Dynamics of intraseasonal oscillations: the role of surface energy fluxes

1
Dynamics of intraseasonal oscillations the role
of surface energy fluxes

• Adam Sobel

2
Collaborators

• Hezi Gildor, Weizmann Institute
• Eric Maloney, Oregon State U.
• Gilles Bellon, Columbia U. (CAOS Alumnus)

3
Intraseasonal OLR variance
Northern Summer
Southern Summer
Lawrence and Webster 2002
4
Northern summer intraseasonal OLR variance tends
to avoid land
5
Southern summer intraseasonal OLR variance tends
to avoid land
6
Emanuel (87) and Neelin et al (87) proposed that
the MJO is driven by wind-induced surface flux
perturbations
Enhanced sfc flux
Mean flow
Perturbation flow
Wave propagation
7
This idea has been somewhat abandoned because the
real MJO does not look quite like the original
WISHE theory
Observed cloudiness and wind from TOGA COARE
Chen, Houze and Mapes 1996
8
But the real MJO does have significant net
surface heat flux variations, roughly in phase
with convection
9
Over land, there can be no significant net flux
variations on intraseasonal time scales - so if
net flux were important to ISO, the observed
variance maps should look as they do!
10
The simplest intraseasonal variability is seen in
a local analysis (Waliser 1996)
Month -1
SST
Month 0
Month 1

Time-varying composites of hot spots -
SSTgt29.5C for a period gt 1 month
Highly reflective cloudiness
11
This has the appearance of a local
recharge-discharge oscillation

• Clear skies, no deep convection,
• light winds, weak surface fluxes, high
• insolation -gt SST warms, lower atmos.
• humidity increases, troposphere becomes
• more unstable to deep convection
• Deep convection breaks out, clouds
• block sun, high winds strong sfc fluxes,
• SST decreases, lower troposphere dries,
• troposphere becomes more stable
• Deep convection stops, cycle starts
• over
• Can be driven by ISO, but doesnt
• necessarily have to be

Stephens et al. 2004
12
We can make a very simple model that has such
a recharge-discharge oscillation (Sobel and
Gildor 2003)
with
Simple Betts-Miller convection
Linear cloud-radiative feedback, SW and LW cancel
at TOA
Sfc wind constant for starters (will relax this)
13
The growth rate in this model is sensitive to
parameters, period isnt - it is robustly
intraseasonal
frequency
growth rate
0.1 d-1 60 d period
r (cloud-rad/sfc flux feedback)
tc (convective time scale)
mixed layer depth
14
Nonlinear solution in unstable regime looks much
like observed hot spot evolution
precipitation
evaporation (perturbed only due to sfc.
humidity difference)
atmospheric rad. cooling
surface shortwave ( wind induced evaporation)
SST
15
A more realistic system may be one that is stable
(in a single column) but forced by a traveling
ISO disturbance. Amplitude in this system depends
non-monotonically on mixed layer depth.
Amplitude as function of mld
Linear calculation forced in atmospheric
temperature equation
16
At least one GCM behaves similarly (Maloney and
Sobel 2004) - MJO amplitude vs. mixed layer depth
SST
Precip
Simple model (amplitude is max-min)
GCM (amplitude is std. dev.of filtered data)
Mixed layer depth -gt
(these simple model experiments are nonlinear,
and forced by oscillations in surface wind speed)
17
Wet land is like a mixed layer of zero depth
(swamp). Thus if MJO is dependent on surface
energy fluxes (turbulent, radiative, or both) it
should weaken over land as observed.
18
However, the GCM result is very dependent on
convective scheme.
sfc LH flux (MS04)
Precip (Maloney 2002)
Precip (Maloney Sobel 2004)
19
What about the Northern Summer ISO - monsoon
active and break periods?
latitude
Time
Sikka and Gadgil 1980
20
There is both northward and eastward propagation
Wang et al. 2006
21
The northward part appears to be somewhat
independent, justifying axisymmetric models
Nanjundiah et al. 1992
22
A recent model focuses on the vertical shear
mechanism - essentially dynamical, rather than
thermodynamic
Jiang et al. 2004
23
We (Bellon and Sobel 2007) use the QTCM2 (Sobel
and Neelin 2006, building on Neelin and Zeng
2000)
Vertical structure
Mass conservation (pt-pb) ?yv0(t,y) - pb
?yvb(t,y)
24
The model is similar in some but not all ways to
those of Jiang et al. (2004), Drbohlav and Wang
(2005)
Parameterizations Convection Betts-Miller (a
quasi-equilibrium scheme) Radiation newtonian
cooling towards a uniform temperature.Aquaplanet,
axisymmetric, on the ß-planeForcing
25
This model produces a nice intraseasonal
northward-propagating oscillation, robustly to
parameters
Precipitation (mm/d)
time
Latitude (1000s km)
26
Wind-induced sfc fluxes are crucial to the model
instability, hence dependence on mixed layer
depth is as before
Period growth rate from linear model
1/growth rate
period
Mixed layer depth
27
If this model were relevant to reality, it would
imply damping of ISO over land as observed
28
Summary

• Simple models of several types have intraseasonal
  oscillations that depend on surface fluxes
  these oscillations are damped without ocean heat
  storage
• At least one GCM works similarly (though at least
  one other doesnt)
• Observed ISO (at least in SH summer) has
  substantial net surface energy flux anomalies in
  more or less correct phase to drive the
  oscillation
• Observed variance of ISO is maximum over ocean,
  minimum over land, in both seasons and hemispheres

29
Concluding remarks

• We suggest it is likely that surface fluxes
  (turbulent and radiative) are important to the
  energetics of the ISO.
• This is testable in models. To what extent is it
  true in different models, and does the answer
  correlate with goodness (or other properties) of
  ISO simulation?
• Even if true, it would neither mean we deeply
  understand the ISO, nor that we could simulate or
  predict it.
• Still, if we could decide conclusively on this it
  might be a step forward

30
(No Transcript)
31
Mean states
Results two limit cycles
Limit Cycle 1
Limit Cycle 2
CMAP July, 80E-90E