A Model El Nino – Southern Oscillation

A Model El Nino –
Southern Oscillation
Stephen E. Zebiak and Mark A. Crane
Monthly Weather Review
Volume 115
October 1987
Presentation by Andrew Condon
Rosenstiel School of Marine and Atmospheric Science
University of Miami
Talking Points
Model Description
- Atmosphere
- Ocean
- Coupling
Results (Standard Case)
Model Sensitivities
- Atmospheric Parameterizations
- Oceanic Parameterizations
Influences of the Annual Cycle
Elements of the Model Oscillation
El Niño: A phenomenon in the equatorial Pacific Ocean characterized
by a positive sea surface temperature departure from normal (for
the 1971-2000 base period) in the Niño 3.4 region (120°W-170°W,
5°N-5°S) greater than or equal in magnitude to 0.5°C, averaged
over three consecutive months.
The Southern Oscillation: an oscillation in the surface pressure
(atmospheric mass) between the southeastern tropical Pacific and
the Australian-Indonesian regions. When the waters of the eastern
Pacific are abnormally warm (an El Niño event) sea level pressure
drops in the eastern Pacific and rises in the west. The reduction in
the pressure gradient is accompanied by a weakening of the lowlatitude easterly trades.
Major El Niño events since 1950: 1957-58, 1965, 1968-69, 1972-73,
1976-77, 1982-83, 1986-87, 1991-92, 1994-95, and 1997-98.
Image Source:
Following the 1982-83 El Nino a large scientific effort was undertaken
to better understand this event
► Key findings from these studies include: the importance in tropical
Pacific SST anomalies in producing observed atmospheric anomalies
during ENSO and the observed Pacific SST and sea level anomalies
during ENSO result primarily from the influence of surface wind stress
► Most studies prior to this had focused on one aspect or the other (i.e.
atmosphere or ocean) but few dealt with the coupled system. Those
that did were highly parameterized and did a poor job of coupling the
Model Description
Circulation is forced by a
heating anomaly that depends
partly on local heating
associated with SST anomalies
and partly on low-level moisture
convergence (found in obs and
GCM studies to be important,
not usually incorporated in
- Nonlinear convergence feedback
is incorporated using an
iterative procedure in which the
heating at each iteration
depends on the convergence
field from the previous iteration
Rectangular basin extending
from 124°E to 80°W and from
29°N to 29°S.
- Frictional layer to account for
wind driven currents
- Surface currents are generated
by spinning the model up with
monthly climatological mean
- There is 3D temperature
advection by the specified mean
current and an anomalous
current produced in model
- Heat flux anomaly is
proportional to SST anomaly,
always acting to adjust to the
climatological mean state
Model Description
► Ocean
component is forced by surface wind stress
► Allow time dependence only in moisture
convergence component of heating
► Heating related directly to SST gives a wind
response in equilibrium with the SST field on a
time scale of 10 days
► Heating due to internal moisture convergence
operates in a time stepping sense and forces a
wind field adjustment on a time scale of 1 month
90 year model run
initiated with a westerly
wind anomaly in the
region 145°E to 170°W
for a period of four
► Quasi-regular oscillations
with a favored period of
3-4 years
► Peak in June or around
the end of the year with a
total duration of between
14 and 18 months
► Peak first in the east and
later in the central region
Observations from
Rasmusson and
Carpenter from 1921-76.
► Irregular oscillations with
a preference for a 3-4
year period
► Major events have a
duration of somewhat
longer than a year
► Model was unable to
reproduce initial coastal
warming which is usually
observed prior to the
major central Pacific
Time series of area
averaged wind anomalies
During major warm
events the two indices
vary in a similar fashion
indicating a large scale
coherent wind forcing
Western Pacific zonal
wind anomalies are
weaker than observed
and switch from westerly
to easterly later than
► Following
showcase a major
warm event in three
month intervals
beginning in the end of
year 30 and going to
year 32
► December of year 30,
no appreciable
anomalies in either
SST or winds
March (31)
region of warm SST
has developed in the
equatorial zone east
of 170°W with a max
near 130°W
► Small westerly wind
anomalies near
130°W to 160°W
June (31)
Warm event is well
SST anomalies exceeding
1°C in the eastern
equatorial Pacific
Sizeable western wind
anomalies in the central
Model warming tends to
occur uniformly rather
than initially at the coast
September (31)
Expansion and amplification of
SST and wind anomalies in the
fall of an ENSO year
Warm anomalies as far
westward as 160°E and eastern
Pacific anomalies exceed 2°C
Large westerly wind anomalies
cover the whole equatorial
central Pacific
A region of small negative SST
anomaly and easterly wind
anomaly has developed in the
Western Pacific
December (31)
Peak temperature
anomalies occur with a
max at the coast and
another one near 140°W
SST anomaly has
expanded meridionally
Westerly wind anomalies
of about 2 m s-1 in the
central Pacific
Easterly anomalies in the
western Pacific in obs, not
shown in the model
March (32)
Temperature anomalies
begin to decrease,
especially at the east coast
► Single max in the eastern
central Pacific
► Large westerly wind
anomalies in the central
► Increasing easterly
anomalies farther to the
June (32)
Eastern ocean is still
warm, but temperatures
are decreasing rapidly
► Westerly wind anomalies
have decreased and
receded westward
► Stronger easterlies are
evident in the east
► Composites show cold SST
anomalies and poleward
wind anomalies in the
eastern ocean at this time
September (32)
Dramatic change in winds
and SST amounting to a
rapid termination of the
warm event
► Equatorial eastern and
central ocean is cold
► Winds are primarily
meridional and directed
► Agrees well with composites
for this time period
► All major events evolve in a
similar fashion. Some
smaller amplitude anomalies
develop differently and do
not conform to this scenario
Model thermocline depth anomaly along
the equator between years 30 and 45
Measure of heat content of the upper
Major warm episodes (31 and 41) are
characterized by anomalously high heat
content in the east and low heat content
in the west
Periods preceding major warm events
are characterized by above-normal heat
content at all longitudes
Periods immediately following warm
events show a deficit of heat content
Rise in equatorial heat content precedes
the development of equatorial westerlies
and positive SST anomalies in the
eastern ocean; that is it precedes the
ENSO event
Strong and sustained westerly wind
anomalies in the central Pacific during
warm phase
Occurs in phase with the anomalously
high heat contents in the eastern ocean
and low heat content in the western
At times (38 and 44) westerly anomalies
appear in a fashion similar to the onset
of ENSO, but the presence of strong
easterly anomalies in the eastern Pacific
contrast them
The easterly anomaly grows with the
westerly anomaly and development
ceases shortly thereafter
Terminated events tend to start early in
the year (Jan-March), may be influenced
by annual cycle
ENSO summary
Large scale pattern of equatorial westerly wind anomalies in the
central Pacific and equatorial SST anomalies that extend across most of
the basin and decrease in amplitude from east to west
The climatologically mean state includes easterly trade winds blowing
across the eastern and central ocean
Easterly stress induces an equatorial upwelling and sets up a zonal tilt
to the thermocline
Cold sub-thermocline water is far removed from the surface in the west
and very near the surface in the east
Sets up a sub-surface temperature anomaly that is largest in the east
and smaller toward the west
Atmospheric response is equatorial westerly wind anomalies that span
nearly the entire region of SST anomalies
Westerly wind deepens the eastern ocean thermocline, suppresses
equatorial upwelling, and sets up eastern current anomalies which
reinforce the temperature anomaly pattern
Model Sensitivities
Atmospheric Parameterization
25-year run based on initial
conditions present at the start
of year 31, only considering
NINO3 SST index
Coefficient of heating term
proportional to SST anomalies
increased 10%
Results in a large increase in
amplitude of events, but similar
Coefficient of heating
proportional to moisture
convergence is increased by 7%
Net impact on the larger scale
structure is minimal
Model Sensitivities
Oceanic Parameterizations
Cases 1-4 both the amplitude and time
scale of the oscillations increase
Reduced thermal dissipation allows
larger SST anomalies
Larger drag coefficient produces
greater wind stress forcing resulting in
larger anomalies
Stronger mean upwelling yields larger
SST anomaly in response to
thermocline displacements
Smaller equivalent depth produces
larger thermocline variations and larger
subsurface temperature anomalies
5 decreases the coupling strength,
resulting in smaller oscillations with a
shorter period
6 indicates low sensitivity
A sizeable decrease in background
ocean dissipation produces no change
in the solution, however other
experiments with larger dissipation
result in reduced amplitude
Influences of the Annual Cycle
Warm events tend to amplify sharply
during the northern summer, reach peak
amplitude around the end of the year and
diminish the following year
To examine annual cycle influence in the
model, experiments were performed using
initial conditions from Jan of year 31 and
the annual cycle was turned off at a given
subsequent month by holding the mean
fields fixed from then on
April (0) shows the growth of warm event
is retarded considerably (amplitude
increases more slowly, reaches a max later
and decreases sharply)
August (0) shows the amplitude continues
to rise for many months, peaks later and
larger than with annual cycle
December (0) shows a steady decline into
the middle of year 1
July (+1) shows growth of negative
anomalies continues longer and leads to
larger anomalies during year 2+
Annual Cycle
Summer period is most favorable for
rapid growth of both positive and
negative anomalies (mean winds, mean
upwelling and mean SST gradient are all
Spring period is least favorable for
anomaly growth (weak trade winds and
associated upwelling and SST gradient)
Small anomalies present during the
spring of an ENSO year amplify rapidly
during the summer and fall
During winter the coupling strength
decreases significantly and becomes
insufficient to maintain the large
Hesitation period in the spring followed
by an increase in coupling strength and
demise of event in the summer
Second experiment in which cycle turned
off July (0) and continues for 25 years
shows the solution settles into periodic
Elements of the model oscillation
The subsurface temperature is made more
or less sensitive to changes in the areaaveraged heat content
When made completely insensitive to
changes in heat content the ENSO
oscillation stops and is replaced by an
annual oscillation
When exactly half of the actual heat
content fluctuation is allowed an oscillation
with a period of 5-6 years develops
When the effect is enhanced by a factor of
3 the period is 1-2 years
The oscillatory character of the coupled
system depends on the effect of variations
in net equatorial heat content
The 3-4 year period preference reflects a
time delay between dynamical changes in
the eastern ocean and associated largescale fluctuations in equatorial wind stress
Without anomalous external forcing the model produces recurring warm
events that are irregular in both amplitude and spacing, but favor a 3-4
year period
Largest growth occurs during the northern summer and fall and termination
during the following spring and summer
Signature of model warm events include equatorial westerly wind anomalies
in the central Pacific and large SST anomalies in the eastern Pacific
All parameter changes that amounted to increasing (decreasing) the
strength of the atmosphere-ocean coupling tended to produce larger
(smaller) amplitudes and longer (shorter) periods
There is a buildup in heat content prior to the onset of a warm episode and
a rapid decrease during the course of the event
The variations in heat content is due to a phase lag between wind stress
and thermocline motions, this is why interannual oscillations occur in the
coupled system
Above normal heat content is a necessary precondition for the onset of a
warm episode.
All mechanisms essential to the ENSO cycle are contained in the Pacific
region and they are largely controlled by deterministic processes in the
tropical Pacific atmosphere-ocean system
Additional References
► http://www.nws.noaa.gov/ost/climate/STIP/
► http://ess.geology.ufl.edu/usra_esse/el_nino