signature*of*the*atlan0c*meridional ... · signature*of*the*atlan0c*meridional*...
TRANSCRIPT
Signature of the Atlan0c Meridional Overturning Circula0on in the North
Atlan0c Dynamic Sea Level
Jianjun Yin Department of Geosciences
University of Arizona
2013 US AMOC/UK RAPID Mee3ng
Paul Goddard (UA), Stephen Griffies (GFDL) and Ronald Stouffer (GFDL)
Outline
• Mean dynamic sea level (DSL) in the North AtlanDc and along the U.S. East Coast
• Sea level rise (SLR) observaDons -‐ Satellite alDmetry -‐ Tide gauges
• GFDL ESM2M simulaDons and projecDons -‐ Two modes of the DSL variability and change -‐ AMOC-‐DSL relaDonship -‐ 21st century projecDons of the DSL along the U.S. East Coast
• Conclusions
Terminology • Dynamic sea level (η): the sea surface
deviaDon from the geoid (z=0 in climate
models).
barotropic
baroclinic
H z
fdzzdHgg Vk ×−ʹ′∇−=∇ ∫ ∫
−
0 0
0
ρρ
η
• Steric sea level (h): sea level gradient
induced by of ocean density variaDon
C W
∫−
−−=0
00
]),,([1),,(Z
dzPSTtyxh ρρρ
• Thermosteric sea level: temperature effect
• Halosteric sea level: salinity effect
∫−
−−=0
000
]),,([1),,(Z
thermo dzPSTtyxh ρρρ
∫−
−−=0
000
]),,([1),,(Z
halo dzPSTtyxh ρρρ
Mean DSL in the North AtlanDc
Subpolar water mass m
JPL AVISO 1992-‐2010
PHC T, S
climatology
relaDve to 1500 m
Subtropical water mass Cape Ha\eras
Dynamic sea level Steric sea level
A
B
Offshore and Alongshore Gradient
Lines: Mean DSL along the East Coast of U.S. Rectangles: Mean DSL difference across the Gulf Stream (∆η)
• Cross-‐current DSL difference: Cape Ha\eras Upstream: 0.2-‐0.5 m Downstream: 0.9-‐1.1 m • Alongshore gradient -‐ 0.6 m drop from Florida to Maine -‐ Strong gradient north of Cape Ha\eras • Steric effect associated with
distribuDon of different water masses
Temperature and Salinity Effect
• PHC, upper 1500 m • The DSL gradient is mainly set up by the temperature variaDon • Salinity effect parDally compensates
Thermosteric Halosteric
m
Outline
• Mean dynamic sea level (DSL) in the North AtlanDc and along the U.S. East Coast
• Sea level rise (SLR) observa0ons -‐ Satellite alDmetry -‐ Tide gauges
• GFDL ESM2M simulaDons and projecDons -‐ Two modes of the DSL variability and change -‐ AMOC-‐DSL relaDonship -‐ 21st century projecDons of the DSL along the U.S. East Coast
• Conclusions
Decadal Trend of the North AtlanDc DSL
mm / year
mm / year
1993-‐2002
2003-‐2012
Covarying dipole between the Gulf Stream and subpolar gyre (Hakkinen and Rhines, 2004; Zhang, 2008; Lorbacher et al., 2010)
Contrast across the Gulf Stream and Cape Ha\eras North – high South – low (Yin et al., 2009; Sallenger et al., 2012; Ezer et al., 2013)
(CSIRO alDmetry data)
Southeast
Mid-‐AtlanDc
New England and Canada
U.S. East Coast SLR Regimes
Coastal SLR Pa\erns
Long-‐term annual Dde gauge data (PSMSL) Key West – Southeast Norfolk – Mid-‐AtlanDc Boston – New England
2009
U.S. East Coast SLR PaQerns 1950-‐2012 Middle-‐High SLR faster in Mid-‐AtlanDc SLR rate decreased to the north and south 1993-‐2012 North-‐High South-‐Low SLR faster north of Cape Ha\eras SLR slower south of Cape Ha\eras GPS correc3on: Woppelmann et al. 2009 GIA correc3on: Pel3er 2004
AMOC vs SLR
Correla0on Annual mean: -‐0.8 Monthly data: -‐0.5 for 2 month lag
AMOC 3me series from McCarthy et al., 2012
PSMSL 3de gauge data at Boston
Anomalous 2009
Long-‐term annual Dde gauge data
SLR in a parDcular year:
New England and Canada
Mid-‐AtlanDc
Southeast
2)1()1()( −−+
=tSLtSLtSLR
Outline
• Mean dynamic sea level (DSL) in the North AtlanDc and along the U.S. East Coast
• Sea level rise (SLR) observaDons -‐ Satellite alDmetry -‐ Tide gauges
• GFDL ESM2M simula0ons and projec0ons -‐ Two modes of the DSL variability and change -‐ AMOC-‐DSL relaDonship -‐ 21st century projecDons of the DSL along the U.S. East Coast
• Conclusions
GFDL ESM2M • Delworth et al., 2006; Dunne et al., 2012
• Atmosphere: 2.5ox2o, 24 levels; Ocean: 1ox(1o-‐1/3o), 50 levels
• Free ocean surface and real freshwater flux
• Explicit simulaDon of global carbon cycle
• CMIP5 runs: historical (1861-‐2005); RCP projecDon runs (2005-‐2100)
• Dynamic sea level rise analysis
t)y,(x,hΔgρ
t)y,(x,Δpt)y,(x,0
b ʹ′+=Δη
Mass redistribuDon Local steric
ΔhΔht)y,(x,hΔ
ΔhdAA1(t)Δh
Δρdzρ1t)y,Δh(x,
global
η
H0
−=ʹ′
=
−=
∫
∫−
(Global steric)
(Local steric)
20th Century Mode (DSL)
21st Century Mode (DSL)
(m)
20th Century Mode (obs) EOF1 alDmetry (1993-‐2012)
EOF1 Ocean Heat Content (1955-‐2012) Trend of OHC (1955-‐2012)
Trend of wind stress and Sverdrup streamfuncDon (1948-‐2007)
CSIRO data
NOAA NCDC upper 700 m data NOAA NCDC data
Sv / yr
(Pa/yr) CORE forcing data
1018 J / yr
21st Century Mode Dynamic
Local Steric Bo\om Pressure
Buoyancy Flux and Wind Stress
m
m
m
10-‐8 m
2 s -‐3
-‐ GFDL ESM2M -‐ RCP8.5 -‐ 2091-‐2100 minus 1986-‐2005
Mechanism
Buoyancy flux: α: thermal expansion coefficient β: saline contracDon coefficient QHF: net air-‐sea heat flux cp: specific heat for seawater E: evaporaDon P: precipitaDon R: runoff
)](/)[/( 00 RPEScQgB pHF −−−= βραρ
AMOC vs DSL
• High correlaDon between the AMOC weakening and dynamic SLR on the U.S. East Coast • Dynamic SLR superimposed on global mean SLR, leading to high vulnerability of the U.S. East
Coast to future SLR
Ø Elevated storm surges, beach erosion,
inundaDon of low-‐lying area, damages to
coastal infrastructure and ecosystems
Impact
Conclusions • The SLR along the U.S. East Coast switched from a middle-‐high pa\ern
(faster in Mid-‐AtlanDc) during the 20th century to a north-‐high south-‐low pa\ern (separaDng at Cape Ha\eras) during the past decades.
• GFDL ESM2M suggests two disDnct modes of the DSL variability and change in the 20th and 21st century in the North AtlanDc.
• The middle-‐high pa\ern of coastal sea level rise during the 20th century was induced by northward ship of the Gulf Stream.
• The north-‐high south-‐low pa\ern during the 21st century is mainly caused by the significant decline of the cross-‐Gulf-‐Stream density contrast, i.e., the baroclinic process associated with the AMOC weakening.
• The northeast coast of the U.S. is parDcularly vulnerable to future sea level rise and storm surge.