Transcript Slide 1

Extreme sea level drops in the western tropical Pacific—
Causes, coastal impacts, and future projections
Matthew J. Widlansky
International Pacific Research Center
In collaboration with:
Axel Timmermann (IPRC)
Mark Merrifield (JIMAR, UHSLC)
Shayne McGregor (UNSW Sydney)
Malte Stuecker (UH Met. Dept)
Wenju Cai (CSIRO Australia)
Aerial view of Olosega Village,
courtesy
National Park of American Samoa
During strong El Niño events, sea level
drops around some tropical western Pacific
islands by up to a foot (30 cm)
Samoans call these events ‘taimasa’
(pronounced [kai’ ma’sa])
in reference to the foul smelling tide
caused by coral die-offs
Future extreme low sea level events may
become more frequent
Snapshot of a key
climate feature
South Pacific
Convergence Zone
(SPCZ)
Image from 8 February 2012
MTSAT-2 visible channel, Digital Typhoon, National Institute of Informatics
Recent literature on the SPCZ
(Widlansky et al., 2013 Nature Climate Change)
Underlying SST gradients influence the SPCZ
Observed rainfall and SST climatology during DJF
28°C
(mm day-1)
26°C
NOAA SST
GPCP rainfall
• Tropical SPCZ adjacent the meridional SST gradient (equatorial)
(e.g., Lindzen and Nigam, 1987 J. Atmos. Sci.)
• Subtropical SPCZ transects the meridional SST gradient (mid-latitudes) and is west
of maximum zonal SST gradient
(2011 Climate Dynamics)
Interannual variability of the SPCZ
Observed rainfall and SST climatology during DJF
(mm day-1)
Extreme
El Niño
28°C
26°C
El Niño
La Niña
NOAA SST
GPCP rainfall
(2011 Climate Dynamics)
Extreme zonally oriented SPCZ event:
4 January 1998
GMS-5
IR water vapor
6.70-7.16 μm
Susan
(125 kts)
Katrina
Some islands
experienced
droughts while
others more
tropical cyclones
Ron
(Tonga: 67%
damaged)
(28 days)
In Samoa,
prolonged low sea
levels exposed
shallow reefs
Tide gauge observations (tropical western Pacific)
UH Sea Level Center data
Interhemispheric sea level seesaw (r = 0.60) at lag 6 months
Very low sea levels, or ‘taimasa’, affect South Pacific islands mostly
during strong El Niño events (Widlansky et al., in review)
Shallow reef response to sea level variability
Normal conditions
El Niño Taimasa
d>h
d<h
Top portions of coral heads die off, creating what are known as microatolls
(e.g., Woodroffe and McLean, 1990 Nature)
Flat top Porites
coral
photo courtesy
National Park of
American Samoa
What causes extreme sea level drops?
Wind-stress variability associated with ENSO
Equatorial Pacific
(10°S-10°N, 100°E-60°W)
wind stress anomaly
(McGregor et al., 2012 J. Climate)
26%
15%
• 3 highest peaks of PC1 classified strong El Niño events, matching lowest
sea levels in the Southwest Pacific
• PC2 abruptly switches from negative to positive, especially after 1982/83
and 1997/98 El Niño peaks
Regressions: Sea surface height and wind stress
Zonal sea level gradient
Canonical sea level response to El Niño
Meridional sea level gradient
Most pronounced during strong El Niño
(e.g., Wyrtki, 1984 JGR)
(e.g., Alory and Delcroix, 2002 JGR)
Equatorially symmetric wind stress
pattern associated with ENSO
Southward shifted westerly wind
anomaly east of Dateline
(Stuecker et al., 2013 Nature Geoscience)
(McGregor et al., 2012 J. Climate)
Shading:Wind
Vectors:
Sea surface
stress (ERA
height
interim)
(ECMWF ORAs4)
Vectors:
Contours:
Tide
Gauge
Wind
Wind-stress
Stations:
stress (ERA
(UHSLC)
curl interim)
(negative, SH cyclonic)
Blue contours: Negative wind-stress curl (SH cyclonic)
Regressions: Near-surface current anomalies reverse
Zonal sea level gradient
Meridional sea level gradient
Strengthened Equatorial Counter Current, Current anomalies reverse, sea levels
drainage of West Pacific Warm Pool
return to normal in northwestern Pacific
(e.g., Wyrtki, 1984 JGR)
Sea levels remain depressed south of
5°N; i.e., the meridional seesaw.
(Alory and Delcroix, 2002 JGR)
Shading: Sea surface height (ECMWF ORAs4)
Vectors: Near-surface current, 5–56 m average (ORAs4)
Regressions: SPCZ collapses equatorward
Zonal SPCZ events are associated with prolonged
extreme sea level drops
(PC1 & PC2 > 0)
Shading: Rainfall (GPCP)
Blue contours: Pacific rainbands enclosed by 5 mm d-1
rainfall annual climatology
Asymmetric western Pacific sea level response
r = 0.74
at lag 3 months
Combination-mode: ENSO (fE) + Annual Cycle (1)
Near-annual combination tones appear in PC2 of surface winds
(Stuecker et al., 2013 Nature Geoscience)
& western Pacific sea level gradient
2–7 years
~15
months
~9
months
Shallow-water model (1.5-layer) hindcast simulations
‘Zonal seesaw’ of tropical Pacific
thermocline depth and sea levels
associated with ENSO
‘Meridional seesaw’ characterized
by 9 and 15 month spectral energy
(Wang, Wu, & Lukas, 1999 J. Meteorol. Soc. Jpn.)
Combination-mode and prolonged sea level drops
Southwest Pacific
PC2 experiment forced with the southward westerly wind shift
—essentially nonlinear interaction between annual cycle & ENSO—
sufficient to prolong below-normal sea levels
PC2 correlated with central Pacific sea level
Correlation with observed sea surface height
Central Pacific
r = 0.73
at lag 1 month
Key Points
1) Extremely low sea levels—capable of damaging shallow coral
reefs—persist long after termination of strong El Niño events
in the tropical southwestern and central Pacific.
2) Sea level drops are related to interaction of El Niño with the
forced annual cycle and associated seasonal development of
the South Pacific Convergence Zone.
3) Hindcast experiments suggest potential predictability of
future extreme sea level drops once El Niño has developed to
a certain intensity threshold.
How will strong El Niño events and
extreme sea level drops respond to
climate change?
Recent work by Cai & coauthors…
(Nature 2012)
Observed climatology (1981–2005)
Shading: Rainfall (GPCP)
Blue contours: Rainfall > 5 mm d-1
Green contours: Warm pool > 27.5°C
Red line: Zonal SPCZ position
CMIP5 projection (2074–2098 minus 1981–2005)
RCP8.5 W m-2 (31 models)
Blue contours: CTRL Rainfall > 5 mm d-1
Green contours: CTRL Warm pool > 27.5°C
Shading: Warming Vectors: Surface wind change
Defining a “zonal SPCZ event”: PC1 > 1 and PC2 > 0
Second principal component
GPCP rainfall
Zonal SPCZ
La Niña
Neutral
Moderate El Niño
First principal component
CMIP5 projections
Considering only models able to simulate nonlinear
behavior of the SPCZ (12 out of 31 models)
Climate
Change?
Number of zonal SPCZ events increases from Control
to Climate Change period
Meridional SST gradient & zonal SPCZ events
Observed rainfall and SST climatology during DJF
(mm day-1)
Box 1
28°C
26°C
Box 2
~ GPCP rainfall
~ NOAA SST
1997/98
El Niño
= [Box 1 SST – Box 2 SST]
Smaller future meridional SST gradient
Maximum equatorial warming is a robust response
to greenhouse warming
(e.g., Xie et al., 2010 J. Climate)
Box 1
Box 2
SST trend pattern (departure from tropical mean)
21st century projection (RCP 4.5 W m-2, 20 models)
Pacific island communities experience extreme weather
–droughts or floods, tropical cyclones, & sea level drops–
during zonal SPCZ events
Southwest Pacific
Simulated sea surface heights
Southwest Pacific
21st century projection?
RCP 4.5
*very preliminary
(1 model/1 run)
MIROC5 ocean-atmosphere GCM
Should future increased frequency of El Niño Taimasa occur,
a higher likelihood of prolonged low sea level events is
perceivable
Plans for further study:
Simple shallow-water ocean model
1) Hindcast experiments to confirm seasonal
predictability of El Niño Taimasa events
Use a sophisticated coupled ocean-atmosphere
model (e.g., NCAR CESM)
Future simulation from one model
2) Assess changing frequency of El Niño Taimasa
Examine ensemble of CMIP5 future climate
change experiments
3) Biogeographic characterization of near-shore reef
and community relevance
• Compile bathymetric data
• Partner with coral experts to determine growth
behaviors in response to sea level variability
Sketch outline of a reef flat
Last idea: Simulate shallow reef growth & decay
Normal conditions
El Niño Taimasa
(d > h) = 1
(d < h) = 1
Parametric coral model
Height
change
Growth
Decay or erosion
Coral growth constraints:
1) Water temperature below critical temperature (Tc – T)… warming
Non-branching coral
2) Aragonite saturation state (arag)… ocean acidification
(Porites)
3) Water depth above coral (d – h)… tides, El Niño, & sea level rise
Species specific parameters:
1) Growth rate constant (0)
2) Height dependent growth function (f)
3) Decay rate of exposed coral ()
Collaborate with coral experts to simulate future reef vulnerability
caused by climate extremes & communicate “Taimasa threat index”
Matthew Widlansky
[email protected]
http://iprc.soest.hawaii.edu/users/mwidlans/
Thank you
Increased number of zonal SPCZ events
Flux adjusted perturbed physics experiments with HadCM3 model
(12 out of 17 experiments considered)
1
Greenhouse warming is likely to cause:
1) More summers with small meridional SST gradients
2) Increased frequency of zonal SPCZ events
2