Dynamics
of monsoon-induced biennial variability in ENSO
K.-M. Kim and K.-M. Lau
NASA/Goddard
Space Flight Center, Greenbelt, Maryland
Geophysical
Research Letter
(in
Press)
Abstract. The mechanism of the
quasi-biennial tendency in ENSO-monsoon coupled system is investigated using an
intermediate coupled model. The monsoon wind forcing is prescribed as a
function of SST anomalies based on the relationship between zonal wind
anomalies over the western Pacific to sea level change in the equatorial
eastern Pacific. The key mechanism of quasi-biennial tendency in El Niņo
evolution is found to be in the strong coupling of ENSO to monsoon wind forcing
over the western Pacific. Strong boreal summer monsoon wind forcing, which lags
the maximum SST anomaly in the equatorial eastern Pacific approximately 6
months, tends to generate Kelvin waves of the opposite sign to anomalies in the
eastern Pacific and initiates the turnabout in the eastern Pacific. Boreal
winter monsoon forcing, which has zero lag with maximum SST in the equatorial
eastern Pacific, tends to damp the ENSO oscillations.
1.
Introduction
The presence of a distinctive tropospheric
biennial oscillation (TBO) associated with El Niņo and Asia-Australian monsoon
is well known. Lau and Sheu [1988]
found a biennial peak in a global rainfall pattern associated with El Niņo. Rasmusson et al. [1990] and Ropelewski et al. [1992] identified
spatial patterns of sea surface temperatures and wind associated with the
quasi-biennial variability, similar to that of El Niņo. Numerous studies have
also reported the presence of strong TBO variability in the Asian summer
monsoon, involving possibly both regional and Indo-Pacific basin-scale coupled
ocean-atmosphere processes [Yasunari, 1990;
Meehl, 1993, 1997; Shen and Lau 1995; Goswami, 1995]. Recently Chang
and Li [2000] suggested that TBO may stem from processes within the Indian
ocean and the maritime continent. Clarke
and Shu [2000] suggested the western Pacific wind anomalies may phase-lock
the ENSO to the seasonal cycle.
It is also well known that the Asian summer
monsoon is negatively correlated with an El Niņo [Rasmusson and Carpenter, 1982, Webster
and Yang, 1992, Lau and Yang,
1996 and many others]. Wainer and Webster
[1996] argued that the interannual variation of the summer monsoon may
contribute to irregularities of El Niņos. Chung
and Nigam [1999] showed that, based on results from an intermediate
ocean-atmosphere coupled model, that monsoon forcing may increase the frequency
of the occurrence of El Niņo's. Yasunari
and Seki [1992] and Shen and Lau
[1995] suggested that the biennial tendency is likely to be a fundamental time
scale involved in monsoon-ENSO interaction. However, it has never been make
clear how the ENSO itself may be affected by the monsoon and why are there
strong TBO signals in both the ENSO and monsoon. Recently, Lau and Wu [2000] showed
the observational evidence of strong monsoon-ENSO interaction in El Niņo's that
exhibit strong biennial tendency. They hypothesized that the biennial
variability in El Niņo is induced by strong monsoon forcing in the equatorial
western Pacific. The objective of this paper is providing a theoretical basis
of that hypothesis.
2.
Western Pacific wind anomalies
Recently Weisberg and Wang [1997] and Wang
et al [1999] showed that the far western Pacific and the central Pacific
constitute two separate surface wind systems in governing ENSO variability.
This situation is illustrated in Fig. 1, which shows the lagged correlation of
monthly sea level anomalies in the equatorial eastern Pacific (5°S-5°N, 140°W-120°W) with 1000 mb wind
in the equatorial western Pacific (5°S-5°N,
125°E-145°E, solid line) and
central Pacific (5°S-5°N, 180°E-160°W, dashed line). The
wind data is from the National Center for Environmental Prediction (NCEP)
reanalysis [Kalnay et al. 1996] and
the sea level data is from the NCEP ocean assimilation data [Ji et al. 1995] from January 1980 to
August 1998. It is clear that the central Pacific wind has the highest
simultaneous correlation with the sea level variability in the eastern Pacific
with westerly wind coinciding with the maximum sea level height. In contrast,
the western Pacific wind has the highest correlation with the westerly
(easterly) wind leading the sea level rise (fall) in the eastern Pacific by
approximately six months. There is also a significant negative correlation at
about 5-6 month lag. The correlation suggests a biennial tendency in the
surface wind in the western Pacific with respect to the thermocline variability
in the eastern Pacific. However, it does not necessarily mean that the surface
wind is directly forced from the eastern Pacific. The time delay of western
Pacific wind anomalies may stem from air-sea interaction associated with the
development of off-equatorial SST anomaly in the western Pacific (Wang et al.
1999). Alternatively, the wind
anomalies may be related to the excitation of an anomalous West Pacific
anticyclone (Lau and Wu 2000, Wang et al 1999). Another possibility is the
propagation of wind anomalies from the Indian Ocean (Barnett 1984, Shen and Lau
1995). All these mechanisms are likely be linked to anomalies in the boreal
summer and winter Asian monsoons.
3.
Model Experiments
We use a modified intermediate coupled
ocean-atmosphere model of Zebiak and Cane
[1987] to elucidate the fundamental dynamics of biennial variability in
monsoon-ENSO interaction. This model has been used in various theoretical
studies of coupled ocean-atmosphere interaction and in experimental
seasonal-to-interannual prediction [Cane
et al., 1986; Tziperman et al.,
1998; Chen et al., 1998]. While this
model lacks detailed physics of the atmospheric and oceanic processes, it does
capture the essential dynamics of El Niņo evolution [Battisti, 1988; Zebiak and
Cane, 1987]. One of the major limitations of the model is the lack of the
influence from monsoon processes to the west of the dateline. As a result, the
model has very weak variability of SST and surface wind in the western Pacific,
which are inconsequential to the model ENSO evolution Because the seasonal cycle
is specified in the model, the seasonal variation of monsoon implicitly plays a
role in the model. Both the interannual
and intraseasonal variability of the monsoon may affect ENSO. In this study, we
only focus on the interannual component that is associated with El Nino. Subseasonal component is neglected in this
study because the model is not sensitive to these shorter time scale forcings.
Because of limited domain and model physics, monsoon effects cannot be fully included. Therefore the monsoon wind forcing over the equatorial western Pacific is only considered. Based on the previous studies and the observational evidence shown in Fig. 1, the monsoon wind forcing is parameterized as a linear relation between NINO3 SSTA (TEPAC) and monsoon wind forcing as follows.
U mons = -a F(TEPAC
(t -d)) (1)
Where
a
denotes the magnitude of the coupling between the NINO3 SSTA and the imposed
monsoon wind (Umons). A
value of a=1.0
corresponds to about 2 m/sec maximum wind speed over the western Pacific, when
the Nino3 SSTA is 3K. The time lag d, between the Nino3 SSTA and the monsoon wind
forcing is based on the relationship shown in Fig. 1. The spatial function F, is prescribed as a Gaussian profile
in the meridional direction with half-width about 10 degree within the equator,
and zonal wind patch between the western edge of the ocean domain(130°E) and the eastern
edge of the warm pool(160°E).
As we shall show, the present results are insensitive to the details of the
spatial distribution, but more to the coupling coefficient, a
and the time lag, d.
As discussed previously, this lagged relationship may be attributed to the wind
forcing from the two monsoons, i.e. zero lag for the boreal winter and 6 months
lag for the boreal summer. In the following discussion, we will describe the
temporal variations of El Nino, using SST anomalies averaged over the eastern
Pacific (NINO3 SSTA). For propagating features, we use thermocline depth
anomalies which is the only model variable corresponding to sea level.
4.
Results
We begin with the discussion of the impact
of the time delay, and the strength of coupling in the prescribed monsoon
forcing on the evolution of ENSO. For the control experiment (a=0),
we have chosen typical model parameters so that the model exhibits an ENSO
oscillation with pronounced cycle at 3.4 years. As the coupling increases, for
d=0,
the model ENSO oscillations are strongly damped (Fig. 2a). In contrast, for
d=
6 months, the oscillations evolve with increasing frequency locking towards a
biennial periodicity (Fig. 2b). It is clear that by the time a=2
(which about doubles the typical present-day value), the model ENSO is locked
into a pronounced limit cycle with periodicity of exactly two years. The
aforementioned frequency evolution is also apparent in the spectra, as a
function of a
(Fig. 2 c and d). It is interesting to note that the frequency modulation
appears to occur in steps, typical of nonlinear systems. When the coupling
coefficient is weak at the range of 0.1 to 0.5, the dominant periodicity is 3
years. A bifurcation occurs at about a= 0.5, when
multiple periodicities are excited. For a>1.0,
increasing frequency locking to the biennial time scale is apparent.
Figure 3 shows the composite time-longitude
sections of surface wind and thermocline depth anomalies along the equator for
the control (a=0, d=0) and for a=1.5, d=6. Apparent in the control is an intrinsic
oscillation of warm (El Niņo) and cold (La Nina) state with a periodicity of
approximately 42 months. Thermocline anomalies propagate eastward and are
strongest in the two ends of the ocean domain. Strong coupling of wind and
thermocline occur only in the eastern Pacific, with low-level convergence
(divergence) overlying positive (negative) thermocline anomalies. As stated
previous, in the control, the model surface winds are weak in the western
Pacific and play no role in the generation of the thermocline anomalies. When
an interactive monsoon wind with a time lag of 6 months (with respect to the eastern
Pacific thermocline/SST anomaly) is introduced, the model intrinsic oscillation
equilibrates to a periodicity of approximately 28 months. In this case, both
wind system in the western and the eastern Pacific play important role in the
oscillation. Over the eastern Pacific the coupling wind/thermocline mechanism
is the same as in the control. However, in the western Pacific, strong phase
locking of the anomalies with the seasonal cycle occurs. Here, maximum
thermocline anomalies occur in and around boreal winter about 6 months after
the maximum monsoon wind forcing in the boreal summer. The eastward propagation
of the coupled thermocline and wind anomalies is very pronounced, and takes
approximately 12 months to complete. As a result, the entire system oscillates
at a cycle of approximately 2 years.
Numerous additional experiments have been carried out with different variations of the monsoon forcing and time lags. One set of experiments is with the monsoon forcing invoked only for December through February (DJF) with zero lag, and for June through August (JJA) with 6 months lags. Yet another set was carried out with imposed stochastic wind forcing added to (1). These variations in wind forcing do not change the nature of the aforementioned results.
5.
Summary and Discussions
The present results illustrate a simple,
basic mechanism in which monsoon wind forcing in the tropical western Pacific
may affect ENSO variation. In the control, the model ENSO cycle can be
described in terms of the delay oscillator mechanism [Suarez and Schopf, 1988; Battisti
1988]. Coupled wind and thermocline/SST anomalies are generated continuously by
the action of strong air-sea interaction in the eastern Pacific. The associated
cyclonic wind stress curl spawns equatorially trapped westward propagating
oceanic Rossby waves, which are reflected as an oceanic Kelvin wave at the
western boundary. The upwelling (downwelling) Kelvin wave propagates eastward
and generates negative (positive) thermocline anomalies which replaces the existing
positive (negative) anomalies and perpetuates the model El Niņo cycle. When the
monsoon wind forcing is applied with zero lag, an eastward propagating
equatorial Kelvin wave is generated producing thermocline perturbation of the
opposite sign to an anomaly already existing in the eastern Pacific,
effectively annihilating the anomaly in the eastern Pacific. The ENSO
oscillation is therefore strongly damped as illustrated in Fig. 4a, which
describes the situation that occurs during the boreal winter.
In the case of monsoon wind forcing with
6-month lag, the delayed oscillator mechanism is strongly modulated by wind
forcings in the western Pacific. The sequence of events is shown schematically
in Fig. 4b. Here, maximum thermocline anomaly develops alternatively in the
eastern and western Pacific, at about 12 months apart, around boreal winter. In
the western Pacific, the thermocline anomaly develops, about 6 months after the
maximum monsoon wind forcing, which occurs in the intervening boreal summer season.
In the transition, say from cold to warm phase in JJA(1) in the eastern
Pacific, the monsoon easterly wind
forces the thermocline anomaly to increase in the far western Pacific, and in
doing so sends an upwelling Kelvin wave to the east, initiating a turnabout in
the eastern Pacific. This upwelling Kelvin wave enhances and supplants the
negative Kelvin wave generated by Rossby wave-reflection at the western
boundary by the delayed oscillator mechanism, eventually leading to the
establishment of the cold phase in DJF(2).
This is followed by the development of westerly wind anomalies in the
western Pacific 6 months later in JJA(2).
The biennial cycle then repeats. Hence the imposed monsoon forcing cause
the coupled monsoon-ENSO system to equilibrate toward a biennial see-saw
oscillation across the Pacific, with the monsoon acting as a
"pace-maker" for the model ENSO oscillation.
It is worth pointing out that the results presented here are highly idealized, so that detailed comparison with actual observations is not warranted. In the model coupled system, the ENSO oscillation period is either 3-4 year or quasi-biennial depending on the strength of the monsoon-ENSO coupling. In reality, both periodicities can occur simultaneously and many other factors can modulate or induce irregularities in ENSO cycles. Nonetheless, the basic mechanisms proposed here is specifically relevant to monsoon forcing and can be further tested in more sophisticated coupled models.
Acknowledgements. This
work is supported by the NASA Earth Science Enterprise Modeling and Data
Analysis Program. The authors would like to thank Dr. G. Meehl for providing
useful comments to this paper.
References
Barnett, T.P., 1991: The interaction of multiple
timescales in the tropical climate system.
J. Climate, 4, 269-285.
Battisti, D. S., 1988: Dynamics and
thermodynamics of a warming event in a coupled tropical atmosphere ocean model,
J. Atmos. Sci., 45, 2889-2919.
Cane, M. A., S. E. Zebiak, and S. C. Dolan,
1986: Experimental forecasts of El Niņo. Nature,
321, 827-832.
Chang, C.-P. and T. Li, 2000: A theory for the
tropospheric biennial oscillation. J.
Atmos. Sci. 57, 2209-2224.
Chen, D.K., M.A. Cane, S.E. Zebiak, A. Kaplan,
1998: The impact of sea level data assimilation on the Lament model prediction
of the 1997/98 El Niņo. Geophys. Res.
Lett., 25, 2837-2840.
Chung, C. and S. Nigam, 1999: Asian summer
monsoon-ENSO feedback on the Cane-Zebiak model ENSO. J. Climate, 12, 2787-2807.
Clarke, A. J. and L. Shu, 2000: Quasi-biennial
winds in the far western equatorial Pacific phase-locking El Niņo to the
seasonal cycle. Geophys. Res. Lett., 27, 771-774.
Goswami, B. N., 1995: A multiscale interaction
model for the origin of the tropospheric QBO. J. Climate, 8, 524-534.
Ji, M., A. Leetmaa, and J. Derber, 1995: An
ocean analysis system for seasonal to interannual climate studies. Mon. Wea. Rev., 123, 460-481.
Kalnay, E., M. Kanamitsu, R. Kistler, W.
Collins, D. Deaven, L. Gandin, M. Iredell, S. Saha, G. White, J. Woollen, Y.
Zhu, M. Chelliah, W. Ebisuzaki, W. Higgins, J. Janowiak, K.C. Mo, C.
Ropelewski, J. Wang, and A. Leetmaa, 1996: The NCEP/NCAR 40-year reanalysis
project. Bull. Amer. Meteor. Soc., 77, 437-471.
Lau, K.-M. and P. J. Sheu, 1988: Annual cycle,
quasi-biennial oscillation, and Southern Oscillation in global precipitation. J. Geophys. Res., 93, 10975-10988.
Lau K.-M. and S. Yang, 1996: The Asian monsoon
and predictability of the tropical ocean-atmosphere system. Q. J. Roy. Meteor. Soc., 122, 945-957.
Lau K.-M. and H.-T. Wu, 2000: Intrinsic coupled
ocean-atmosphere modes of the Asian summer monsoon: A re-assessment of
monsoon-ENSO relationships. J. Climate,
accepted.
Meehl, G. A., 1993: A coupled air-sea biennial
mechanism in the tropical Indian and Pacific regions: Role of the ocean. J. Climate, 6, 31-41.
Meehl, G. A., 1997: The south Asian monsoon and
the tropospheric biennial oscillation. J.
Climate, 10, 1921-1943.
Rasmusson, E. M. and T. H. Carpenter, 1982:
Variations in tropical sea-surface temperature and surface wind fields
associated with the Southern Oscillation El-Nino. Mon. Wea. Rev., 110,
354-384.
Rasmusson, E. M., X. Wang, and C. F. Ropelewski,
1990: The biennial component of ENSO variability. J. Mar. Sys., 1, 71-96.
Ropelewski, C. F., M. S. Halpert, and X. Wang,
1992: Observed tropospheric biennial variability and its relationship to the
Southern Oscillation. J. Climate, 5, 594-614.
Suarez, M. J. and P. S. Schopf, P. S., 1988: A
delayed action oscillator for ENSO. J.
Atmos. Sci., 45, 3283-3287.
Shen, S.-H., and K. M. Lau, 1995: Biennial oscillation associated with the East Asian summer monsoon and
tropical sea surface temperature. J.
Meteor. Soc. Japan, 73, 105-124.
Tziperman E., M. A. Cane, S.E. Zebiak, Y. Xue,
and B. Blumenthal, 1998: Locking of El Niņo's peak time to the end of the
calendar year in the delayed oscillator picture of ENSO. J. Climate, 11,
2191-2199.
Wainer, I., and P. J. Webster, 1996: Monsoon/El
Niņo-Southern Oscillation relationships in a simple coupled ocean-atmosphere
model. J. Geophys. Res., 101, 25599-25614.
Wang B., R. Wu, and R. Lukas, 1999: Roles of the
western North Pacific wind variation in thermocline adjustment and ENSO phase
transition. J. Meteor. Soc. Japan, 77, 1-16.
Wang C.,
R. H. Weisberg, and J. I. Virmani, 1999: Western Pacific interannual variability
associated with the El Nino-Southern Oscillation. J. Geophys. Res., 104,
5131-5149.
Webster, P.J. and S. Yang, 1992: Monsoon and
ENSO-Selectively interactive systems. Q.
J. Roy. Meteor. Soc., 118,
877-926.
Weisberg, R. H. and C. Wang, 1997: Slow variability
in the equatorial west-central Pacific in relation to ENSO. J. Climate, 10, 1998-2017.
Yasunari, T., 1990: Impact of Indian monsoon on
the coupled atmosphere/ocean system in the tropical Pacific. Meteorol.
Atmos. Phys., 44, 29-41.
Yasunari, T. and Y. Seki, 1992: Role of the
Asian monsoon on the interannual variability of the global climate system. J. Meteor. Soc. Japan, 70, 177-189.
Zebiak, S. E. and M. A. Cane, 1987: A model
ENSO, Mon. Wea. Rev., 115, 2262-2278.
(Received: September 23, 2000; Accepted: October 24,2000)
Figure 1. Lag correlation of monthly sea level anomalies in the equatorial eastern Pacific (5S-5N, 140W-120W) with 1000 mb wind in the equatorial western Pacific (5S-5N, 125E-145E, solid line) and central Pacific (5S-5N, 180E-160W, dashedline). Positive lags mean that changes in wind lead those in sea level. |
Figure 2. Figure 2. Simulated NINO3 SSTA time series (a) and its power spectra (c) with various range of coupling magnitude (a) for lag=0. (b) and (d) are same as (a) and (c), except for lag=6. Units of (c) and (d) are (K2/f). |
Figure 3. Composite of simulated thermocline depth (color) and wind stress (contour) anomalies averaged between 2N and 2S along the equator for the control run. (b) Same as (a), except for the experiment with lag = 6. |
Figure 4. (a) Schematics showing possible interaction of the boreal winter monsoon with ENSO as a damped oscillator. Shaded patches represent the sea level anomalies. Open arrows indicate zonal wind anomalies. Thin and wavy arrows represent the Rossby and Kelvin wave propagation, respectively. (b) Same as in (a) but for a biennial oscillator induced by the interaction of the monsoons (summer and winter) with ENSO. |