Southern hemisphere circulation shifts in a warming climate
Staten, P. W. , T. Reichler , and J. Lu
2. Center for Ocean‐Land‐Atmosphere Studies, Calverton, MD
The Ferrel cell vs. Hadley cell shift ratio
ratio of shifts
SST, 2.2
CO2, 1.6
O3, 4.8
U850
max
We ask:
1.4
1.2
2.2
• What is the effect of each forcing?
• How will the circulation change in the 21st century?
• How do various forced shifts relate to each other?
• How do forced shifts compare to interannual variability?
• What mechanisms are involved?
Ferrel
cell
center
DJF 2100‐2000 forced response
Hadley cell edge
-2
15
15
60 oS
30 oS
eq 90 oS
60 oS
30 oS
eq
2.4
5
o
‐5o
Ferrel
cell
center
The seasonality of shifts
Ferrel cell center shift
eq
JJAS
a
O3
CO2
SST
all
0
‐1
0
‐5o
5o
Shading represents a histogram
of shifts.
Hadley cell edge
Ferrel cell center
‐1
‐2
‐2
J F MAM J J A S OND J
b
30 oS
‐10 0
10 20 30 40 ‐10 0
10 20 30 40
U/cos(θ) and angular phase speed (ms‐1)
• During DJFM, weak westerlies at high latitudes limit
deposition of momentum except near the subtropical jet
core, keeping the Hadley and Ferrel cells more tightly
coupled in spite of increased phase speeds.
• While linear wave behavior seems clearly evident, the
contraction of the jet from both sides due to increased phase
speeds is not, suggesting other mechanisms may also be
work (see McLandress et al., 2011).
• During JJAS, increased phase speeds are associated with a
poleward shift in eddy activity. Strong high latitude
westerlies allow wave activity to extend far polewardof
Hadley cell.
• SST warming (not shown) produces more EMFC at low
latitudes and slow phase speeds, and less at high latitudes.
Perhaps this is why SST warming doesn't 'favor' the Ferrel
cell so strongly, even during JJAS, when the polar vortex
makes the Ferrel cell more 'sensitive'.
1
1
lat
1.3
U850
max
1.7
Hadley cell edge shift
DJFM
4xCO2 response
80 oS
DJF coupled variability
• Moderate O3 response + strong SST response = continued
poleward jet shift in AM2.1 (even during 2000‐2050).
• These results are likely model dependent, as studies using
CMAM (e.g. Kang et al., 2011; McLandress et al., 2011), and
other models (e.g. Polvani et al., 2010, Son et al., 2010)
suggest a tighter 'competition' between ozone recovery
and other forcings during the coming decades.
o
Ferrel cell center
Having examined forced shifts, we now examine year‐to‐year shifts in
the coupled CM2.1 control simulation.
90 oS
Speculating on mechanisms–
a flux convergence view
We calculate the eddy momentum flux convergence (EMFC)
after Chen & Held (2007), and discuss their proposed
mechanism whereby increased phase speeds are posited to
result in a poleward shift in wave activity.
FC/HC ratio due to coupled variability
100
1000
externally
forced
response
60 oS
-25
-5
5
height (hPa)
15
15
all
1o
-25
-5
CO2
internal
variability
SST O3 CO2
• Shift ratios are consistent across time scales in the coupled
simulation, and are larger for the summer.
• This pattern breaks down when oceanic variability is
omitted, but holds for the uncoupled forced responses.
• The FH/HC ratio varies by forcing.
• For a given forcing, shifts vary approximately linearly, and are
insensitive to the background climatology. Kang & Polvani (2010),
examining the interannual U850/HC slope, also find little dependence
on background climatology.
• Shifts vary by hemisphere and season (not shown).
100
-5
ab
coupled
variability
Each point represents the
response to a given forcing.
SST
1
timescale
(years)
o
Ratios are calculated as the slope
of the first eigenvector of the
covariance matrix.
1.6
1.4
2.0
DJF circulation changes
1000
10
1
‐3.5o
21 century circulation shifts
-25
FC/HC
2
shift
ratio
‐3.5o
st
-5
DJF JJA
10 2
DJF forced response
4
10 0
We investigate whether the Ferrel cell shift / Hadley cell shift ratio
(FC/HC) seen during the 21st century is dependent on the strength of
the forcing, type of forcing, or the background climatology.
• 3000‐year coupled CM2.1 control
• 3000‐year uncoupled AM2.1 control
• 500‐year time‐slice simulations with varied forcings:
• CO2: 143, 286, 380, 720, or 1120 ppm
• SSTs: from years 1870, 2000, 2050, or 2100
• O3: 0.4x reverse depletion, control, 1x or 2x depletion
FC/HC ratio overview
Here we examine the FC/HC ratio as a function of timescale
and season, investigating the role of ocean coupling.
FC/HC ratio due to external forcings
We examine:
O3
3. Dept. of Atmos., Oceanic, and Earth Sciences, George Mason University, Fairfax, VA
10 1
Separating forcings
using AM2 & CM2
[email protected]
10 2
1. Department of Atmospheric Sciences, University of Utah, Salt Lake City, UT
2,3
10 0
C36 Th54A
10
1
10 1
1
J F MAM J J A S OND J
• Poleward shifts due to SST warming consistently
overpower equatorward shifts due to O3 recovery.
• The Ferrel cell shifts more strongly than the Hadley cell
almost all year for all forcings.
• This strong SST effect is in contrast to Kang et al. (2011),
who show relatively little dependence of circulation shifts
on SSTs.
• The interannual FC/HC ratio is similar to the forced FC/HC slopes
above.
• Scatter makes it difficult to say whether interannual ratios most
closely match changes from any particular forcing.
• Our results are similar to and extend those by Kang & Polvani
(2010) for our model. For example, we calculate slopes for both
hemispheres and for DJF (not shown). We find correlations above
0.6 for the FC/HC ratio during both seasons, and over both
hemispheres.
This research is supported through TGLL, a NSF GK-12 program.
Conclusions
• In the GFDL model, the SST response dominates the O3
response, producing a poleward jet shifts during the 21st
century.
• The Ferrel cell generally shifts more than the Hadley cell.
• We find no definitive FC/HC shift ratio, even within one
model.
• The FC/HC shift ratio for different forcings, seasons, and
time‐scales may yield insights into underlying mechanisms.
• The Chen and Held (2007) phase speed mechanism may
explain some of the differences between season and forcing.