Monsoon Interannual Variability

1

• Monsoon Interannual Variability
• Relationship to ENSO and Tropospheric Biennial
  Oscillation (TBO)
• Tim Li
• University of Hawaii

• Introduction - observations
• Monsoon-ENSO relationships
• Biennial and lower-frequency variability of
  Indian monsoon
• Asian-Australian monsoon phase relations
• Seasonal evolving TBO structures
• A monsoon-warm ocean interaction scenario for the
  TBO

2
Webster et al. 1998
3
El Nino structure and time evolution
4
Gilbert Walker 1924 Southern Oscillation
5
El Nino-Southern Oscillation (ENSO)
J. Bjerknes, 1969
6
Normal condition
Mechanism for El Nino growth and phase
transition A positive dynamic ocean-atmosphere
feedback Bjerknes feedback J. Bjerknes 1969
El Nino condition
7
A delayed negative feedback leads to El Nino
phase transition
8
Seasonal differential precipitation and winds
(JA-JF)
Top domains for three sub-monsoon systems IM
westerly, north-south T gradient EAM southerly,
east-west T gradient WNPM hemispheric asymmetric
SST gradient
EAM
IM
WNPM
WNPM
Bottom area-averaged rainfall evolution WNPM
strongest, with a peak phase lag to IM and EAM.
IM
JA-JF
EAM
Wang, Clemens and Liu 2003
9
Wang and LinHo 2002
(Climatology 1979-2001)
7/11
9/15
7/01
6/21
6/11
7/11
6/01
7/01
6/11
5/21
6/21
6/01
8/10
7/20
5/21
5/11
6/01
5/01
6/15
5/21
4/21
5/11
10

• How does the ENSO influence the monsoon?
• Indian monsoon ENSO relationship
• WNP monsoon ENSO relationship
• East Asian monsoon ENSO relationship
• What are other factors that influences the
  monsoon in addition to ENSO?
• TBO
• Indian Ocean SSTA/western Pacific SSTA
• Tropical-mid-latitude interaction

11
Possible role of Indian Ocean IO SSTA warm prior
to a strong Indian Monsoon, but cold after a
strong monsoon. No significant simultaneous
correlation between Indian monsoon and IO SSTA
Lau and Yang 1996
12

• How does the El Nino impact South Asian monsoon?
  Why is the atmosphere response equatorially
  asymmetric to a symmetric forcing? How the SSTA
  off Sumatra is initiated? Why is the Indian
  monsoon anomaly opposite to the WNP monsoon
  anomaly?

12-El Nino composite JJA(0)
13
Gill model
The governing equations
Two special cases I, The heating is symmetric
about the equator and has the form of
II, The heating is antisymmetric about the
equator with the form
The forcing will be assumed localized and to have
the form
For a steady state
where
14
Gill model
Model results For symmetric forcing, there are
two parts of the response. 1,
2,
15
Gill model

• Fig. 1 Solutions for heating symmetric about the
  equator in the region xlt2 for decay factor
  e0.1
• Contours of vertical velocity w ( solid contours
  are 0, 0.3, 0.6, broken contour is -0.1)
  superimposed on the velocity field for the lower
  layer. The field is dominated by the upward
  motion in the heating region where it has
  approximately the same shape as the heating
  function. Elsewhere there is subsidence with the
  same pattern as the pressure field.
• Contours of perturbation pressure p (contour
  interval 0.3) which is everywhere negative. There
  is a trough at the equator in the easterly regime
  to the east of the forcing region. On the other
  hand, the pressure in the westerlies to the west
  of the forcing region, though depressed, is high
  relative to its value off the equator. Two
  cyclones are found on the north-west and
  south-west flanks of the forcing region.
• The meridionally integrated flow showing (i)
  stream function contours, and (ii) perturbation
  pressure. Note the rising motion in the heating
  region (where there is a trough) and subsidence
  elsewhere. The circulation in the right-hand
  (Walker) cell is five times that in each of the
  Hadley cells shown in (c).

16
Gill model
For asymmetric forcing,

• Fig. 2 Solutions for heating antisymmetric about
  the equator in the region xlt2 for decay factor
  e0.1
• Contours of vertical velocity w (contours
  interval 0.3) super-imposed on the velocity field
  for the lower layer. The field is dominated by
  the motion in the heating region where it is
  approximately the same shape as the heating
  function (positive in the northern hemisphere).
  Outside the forcing region, the pattern is the
  same as for the pressure field with subsidence in
  the northern hemisphere and weak upward motion in
  the southern hemisphere. There is no motion for
  xgt2.
• Contours of perturbation pressure p (contour
  interval 0.3) which is positive in the south
  (where there is an anticyclone) and negative
  north of the equator (where there is a cyclone).
  The flow has the expected sense of rotation
  around the pressure centers and flows down the
  pressure gradient where it crosses the equator.
  All fields are zero for xgt2.

17
Gill model
Further solutions
Fig.3 Solution obtained by adding the solutions
shown in the two previous figures, corresponding
to heating which is confined to the region xlt2
and is mainly concentrated to the north of the
equator. (a) contours of vertical velocity w
(contour interval 0.3) showing the dominance of
the heating north of the equator. The flow to the
east of the forcing region is the same as in
Fig.1, this being provided entirely by the
symmetric part of the heating. West of the
forcing, the westerly inflow is concentrated
between the equator and y2. An easterly flow is
found south of the equator. (b) Contours of
perturbation pressure p (contour interval 0.3).
The pattern is dominated by a low on the western
flank of the heating region and by the
equatorial trough. A high is found in the
southern hemisphere.
18
How does the El Nino remotely impact the Asian
monsoon? Large-scale east-west overturning
? Equatorial asymmetric response to a symmetric
El Nino forcing, why?
Anomaly dry AGCM with specified 3D summer mean
flow and diabatic heating
Wang, Wu and Li 2003, J. Climate
19
East Asian Monsoon (EAM)-ENSO Relationship Observ
ational fact EAM meiyu-rainfall significantly
increases 6 months after the peak of ENSO.
By the summer after peak El Nino, the eastern
Pacific SST becomes normal. How can ENSO has a
delayed impacts on EAM?
El Nino developing year
El Nino decay year
20
Leading mode of Monsoon Interannual Variation
850 hPa winds and local SST anomalies (19572001)

• Biennial tendency associated with ENSO turnabout
• Distinct evolutions of anticyclonic anomalies
  over SIO and WNP

Wang, Wu, Li 2003, J. Climate
21
Wang and Zhang 2002, JC
22
A positive thermodynamic air-sea feedback
mechanism Wang et al. 2000
El Nino heating ? atmospheric Rossby wave
response ? cold SSTA in WNP ? anomalous AC ? EAM
23
(significance level 95 shaded)
Strong ENSO composite
eastward displacement of the Low-level
anticyclone
Another possible process that leads to formation
of anomalous Philippine Sea anticyclone
24
Chen, Li and Shih, 2007, J. Climate
divergence
Dry, cold
A
Subsidence
Wet, warm
dry
The zonal asymmetry of divergence, vertical
motion, and precipitable water leads to the
eastward propagation of the low-level anticyclone.
25
Quasi-biennial (2-3-yr) and lower-frequency (LF,
3-7-yr) variabilities of the Indian monsoon
rainfall
Monsoon prediction issue
Q What processes are responsible for the QB and
LF interannual variabilities of the Indian
monsoon rainfall?
Li et al. 2001, GRL
26
1000hPa moisture flux convergence (2-3 yr)
27

• Large-scale east-west circulation
• Land-ocean thermal contrast
• Synoptic wave activity/monsoon trough in WP

28
Li and Zhang 2002, J. Meteor. Soc. Japan
Local processes (IO SSTA moisture transport)
prior to the monsoon season

• Monsoon TBO

Large-scale overturning circulation
Convective activity at WP monsoon trough
Monsoon LFV
Land-ocean thermal contrast prior to the monsoon
onset
29
(No Transcript)
30
(No Transcript)
31
Indian and Australian Monsoon in-Phase
Relationship

• Why does a strong (weak) Australian monsoon
  often follow a strong (weak) Indian monsoon?
• Hypothesis El Nino control (Yanai and Liu 2001)

32
Velocity potential difference fields
33
Zhang and Li 2007
The result suggests that Indian Ocean SSTA may
play a active role in bridging the Indian and
Australian summer monsoons.
34
Monsoon TBO
Yasunari 1990
35
What is the TBO? The tendency of the monsoon
rainfall anomaly to flip-flop in successive
years is referred to as the tropospheric biennial
oscillation (TBO, Meehl 1994, 1997).
Observational studies have shown that the TBO is
manifested over various monsoon regions such as
Indonesia/northern Australia (Nicholls 1978,
Yasunari and Suppiah 1988), East Asia (Lau and
Sheu 1988, Tian and Yasunari 1992, Shen and Lau
1995, Chang et al. 2000ab, Wang and Li 2004) and
India (Mooley and Parthasarathy 1984 Meehl 1987
Rasmusson et al. 1990 Yasunari 1990, 1991).
36
Lagged correlation between ISM and SST in EP and
WP
Yasunari 1990
37
Shen and Lau 1995
38
A review on TBO theories

• TBO theories
• Local air-sea interactions (Nicholls 1978 Meehl
  1987 Clarke et al 1998)
• Inter-basin ocean-atmosphere interactions in the
  tropics (Chang and Li 2000, Li et al. 2001)
• Tropical-midlatitude teleconnection (Meehl
  1994,1997 Tomita and Yasunari 1996 Ogasawara et
  al 1999, among others)

39
Local air-sea interaction theory
Nicholls 1978 MWR
40
Meehl 1987 Monsoon-SST local negative feedback
mechanism
41
Meehl 1987, MWR
42
Chang and Li 2000, J. Atmos. Sci.
43
(No Transcript)
44
Meehl 1987
45
Li et al. 2006, JC
TBO activity centers
Time filter
46
A seasonal-sequence EOF analysis using NCAR/NCEP
reanalysis data (1950-1999) Shading
rainfall Vector 925mb wind 200mb wind first
baroclinic mode structure Data seasonal mean
(DJF, MAM, JJA, SON) fields of       
Precipitation, SLP, SST        925mb U and
V        200mb U and V        850mb
geopotential height
47
TAS feedback in western North Pacific (Wang et
al. 2000)
48
A Season-dependent Air-Sea Feedback in the
Southeast Indian Ocean
Warm
EQ
Sumatra
Cold
Anticyclonic Circulation
Mean southeasterly in boreal summer
? A primary mechanism that leads to the phase
locking of the Indian Ocean Dipole (Li et al.
2003, JAS)
49
What is the origin of the TBO ? Hypothesis
1 TBO is forced by remote forcing from the
eastern equatorial Pacific. Hypothesis 2 TBO is
an air-sea coupled mode in the monsoon region.
? This implies that the QB component of ENSO may
result from the inter-basin teleconnection
between the monsoon/warm ocean and eastern
Pacific.
50
A Monsoon-Warm Ocean Interaction Scenario for the
TroposphericBiennial Oscillation
Li, T., P. Liu, X. Fu, B. Wang, and G. A. Meehl,
2006 Tempo-Spatial Structures and Mechanisms of
the Tropospheric Biennial Oscillation in the
Indo-Pacific Warm Ocean Regions. J. Climate, Vol.
19, No. 13, pages 30703087.
51
TBO mechanism
?
?
QB Monsoon
QB ENSO
LF ENSO
LF Monsoon
Delayed oscillator
52

• IPRC Hybrid coupled atmosphere-ocean GCM (Fu et
  al. 2003, 2004)
• ECHAM4 AGCM
• UH 2.5-layer intermediate ocean model (Wang, Li
  and Chang 1995 Fu and Wang 1999)

50-yr integration
53
Hybrid coupled GCM simulations Summer vs. Winter
54
(No Transcript)
55
Coupled ECHAM4-ocean model 50-yr simulation
56
The hybrid coupled GCM experiments suggest that
the TBO originates from the monsoon-warm ocean
interaction. ?Analogy to the PNA pattern
(internal atmospheric dynamics vs. external
forcing) The TBO is an inherent monsoon mode,
while the El Nino forcing may magnify the signal.
Question Through what processes does the air-sea
interaction in the warm ocean lead to a TBO?
57
(No Transcript)
58

Top TRMM precipitation (mm/day) (annual mean
removed)
Bottom ECHAM4 simulated precipitation (mm/day)
59
Box average (10-15N, 140-160E)
A Total SST tendency B Net heat flux term C
Sum of ocean dynamic (3D advection) terms
60
Simulations from an intermediate coupled model
that suppresses the delayed oscillator
mode (i.e., no El Nino dynamics)
61
A monsoon-warm ocean interaction scenario for the
TBO
62
Conclusions

• During the El Nino developing summer, El Nino
  tends to decrease rainfall over India but enhance
  rainfall over WNP. During the El Nino decaying
  phase, a WNP anticyclone anomaly persists from
  northern winter to subsequent summer, leading to
  enhanced Meiyu rainfall over Central China.
• Indian summer monsoon has two significant
  spectrum peaks on quasi-biennial (2-3-yr) and
  lower-frequency timescales. Physical mechanisms
  determining the two timescale variabilities are
  different. The former is primarily attributed to
  local processes over the tropical Indian Ocean,
  while the latter is attributed to remote forcing
  from the Pacific.
• Asian-Australian monsoon in-phase relationship
  in addition to remote ENSO forcing, eastward
  propagation/expansion of Indian Ocean SSTA may
  contribute to the in-phase relationship.

63
Conclusions

• TBO structure and evolution The major convective
  activity centers associated with the TBO appear
  in SEIO and WNP, accompanied with anticyclonic
  (or cyclonic) circulation patterns with
  distinctive life cycles.
• TBO originates from air-sea interactions in the
  warm ocean. Key processes involved include the
  variability of the WNP monsoon and associated
  cross-equatorial flows, convective activity over
  Southeast Asia/maritime continent and anomalous
  Walker circulation, and ocean dynamic responses
  to anomalous wind stress curl in the western
  Pacific.