SPM
Summary for Policymakers
22
Figure SPM.8 | Maps of CMIP5 multi-model mean results for the scenarios RCP2.6 and RCP8.5 in 2081–2100 of (a) annual mean surface temperature change, (b) average percent change in annual mean precipitation, (c) Northern Hemisphere September sea ice extent, and (d) change in ocean surface pH. Changes in panels (a), (b) and (d) are shown relative to 1986–2005. The number of CMIP5 models used to calculate the multi-model mean is indicated in the upper right corner of each panel. For panels (a) and (b), hatching indicates regions where the multi-model mean is small compared to natural internal variability (i.e., less than one standard deviation of natural internal variability in 20-year means). Stippling indicates regions where the multi-model mean is large compared to natural internal variability (i.e., greater than two standard deviations of natural internal variability in 20-year means) and where at least 90% of models agree on the sign of change (see Box 12.1). In panel (c), the lines are the modelled means for 1986−2005; the filled areas are for the end of the century. The CMIP5 multi-model mean is given in white colour, the projected mean sea ice extent of a subset of models (number of models given in brackets) that most closely reproduce the climatological mean state and 1979 to 2012 trend of the Arctic sea ice extent is given in light blue colour. For further technical details see the Technical Summary Supplementary Material. {Figures 6.28, 12.11, 12.22, and 12.29; Figures TS.15, TS.16, TS.17, and TS.20}
−0.55 −0.5−0.6 −0.4 −0.35−0.45 −0.25 −0.2−0.3 −0.1 −0.05−0.15(pH unit)
109
−20 −10−30−50 −40 0 10 20 30 40 50
(b)
(c)
RCP 2.6 RCP 8.5
Change in average precipitation (1986 −2005 to 2081−2100)
Northern Hemisphere September sea ice extent (average 2081−2100)29 (3) 37 (5)
3932
(d) Change in ocean surface pH (1986 −2005 to 2081−2100)
(%)
(a) Change in average surface temperature (1986 −2005 to 2081−2100)3932
(°C)−0.5−1−2 −1.5 0 1 1.5 2 3 4 5 7 9 110.5
CMIP5 multi-model average 2081−2100
CMIP5 multi-modelaverage 1986 −2005
CMIP5 subset average 2081−2100
CMIP5 subsetaverage 1986 −2005
Emanuele Di Lorenzo PICES, October, 2019
ALASKA MARINE HEATWAVE 2019 HOT OFF THE PRESS
Dillon AmayaTongtong Xu
SPM
Summary for Policymakers
22
Figure SPM.8 | Maps of CMIP5 multi-model mean results for the scenarios RCP2.6 and RCP8.5 in 2081–2100 of (a) annual mean surface temperature change, (b) average percent change in annual mean precipitation, (c) Northern Hemisphere September sea ice extent, and (d) change in ocean surface pH. Changes in panels (a), (b) and (d) are shown relative to 1986–2005. The number of CMIP5 models used to calculate the multi-model mean is indicated in the upper right corner of each panel. For panels (a) and (b), hatching indicates regions where the multi-model mean is small compared to natural internal variability (i.e., less than one standard deviation of natural internal variability in 20-year means). Stippling indicates regions where the multi-model mean is large compared to natural internal variability (i.e., greater than two standard deviations of natural internal variability in 20-year means) and where at least 90% of models agree on the sign of change (see Box 12.1). In panel (c), the lines are the modelled means for 1986−2005; the filled areas are for the end of the century. The CMIP5 multi-model mean is given in white colour, the projected mean sea ice extent of a subset of models (number of models given in brackets) that most closely reproduce the climatological mean state and 1979 to 2012 trend of the Arctic sea ice extent is given in light blue colour. For further technical details see the Technical Summary Supplementary Material. {Figures 6.28, 12.11, 12.22, and 12.29; Figures TS.15, TS.16, TS.17, and TS.20}
−0.55 −0.5−0.6 −0.4 −0.35−0.45 −0.25 −0.2−0.3 −0.1 −0.05−0.15(pH unit)
109
−20 −10−30−50 −40 0 10 20 30 40 50
(b)
(c)
RCP 2.6 RCP 8.5
Change in average precipitation (1986 −2005 to 2081−2100)
Northern Hemisphere September sea ice extent (average 2081−2100)29 (3) 37 (5)
3932
(d) Change in ocean surface pH (1986 −2005 to 2081−2100)
(%)
(a) Change in average surface temperature (1986 −2005 to 2081−2100)3932
(°C)−0.5−1−2 −1.5 0 1 1.5 2 3 4 5 7 9 110.5
CMIP5 multi-model average 2081−2100
CMIP5 multi-modelaverage 1986 −2005
CMIP5 subset average 2081−2100
CMIP5 subsetaverage 1986 −2005
Emanuele Di Lorenzo PICES, October, 2019
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
ALASKA MARINE HEATWAVE 2019 HOT OFF THE PRESS
July-Aug-Sept 2019
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
July-Aug-Sept 2019
C
20192014/151957
Marine HeatWave Index
MHW Index
Sea Surface Temperature Anomalies
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
July-Aug-Sept 2019
C
20192014/151957
Marine HeatWave Index
MHW Index
Sea Surface Temperature Anomalies
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
Average of Extremes
C
20192014/151957
Marine HeatWave Index
MHW Index
Sea Surface Temperature Anomalies
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
Correlation
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
Correlation of MHW Index with SSTa
C
20192014/151957
Marine HeatWave Index
MHW Index
Sea Surface Temperature Anomalies
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
Correlation
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
Correlation of MHW Index with SSTa
C
20192014/151957
Marine HeatWave Index
MHW Index
Sea Surface Temperature Anomalies
What about the Atmospheric Circulation?
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-0.6 -0.4 -0.2 0 0.2 0.4 0.6
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
C
20192014/151957
Marine HeatWave Index
MHW Index
Sea Surface Pressure AnomaliesRegression of
MHW Index on SLPa
mlbar
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-0.6 -0.4 -0.2 0 0.2 0.4 0.6
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
CMarine HeatWave Index
MHW Index
Sea Surface Pressure Anomalies
High Pressure GOA System
1950 1960 1970 1980 1990 2000 2010 2020-4
-2
0
2
4
6
Regression of MHW Index on
SLPa
SLPI Index
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-0.6 -0.4 -0.2 0 0.2 0.4 0.6
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
C
Sea Surface Pressure Anomalies
High Pressure GOA System
1950 1960 1970 1980 1990 2000 2010 2020-4
-2
0
2
4
6
Regression of MHW Index on
SLPa
SLPI Index
dSSTa(t)dt = a ⋅SLPI (t)− SSTa(t)
tdissipation
AR-1 Model
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-0.6 -0.4 -0.2 0 0.2 0.4 0.6
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
Regression of MHW Index on
SLPa
C
Sea Surface Pressure Anomalies
R=0.5
High Pressure GOA System
SLP GOA Index (AR-1 Model)
dSSTa(t)dt = a ⋅SLPI (t)− SSTa(t)
tdissipation
1950 1960 1970 1980 1990 2000 2010 2020-4
-2
0
2
4
6
SLPI Index
AR-1 Model
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
C
20192014/151957
Marine HeatWave Index
MHW Index
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
C
20192014/151957
Marine HeatWave Index
MHW Index
SSTa Anomalies 2013-2019
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
C
20192014/151957
Marine HeatWave Index
MHW Index
SSTa Anomalies 2013-2019
NPGO-like
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
NPGO Index R=0.49 PDO Index R=0.36
CMarine HeatWave Index
MHW Index
SSTa Anomalies 2013-2019
NPGO-like
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
NPGO Index R=0.49 PDO Index R=0.36
CMarine HeatWave Index
MHW Index
SSTa Anomalies 2013-2019
NPGO-likeThe dynamics of Marine HeatWave are not
independent of the North Pacific climate modes
ARTICLESPUBLISHED ONLINE: 11 JULY 2016 | DOI: 10.1038/NCLIMATE3082
Multi-year persistence of the 2014/15
North Pacific marine heatwave
Emanuele Di Lorenzo1* and Nathan Mantua2
Between the winters of2013/14 and 2014/15 during the strong North American drought, th
e northeastPacific exp
erienced
the largest marine heatwaveever record
ed. Here we combine observations with an ensemble of climate model simulations
to show that teleconnections
between the North Pacific and the weak 2014/2015El Niño linked the atmospheric forcing
patterns of this even
t. These teleconnection dynamics from the extratropic
s to the tropics during winter 201
3/14, andthen
back to the extratropics during winter 201
4/15, are a key source of multi-year persistence
of the North Pacific atmosphere.
The corresponding ocean anomalies map onto known patterns of North Pacific decadal va
riability, specifically
the North
Pacific Gyre Oscillation(NPGO) in 2014 and the Pacific Decadal O
scillation (PDO) in 2015. A large ensemble of climate
model simulations predicts that
the wintervariance of
the NPGO-and PDO-like p
atterns increases und
er greenhouse forcing
,
consistentwith other stud
ies suggesting an increase in
the atmospheric extremes that lea
d to drought over North America.
During the fall of 2013a large warm temperature anomaly
developedin the upper ocea
n along the axis of theNorth
Pacific Current. As th
e anomaly spreadover a broa
d region
of the Gulf of Alaska (GOA) during the winter of 2013/14, it
reached a record-breaking amplitude wit
h sea surfacetemperature
anomalies (SSTa) exceedingthree standard deviations
(⇠3 �C)
(Fig. 1a and Supplementary Fig. 1, seeMethods for
a description
of the datasets and definitionof the SSTa indices). T
he onset
and growth of this unusual water
mass anomaly is attributed to
forcing associatedwith a persistent
atmospheric ridge over the
northeast Pacific
1 (Fig. 1b) that is conne
cted to the NorthPacific
Oscillation(NPO), a leading pattern of atmospheric variability
2 .
Extreme amplitude and persistencein the NPO pattern is also
implicated inthe record
drought conditions th
at a�ectedCalifornia
in the winterof 2013/14
3–5 and its expression is a known
precursor
of El Niñoconditions
6,7 . By the summer and fall of 2014
, the warm
anomalies reached the Pacific
coastal boundary of N
orth America,
and although the amplitude in t
he GOA and the northern Calif
ornia
Current System (CCS) were
reduced, record-high SSTa were
found
in the regions of south
ern and Baja California
(Fig. 1c). Inthe winter
of 2014/15,the SSTa ov
er the entire northeast
Pacific re-intensified,
exceedingagain the 3 �C threshold (Fig. 1e and Supplementary
Fig. 1). The record-br
eaking high-temperature and the multi-year
persistenceof this warm anomaly, here referred to as a marine
heatwave8 , have had unpreceden
ted impacts on multiple trophic
levels of the marine ecosy
stem and socio-economically important
fisheries. Associatedecosystem
impacts included low primary
productivity9 , 11 new
warm-water copepod species
to the northern
CaliforniaCurrent sh
elf/slope region
10 , a massive influx of dead or
starving Cassin’s aukle
ts (sea birds) onto Pac
ific Northwest beaches
from October through December 201411 , a large whale unusual
mortality event in the western GOA in 201512 , and a California
sea lion unusual mortality event in California
from 2013–201513 .
Severe, negative socio-econ
omic impacts resulted from the 2015
harmful algal bloom that extend
ed from southern Californiato
southeast Alaska, the la
rgest ever recorded
14 . Toxins produced by th
e
extreme harmful algal bloom contaminated shell
fish inWashington,
Oregon and California,prompting prolo
nged closures for valuable
shellfish fisheries. Although the socio-econ
omic consequences of
this climate event need to be fur
ther evaluated, it is po
ssible that the
northeast Pacific warm
anomaly of 2014–15 is the most ecologic
ally
and economically signif
icant marine heatwave on record.
Althoughprevious studies
1,3,15–17 have documented the onset
and nature of the atmospheric variabilitythat forced
the winter
2013/14 SSTa, thedynamics underlying
the persistenceand re-
intensification of the anom
aly in 2015 are still unclear.
The relative
role of ocean internal dy
namics versus direct atmospheric fo
rcing
in driving theexpression
of the 2015SSTa has n
ot been examined.
It is also unclear if the January
–February–March (JFM) 2014 and
JFM 2015 SSTa patterns (Fig. 1a and e) are dynamically linked,
and if they are, how?There is good evidence that atmospheric
teleconnections of tro
pical originplayed a key role in the winter
2013/14 sea-level pressure anomalies (SLPa)
4,15–17 (Fig. 1b),and
that the variance of this anomaly pattern may intensify under
greenhouseforcing
3,4 , hence leading to more extremes in ocean
temperature and western US precipit
ation. Thisraises the q
uestion
of whethertropical/ex
tratropicalteleconnec
tions werealso impor-
tant in driving theexceptiona
l SSTa in the winterof 2014/15.
Atmospheric forcing of the marine heatwave
To understand the rol
e of atmospheric forcing in driving the
strong
North Pacific warm anomalies, we begin by inspectingmaps of
the seasonal evolution of SSTa and SLPa between JFM 2014 and
JFM 2015 (Fig. 1). The patterns of the peak SSTa in JFM 2014
and 2015 show important spatial di�eren
ces. Whereas in 2014 the
core SSTa are centred
in the GOA (Fig. 1a) and exhibit a N
PGO-
like expression18 or Victoria Pattern
19 , in 2015 the largest warm
anomalies are further to the
east and extend along
the entire Pacific
North American coastal boundary, rese
mbling the expressionof
the PDO20 , also referred to as the ‘ARC’ patte
rn (Fig. 1e). These
di�erencesin SSTa patter
ns are mirrored by a change inthe SLPa
patterns, which exhibit a st
rong dipolesystem in JFM 2014, typic
al
of the NPO2 (Fig. 1b), a
nd a more pronounced single SLPa low
in 2015, resembling the express
ion of a deeperand southeastw
ard
extended Aleutian Low (Fig. 1f). To measure the strength of the
2014 and 2015 anomaly patterns wecompute the av
erage SSTain
© ƐƎƏƖɥMacmillan Publishers LimitedƦɥ/�13ɥ.$ɥ�/1(-%#1ɥ��
341#. All rights reservedƥ1 School of Earth and Atmospheric Sciences, Georgia Institute of Technology, Atlanta, Georgia 30332, USA. 2 Southwest Fisheries Science Center, National
Marine Fisheries Service, National Oceanographic and Atmospheric Administration, 110 Sha�er Road, Santa Cruz, California 95060, USA.
*e-mail: [email protected]
1042
NATURE CLIMATE CHANGE | VOL 6 | NOVEMBER 2016 | www.nature.com/natureclimatechange
Increasing Coupling Between NPGO and PDO Leads
to Prolonged Marine Heatwaves
in the Northeast PacificYoungji Joh1
and Emanuele Di Lorenzo1
1School of Earth and Atmospheric Sciences, Georgia Institute of Technology, Atlanta, GA, USA
Abstract The marine heatwave of 2014/2015 in the Northeast Pacific caused significant impacts on
marine ecosystems and fisheries. While several studies suggest that land and marine heatwaves may
intensify under climate change, less is known about the prolonged multiyear nature (~2 years) of the
Northeast Pacific events. Examination of reanalysis products and a 30-member climate model ensemble
confirms that prolonged multiyear marine heatwaves are linked to the dynamics of the two dominant modes
of winter sea surface temperature variability in the North Pacific, the Pacific Decadal Oscillation (PDO), and
the North Pacific Gyre Oscillation (NPGO). Specifically, we find a significant correlation between winter warm
NPGO anomalies and the following winter PDO arising from extratropical/tropical teleconnections. In the
model projections for 2100 under the RCP8.5 scenario, this NPGO/PDO 1 year lag correlation exhibits a
significant positive trend (~35%) that favors more prolonged multiyear warm events (>1°C) with larger
spatial coverage (~18%) and higher maximum amplitude (~0.5°C for events>2°C) over the Northeast Pacific.
Plain Language Summary Between the winters of 2014 and 2015 the Northeast Pacific
experienced the largest and longest marine heatwave ever recorded in the instrumental record. A
distinguishing feature of this event is themultiyear persistence of the ocean warm anomalies from one winter
to the other. By analyzing and comparing different reanalysis products and an ensemble of climate model
projections for 2100, we find that the observational trend for stronger winter to winter persistence of
anomalies in the Northeast Pacific is consistent with climate model projections under the RCP8.5 radiative
forcing scenario. We link this trend to an increase coupling between the two dominant modes of North
Pacific decadal variability.1. IntroductionThe 2013/2015 marine heatwave of the Northeast Pacific was characterized by the strongest ocean tempera-
ture extremes ever recorded in the North Pacific (Anderson et al., 2016; Baxter & Nigam, 2015; Bond et al.,
2015; Hartmann, 2015; Hobday et al., 2016; Peterson et al., 2016; Wang et al., 2014) and by an unusual persis-
tence that spanned the winters of 2013/2014 and 2014/2015 (Di Lorenzo &Mantua, 2016), culminating in one
of the strongest El Niño events of the twentieth century in the fall/winter of 2015/2016. The progression of
the event followed distinct spatial and temporal winter patterns in the ocean and atmosphere that
closely resemble the two dominant modes of variability of sea surface temperature and sea level pressure
anomalies (SSTa and SLPa). Specifically, the spatial structures of the January-February-March (JFM) SSTa in
2013/2014 and 2014/2015 are captured by the 2nd and 1st principal components of the North Pacific SSTa
(Di Lorenzo & Mantua, 2016) (Figure S1 in the supporting information). In the Northeast Pacific, these modes
are commonly referred to as the North Pacific Gyre Oscillation (NPGO) (Di Lorenzo et al., 2008) and the Pacific
Decadal Oscillation (PDO) (Mantua et al., 1997) (Figure S1). The similarity between the marine heatwave pat-
terns and the mode of Pacific decadal variability suggests that the statistics and persistence of these ocean
extremes are linked to the dynamics underlying the North Pacific modes.
Using historical reanalysis products and a climate model ensemble, this study provides a diagnostic of ocean
extremes statistics in past observations and in future model projections under the radiative forcing scenario
RCP8.5. The goal of this study is to (1) confirm the hypothesis that prolonged ocean extremes events follow
recurrent patterns with a transition from a winter NPGO-like pattern to PDO-like pattern in the following win-
ter and (2) examine how the coupling between these modes via tropical/extratropical teleconnections is
changing under a warmer climate favoring more prolonged winter to winter warm events.
JOH AND DI LORENZO
MARINE HEATWAVES IN NORTHEAST PACIFIC
11,663
PUBLICATIONSGeophysical Research Letters
RESEARCH LETTER10.1002/2017GL075930Special Section:
Midlatitude Marine Heatwaves:
Forcing and ImpactsKey Points:• Multiyear SST warm events in the
Northeast Pacific typically emerge as a
winter NPGO-like warm pattern and
transition to a PDO-like pattern in the
following winter• The coupling between winter NPGO
and the following winter PDO is a
robust climate teleconnection in both
observations and the CESM-LENS over
the period 1920-2100• A stronger NPGO-PDO coupling is
predicted under anthropogenic
forcing in the CESM-LENS and leads
to more prolonged and larger area
multiyear marine heatwavesSupporting Information:
• Supporting Information S1Correspondence to:Y. Joh,[email protected]
Citation:Joh, Y., & Di Lorenzo, E. (2017).
Increasing coupling between NPGO
and PDO leads to prolonged marine
heatwaves in the Northeast Pacific.
Geophysical Research Letters, 44,
11,663–11,671. https://doi.org/10.1002/
2017GL075930Received 4 OCT 2017Accepted 8 NOV 2017Accepted article online 13 NOV 2017
Published online 30 NOV 2017
©2017. American Geophysical Union.
All Rights Reserved.
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
C
Correlation of MHW Index with SSTa
CMarine HeatWave Index
MHW Index
NPGO Index R=0.49 PDO Index R=0.36
-2 -1.5 -1 -0.5 0 0.5 1 1.5 20
1
2
3
4
5
MHW Index PDF -NPGO > 1 STD
MHW Index PDF -NPGO < -1 STD
SST Anomaly
Occ
urre
nces
Amplitude of SST Trend
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
C
Correlation of MHW Index with SSTa
CMarine HeatWave Index
MHW Index
NPGO Index R=0.49 PDO Index R=0.36
-2 -1.5 -1 -0.5 0 0.5 1 1.5 20
1
2
3
4
5
MHW Index PDF -NPGO > 1 STD
MHW Index PDF -NPGO < -1 STD
SST Anomaly
Occ
urre
nces
Amplitude of SST Trend
Problem: small size statistics
Observational Record
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
C
Correlation of MHW Index with SSTa
CMarine HeatWave Index
MHW Index
NPGO Index R=0.49 PDO Index R=0.36
-2 -1.5 -1 -0.5 0 0.5 1 1.5 20
1
2
3
4
5
MHW Index PDF -NPGO > 1 STD
MHW Index PDF -NPGO < -1 STD
SST Anomaly
Occ
urre
nces
Amplitude of SST Trend
MPI-Grand Ensemble: 100 ensemble members
1850-2005 with historical radiative forcing
2006-2100 with RCP8.5
Observational Record
Problem: small size statistics
0 1 2 3-2 10
0.3
0.6
SST Anomaly (MHW Index)
Probability Distribution Function
Pre-Industrial 1850-1879
MPI Model 100 Member Ensemble
0 1 2 3-2 10
0.3
0.6
Present Day 1980-2019
SST Anomaly (MHW Index)
Probability Distribution Function
Pre-Industrial 1850-1879
Average Shift
MPI Model 100 Member Ensemble
0 1 2 3-2 10
0.3
0.6
-NPGO > 1 STD
-NPGO < -1 STD
SST Anomaly (MHW Index)
Probability Distribution Function
Average Shift
MPI Model 100 Member Ensemble
0 1 2 3-2 10
0.3
0.6
-NPGO > 1 STD
-NPGO < -1 STD
SST Anomaly (MHW Index)
Probability Distribution Function
Average Shift
MPI Model 100 Member Ensemble
BlobIndex
Annual
∆95th0.130.260.42
∆99th0.050.130.21
+NPGOPresentDay
-NPGO
0 1 2 3-2 10
0.3
0.6
-NPGO > 1 STD
-NPGO < -1 STD
SST Anomaly (MHW Index)
Probability Distribution Function
Average Shift
MPI Model 100 Member Ensemble
BlobIndex
Annual
∆95th0.130.260.42
∆99th0.050.130.21
+NPGOPresentDay
-NPGO
Changes in Extremes associated with trend is comparable to that of the phases of the decadal modes
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
July-Aug-Sept 2019
Sea Surface Temperature Anomalies
Question: Is the Blob going to continue this winter?
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
July-Aug-Sept 2019
Sea Surface Temperature Anomalies
Question: Is the Blob going to continue this winter?
Empirical Dynamical Model Prediction
dxdt
= Lx + ξ Linear Inverse Model
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
July-Aug-Sept 2019
Sea Surface Temperature Anomalies
Question: Is the Blob going to continue this winter?
Empirical Dynamical Model Prediction
dxdt
= Lx + ξ Linear Inverse Model
By solving the LIM system, we obtain
x̂(t + τ) = exp(Lτ)x(t) = G(τ)x(t)
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
July-Aug-Sept 2019
Sea Surface Temperature AnomaliesEmpirical Dynamical Model Prediction
dxdt
= Lx + ξ Linear Inverse Model
By solving the LIM system, we obtain
x̂(t + τ) = exp(Lτ)x(t) = G(τ)x(t)
As data consist of SSTA and SLPA, our model system is
[ ̂s(t + τ)p̂(t + τ)] = G(τ = 6months)[s(t)
p(t)] SSTASLPA
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
July-Aug-Sept 2019
Sea Surface Temperature AnomaliesEmpirical Dynamical Model Prediction
dxdt
= Lx + ξ Linear Inverse Model
By solving the LIM system, we obtain
x̂(t + τ) = exp(Lτ)x(t) = G(τ)x(t)
[ ̂s(t + τ)p̂(t + τ)] = G(τ = 6months)[s(t)
p(t)]As data consist of SSTA and SLPA, our model system is
SSTASLPA
Initialize September
Forecast March
Jan-Feb-March
6 Months Prediction
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
July-Aug-Sept 2019
Sea Surface Temperature AnomaliesEmpirical Dynamical Model Prediction
Initialize September
Forecast March
Jan-Feb-March
6 Months Prediction
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
July-Aug-Sept 2019
Sea Surface Temperature AnomaliesEmpirical Dynamical Model Prediction
Initialize September
Forecast March
Jan-Feb-March
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
Marine HeatWave Index
MHW Index
Empirical Dynamical Model Prediction
Initialize September
Forecast March
Jan-Feb-March
6 Months Prediction
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
July-Aug-Sept 2019
Sea Surface Temperature AnomaliesEmpirical Dynamical Model Prediction
Initialize September
Forecast March
Jan-Feb-March
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
Marine HeatWave Index
MHW Index1950 1960 1970 1980 1990 2000 2010 2020
-3
-2
-1
0
1
2
3
4
2015 Blob
Winter Average Jan-Feb-March
Empirical Dynamical Model Prediction
Initialize September
Forecast March
Jan-Feb-March
6 Months Prediction
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
July-Aug-Sept 2019
Sea Surface Temperature Anomalies
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
Marine HeatWave Index
MHW Index1950 1960 1970 1980 1990 2000 2010 2020
-3
-2
-1
0
1
2
3
4
2015 Blob
2015 Prediction
Empirical Dynamical Model Prediction
Initialize September
Forecast March
Jan-Feb-March
6 Months Prediction
Winter Average Jan-Feb-March
Skill R=0.5
Cross-Validation
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
July-Aug-Sept 2019
Sea Surface Temperature Anomalies
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
Marine HeatWave Index
MHW Index1950 1960 1970 1980 1990 2000 2010 2020
-3
-2
-1
0
1
2
3
4
Empirical Dynamical Model Prediction
Initialize September
Forecast March
Jan-Feb-March
6 Months Prediction
Winter Average Jan-Feb-March
Skill R=0.5
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Winter 2015 ObservedWinter 2015 Prediction
Sea Surface Temperature Anomalies
2015 Blob
2015 Prediction
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
July-Aug-Sept 2019
Sea Surface Temperature Anomalies
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
Marine HeatWave Index
MHW Index1950 1960 1970 1980 1990 2000 2010 2020
-3
-2
-1
0
1
2
3
4
2015 Blob
2015 Prediction
Empirical Dynamical Model Prediction
Initialize September
Forecast March
Jan-Feb-March
6 Months Prediction
Winter Average Jan-Feb-March
Skill R=0.5 ?
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
C
July-Aug-Sept 2019
Sea Surface Temperature AnomaliesEmpirical Dynamical Model Prediction
Initialize September
Forecast March
Jan-Feb-March
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
Marine HeatWave Index
MHW Index1950 1960 1970 1980 1990 2000 2010 2020
-3
-2
-1
0
1
2
3
4
2015 Blob
2015 Prediction
Empirical Dynamical Model Prediction
Initialize September
Forecast March
Jan-Feb-March
6 Months Prediction
Winter Average Jan-Feb-March
Skill R=0.52020 Prediction
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2degree C
WINTER
2020
Sea Surface Temperature Anomalies
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
Marine HeatWave Index
Observed 6m Prediction
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
2015 Blob
2015 Prediction
Prediction R=0.5
2020 Prediction
PredictionNorth
America
Asia Alaska
C
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2degree C
SUMMER
2019
Sea Surface Temperature Anomalies
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
Marine HeatWave Index
Observed 6m Prediction
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
2015 Blob
2015 Prediction
Prediction R=0.5
2020 Prediction
North America
Asia Alaska
C
-220 -200 -180 -160 -140 -120 -100
0
20
40
60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2degree C
WINTER
2020
Sea Surface Temperature Anomalies
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
Marine HeatWave Index
Observed 6m Prediction
1950 1960 1970 1980 1990 2000 2010 2020-3
-2
-1
0
1
2
3
4
2015 Blob
2015 Prediction
Prediction R=0.5
2020 Prediction
PredictionNorth
America
Asia Alaska
C