North Atlantic Heat Content and Transport Trends based on Ocean Analyses

1

• North Atlantic Heat Content and Transport Trends
  based on Ocean Analyses
• Presented at ECCO Meeting, MIT, February, 2006
• Matt Harrison, Ants Leetmaa, Tony Rosati, Tom
  Delworth
• NOAA/GFDL

2
Outline

• Trends from 1993-2004.
• Heat content
• Attribution to surface heat flux and ocean
  transport changes using ECCO 1 degree analysis.
• North Atlantic Overturning.

3
Change in 0/750m heat content1993 to 2004
(Willis)
4
1993-2004 heating rate based on XBT plus
Altimeter (Watts/m2) assuming no change in heat
content below 750m. (data courtesy of J. Willis)
5
1993-2004 heating rate from ECCO 1 degree run.
Left, 0-750m Right, total depth (Watts/m2).
4.1 Watts/m2
6
1990-2001 heating rate from 1/2 degree SODA
analysis (Watts/m2)
7
1993-2004 heating rate from GFDL 3D-Var analysis
(Watts/m2)
8
Heat Content Summary
9
Comments about heat content trends

• Ocean analyses vary from 3-6 Watts/m2 trend in
  North Atlantic ocean heat content during 1990's.
  Estimates based on altimetry and in-situ data
  (Willis) 1.4 Watts/m2 trend.
• Qualitative agreement with enhanced warming
  poleward of 50N.
• How does this compare with independent data?
• Can we attribute these changes to forcings, i.e.
  change in surface heat input versus changes in
  ocean circulation?

10
ECCO mass transport streamfunction trend
(Sv/decade)
Subpolar Gyre Transport
11
(No Transcript)
12
(No Transcript)
13

• EOF1 of SSH (left, blue) along with along-isobath
  current observations at 200m (red) and 1000m
  (green) along Labrador Coast
• Velocity fluctuations are coherent to gt 2000m

14
ECCO Surface Heat Flux Trend
15
(No Transcript)
16
(No Transcript)
17
North Atlantic Overturning

• Bryden et al, 2005 infer from 6 transects taken
  between 1957 and 2004 that the North Atlantic
  overturning has been slowing over this time
  period.

18
ECCO Atlantic Mass Overturning Streamfunction at
26N
19
GFDL Unconstrained Atlantic Mass Overturning
Streamfunction at 26N starting from Levitus
20
Attribution

• ECCO shows a slowdown in subpolar gyre strength,
  consistent with warming at high latitudes.
• Associated with reduction in surface cooling in
  this region.
• This is consistent with trends in altimetry and
  current moorings.
• Slowdown in overturning circulation at 26N is
  consistent with Bryden conjecture but this needs
  to be examined more closely.

21
(No Transcript)
22
Challenges for ocean state estimation

• Separating real physical shifts from regional
  numerical drift ... (Stammer et al, 2000)
• Better physical models sea ice, river runoff,
  narrow passages (e.g. Gibralter, Bering Sea),
  resolution, numerics, etc...
• Improved assimilation methodologies.
• Including ARGO salinity in the 21st Century.
• Attribution.

23

• Lavender et al, 2005 used subsurface drifters
  deployed during WOCE to study the mean
  circulation (ARGO precursor).
• Hydrographic-based estimates of circulation
  suffer from uncertain level of no motion.
• 578 float years of data from Nov 1994 July,
  ..2002. 70 of data from 1997-1999.
• Depths of 400m (39 floats), 700m (123), 1400m
  (49).
• Eddy sampling?

24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31

• EOF1 of SSH (11 variance)
• 4-9 cm/decade in subpolar gyre

32
(No Transcript)
33
(No Transcript)