://www.esrl.noaa.gov/psd/data/
20thC_Rean/ and http://www.esrl.
noaa.gov/psd/data/gridded/data.
ncep.reanalysis.html. AMIP simulations
from the NOAA website http://www.
esrl.noaa.gov/psd/repository/alias/
facts/ were crucial to this analysis. This
work was supported by NSF grant
AGS-0960497. The author gratefully
acknowledges colleagues, Nick Bond,
Paulo Ceppi, Greg Johnson, Todd
Mitchell, Richard Seager, LuAnne
Thompson, and Mike Wallace, whose
generous advice and comments
greatly improved this contribution.
The comments of two anonymous
reviewers of an earlier version of this
work were extremely helpful.
The Editor thanks Andrew Hoell and an
anonymous reviewer for their assistance
in evaluating this paper.

HARTMANN

Midlatitude seasonal weather and climate anomalies receive a large contribution from internal atmospheric
variability that is unrelated to any interactions with the SST [e.g., Deser et al., 2012]. The effect of ocean heat
anomalies and associated SST anomalies on the seasonal weather patterns can thus be negated by other
variability in any particular month. Nonetheless, the systematic connection between NPM SST anomalies and
winter weather in the past, the extreme amplitude of the SST anomalies in 2013–2014, and the similarity of
the 2014 winter anomalies to the historical pattern all suggest that the warm SST in the tropical and North
Pacific likely made a significant contribution to the cold winter in central North America in 2013–2014.
Specified SST simulations with three models confirm the causal relationship between the NPM SST mode and
winter weather in North America.

References
Alexander, M. A., I. Blade, M. Newman, J. R. Lanzante, N. C. Lau, and J. D. Scott (2002), The atmospheric bridge: The influence of ENSO
teleconnections on air-sea interaction over the global oceans, J. Clim., 15(16), 2205–2231.
Barnes, E. A. (2013), Revisiting the evidence linking Arctic amplification to extreme weather in midlatitudes, Geophys. Res. Lett., 40, 4734–4739,
doi:10.1002/grl.50880.
Barsugli, J. J., and D. S. Battisti (1998), The basic effects of atmosphere-ocean thermal coupling on midlatitude variability, J. Atmos. Sci., 55(4),
477–493.
Blunden, J., and D. S. Arndt (2014), State of the climate in 2013, Bull. Am. Meteorol. Soc., 95(7), S1–S257.
Bond, N. A., J. E. Overland, M. Spillane, and P. Stabeno (2003), Recent shifts in the state of the North Pacific, Geophys. Res. Lett., 30(23), 2183,
doi:10.1029/2003GL018597.
Bretherton, C. S., M. Widmann, V. P. Dymnikov, J. M. Wallace, and I. Blade (1999), The effective number of spatial degrees of freedom of a
time-varying field, J. Clim., 12, 1990–2009.
Cayan, D. R. (1992), Latent and sensible heat flux anomalies over the northern oceans: The connection to monthly atmospheric circulation,
J. Clim., 5, 354–369.
Ceballos, L. I., E. Di Lorenzo, C. D. Hoyos, N. Schneider, and B. Taguchi (2009), North Pacific Gyre Oscillation synchronizes climate fluctuations
in the eastern and western boundary systems, J. Clim., 22(19), 5163–5174, doi:10.1175/2009jcli2848.1.
Chhak, K. C., E. Di Lorenzo, N. Schneider, and P. F. Cummins (2009), Forcing of low-frequency ocean variability in the northeast Pacific, J. Clim.,
22(5), 1255–1276, doi:10.1175/2008jcli2639.1.
Compo, G. P., et al. (2011), The Twentieth Century Reanalysis Project, Q. J. R. Meteorol. Soc., 137(654), 1–28, doi:10.1002/qj.1776.
Cummins, P. F., and H. J. Freeland (2007), Variability of the North Pacific current and its bifurcation, Progr. Oceanogr., 75(2), 253–265,
doi:10.1016/j.pocean.2007.08.006.
Davis, R. E. (1976), Predictability of sea-surface temperature and sea-level pressure anomalies over the North Pacific Ocean, J. Phys.
Oceanogr., 6(3), 249–266.
Deser, C., and M. L. Blackmon (1995), On the relationship between tropical and North Pacific sea-surface temperature variations, J. Clim., 8(6),
1677–1680.

©2015. American Geophysical Union. All Rights Reserved.

8

 Geophysical Research Letters

10.1002/2015GL063083

Deser, C., A. Phillips, V. Bourdette, and H. Y. Teng (2012), Uncertainty in climate change projections: The role of internal variability, Clim. Dyn.,
38(3–4), 527–546, doi:10.1007/s00382-010-0977-x.
Di Lorenzo, E., et al. (2008), North Pacific Gyre Oscillation links ocean climate and ecosystem change, Geophys. Res. Lett., 35, L08607,
doi:10.1029/2007GL032838.
Ding, Q. H., J. M. Wallace, D. S. Battisti, E. J. Steig, A. J. E. Gallant, H. J. Kim, and L. Geng (2014), Tropical forcing of the recent rapid Arctic
warming in northeastern Canada and Greenland, Nature, 509(7499), 209–212, doi:10.1038/nature13260.
Eade, R., D. Smith, A. Scaife, E. Wallace, N. Dunstone, L. Hermanson, and N. Robinson (2014), Do seasonal-to-decadal climate predictions
underestimate the predictability of the real world?, Geophys. Res. Lett., 41, 5620–5628, doi:10.1002/2014GL061146.
Francis, J. A., and S. J. Vavrus (2012), Evidence linking Arctic amplification to extreme weather in mid-latitudes, Geophys. Res. Lett., 39, L06801,
doi:10.1029/2012GL051000.
Frankignoul, C., N. Sennechael, Y. O. Kwon, and M. A. Alexander (2011), Influence of the meridional shifts of the Kuroshio and the Oyashio
Extensions on the atmospheric circulation, J. Clim., 24(3), 762–777.
Horel, J. D., and J. M. Wallace (1981), Planetary-scale atmospheric phenomena associated with the Southern Oscillation, Mon. Weather Rev.,
109, 813–829.
Huber, M., and R. Knutti (2014), Natural variability, radiative forcing and climate response in the recent hiatus reconciled, Nat. Geosci., 7(9),
651–656, doi:10.1038/ngeo2228.
Kalnay, E., et al. (1996), The NCEP/NCAR 40-Year Reanalysis Project, Bull. Am. Meteorol. Soc., 77(March), 437–471.
Kim, B.-M., S.-W. Son, S.-K. Min, J.-H. Jeong, S.-J. Kim, X. Zhang, T. Shim, and J.-H. Yoon (2014), Weakening of the stratospheric polar vortex by
Arctic sea-ice loss, Nat. Commun., 5, doi:10.1038/ncomms5646.
Kosaka, Y., and S. P. Xie (2013), Recent global-warming hiatus tied to equatorial Pacific surface cooling, Nature, 501(7467), doi:10.1038/
nature12534.
Kushnir, Y., W. A. Robinson, I. Blade, N. M. J. Hall, S. Peng, and R. Sutton (2002), Atmospheric GCM response to extratropical SST anomalies:
Synthesis and evaluation, J. Clim., 15(16), 2233–2256.
Kwon, Y. O., M. A. Alexander, N. A. Bond, C. Frankignoul, H. Nakamura, B. Qiu, and L. Thompson (2010), Role of the Gulf Stream and Kuroshio-Oyashio
systems in large-scale atmosphere-ocean interaction: A review, J. Clim., 23(12), 3249–3281.
Lau, N. C., and M. J. Nath (1994), A modeling study of the relative roles of tropical and extratropical SST anomalies in the variability of the
global atmosphere-ocean system, J. Clim., 7(8), 1184–1207.
Lau, N. C., and M. J. Nath (1996), The role of the “atmospheric bridge” in linking tropical Pacific ENSO events to extratropical SST anomalies,
J. Clim., 9(9), 2036–2057.
Liu, Z. Y., and M. Alexander (2007), Atmospheric bridge, oceanic tunnel, and global climatic teleconnections, Rev. Geophys., 45, RG2005,
doi:10.1029/2005RG000172.
Mantua, N. J., and S. R. Hare (2002), The Pacific Decadal Oscillation, J. Oceanogr., 58(1), 35–44.
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. Am. Meteorol. Soc., 78(6), 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. Clim., 11(12), 3112–3127.
Neelin, J. D., D. S. Battisti, A. C. Hirst, F.-J. Fei, Y. Wakata, T. Yamagata, and S. E. Zebiak (1998), ENSO theory, J. Geophys. Res., 103(C7),
14,261–14,290, doi:10.1029/97JC03424.
Newlin, M. L., and M. C. Gregg (2014), Global oceans [in “State of the Climate in 2013”], Bull. Am. Meteorol. Soc., 95(7), S51–S78.
North, G., T. Bell, R. Cahalan, and F. Moeng (1982), Sampling errors in the estimation of empirical orthogonal functions, Mon. Weather Rev.,
110, 699–706.
Palmer, T. N., and D. L. T. Anderson (1994), The prospects for seasonal forecasting—A review paper, Q. J. R. Meteorol. Soc., 120(518), 755–793.
Qiu, B. (2000), Interannual variability of the Kuroshio Extension system and its impact on the wintertime SST field, J. Phys. Oceanogr., 30(6),
1486–1502.
Rasmusson, E. M., and J. M. Wallace (1983), Meteorological aspects of the El Niño/Southern Oscillation, Science, 222, 1195–1202.
Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan (2003), Global analyses of sea surface
temperature, sea ice, and night marine air temperature since the late nineteenth century, J. Geophys. Res., 108(D14), 4407, doi:10.1029/
2002JD002670.
Seager, R., and M. Hoerling (2014), Atmosphere and ocean origins of North American droughts, J. Clim., 27(12), 4581–4606.
Seager, R., Y. Kushnir, N. H. Naik, M. A. Cane, and J. Miller (2001), Wind-driven shifts in the latitude of the Kuroshio-Oyashio Extension and
generation of SST anomalies on decadal timescales, J. Clim., 14(2001), 4249–4265.
Seager, R., M. Hoerling, S. Schubert, H. Wang, B. Lyon, A. Kumar, J. Nakamura, and N. Henderson (2014), Causes and predictability of the
2011–14 California drought, Rep., 40 pp., doi:10.7289/V7258K7771F.
Smirnov, D., M. Newman, and M. A. Alexander (2014), Investigating the role of ocean-atmosphere coupling in the North Pacific Ocean,
J. Clim., 27(2), 592–606.
Smith, T. M., R. W. Reynolds, T. C. Peterson, and J. Lawrimore (2008), Improvements to NOAA’s historical merged land-ocean surface
temperature analysis (1880–2006), J. Clim., 21(10), 2283–2296.
Trenberth, K. E., and J. T. Fasullo (2013), An apparent hiatus in global warming?, Earth’s Future, 1(1), 19–32.
Vimont, D. J., D. S. Battisti, and A. C. Hirst (2001), Footprinting: A seasonal connection between the tropics and mid-latitudes, Geophys. Res.
Lett., 28(20), 3923–3926, doi:10.1029/2001GL013435.
Vimont, D. J., J. M. Wallace, and D. S. Battisti (2003), The seasonal footprinting mechanism in the Pacific: Implications for ENSO, J. Clim., 16(16),
2668–2675.
Vivier, F., K. A. Kelly, and L. Thompson (1999), Contributions of wind forcing, waves, and surface heating to sea surface height observations in
the Pacific Ocean, J. Geophys. Res., 104(C9), 20,767–20,788, doi:10.1029/1999JC900096.
Wang, F. Y., Z. Y. Liu, and M. Notaro (2013), Extracting the dominant SST modes impacting North America’s observed climate, J. Clim., 26(15),
5434–5452, doi:10.1175/jcli-d-12-00583.1.
Wang, S. Y., L. Hipps, R. R. Gillies, and J.-H. Yoon (2014), Probable causes of the abnormal ridge accompanying the 2013–2014 California
drought: ENSO precursor and anthropogenic warming footprint, Geophys. Res. Lett., 41, 3220–3226, doi:10.1002/2014GL059748.
Zhang, Y., J. M. Wallace, and D. S. Battisti (1997), ENSO-like interdecadal variability: 1900–93, J. Clim., 10(5), 1004–1020.

HARTMANN

©2015. American Geophysical Union. All Rights Reserved.

9