Transcript Document

El Nino – Southern Oscillation (ENSO)
圣婴现象和南方涛动
Mechanism, Prediction & Impacts
December 1982
SST Anomaly
The white areas off the tropical coasts of South and North America
indicate the pool of warm water
• El Niño/La Niña-Southern Oscillation, or ENSO, is a
quasiperiodic climate pattern that occurs across the
tropical Pacific Ocean roughly every five years.
• It is characterized by variations in the temperature of the
surface of the tropical eastern Pacific Ocean—warming or
cooling known as El Niño and La Niña respectively—and
air surface pressure in the tropical western Pacific—the
Southern Oscillation.
• The two variations are coupled: the warm oceanic phase,
El Niño, accompanies high air surface pressure in the
western Pacific, while the cold phase, La Niña,
accompanies low air surface pressure in the western
Pacific.
• Mechanisms that cause the oscillation remain under
study.
Discovering the Southern Oscillation
ENSO normal state
● Normal equatorial winds warm as they flow westward across the Pacific
● Cold water is pulled up along west coast of South America
● Warming water is pushed toward west side of Pacific
El Niño state
 Sea surface is warm in central and eastern Pacific
 Less cold water is pulled up along west coast of South America
 Hot air rises in central Pacific, travels east and west before cool
La Niña state
Warm water accumulates in far western Pacific.
Equatorial water is cooler than in the normal state
NINO3.4
NINO3
NINO3.4 = ENSO index, measuring
average SST anomaly within box
5S-5N; 170W-120W
NINO3 (or NINO3.4) is measure
of oceanic part of ENSO.
Southern Oscillation Index* (SOI)
is measure of atmospheric part
of ENSO.
These two indices are highly
correlated.
*Traditional Version:
SOI = SLPTahiti - SLPDarwin
(there’s also and equatorial version)
Correlation > 0.9
Although they have
similarities…
ALL El Niño events
are unique
Mechanism - How it works:
First understand the mean state
Tropical Pacific – Average State
Walker Circulation
Coupled Behavior in tropical Pacific
1 pocean h
 

 x
x
SST  pair  

SST
Gradient
x
Winds

T
T
dT
dT
 u  v
w
 T
t
x
dy
dz
  
dT 
 dT 
dT 

SST  w   w
 w  


 dz 
 dz 
dz 
Upper Ocean
Structure
(Thermocline)
Pacific Ocean Temperatures along Equator
Based on these observations
of equatorial temperatures:
1) Is eastern Pacific thermocline
deeper or shallower than
normal?
2) What direction are the
zonal wind anomalies?
3) Will eastern Pacific SSTs
get warmer or colder?
http://www.pmel.noaa.gov/tao/jsdisplay or http://www.tao.noaa.gov
What is the direct (i.e. oceanic) impact of
El Niño events on CO2 variability?
“There is thus ample reason
for a never-ending succession of
alternating trends by air-sea
interaction in the equatorial belt,
but just how the turnabout between trends
takes place is not yet quite clear.”
J. Bjerknes
1969
Klaus Wyrti in early 1970s
shows through observations
of sea level that changes in
upper ocean structure are
related to ENSO variability,
that can influence the initiation
of El Nino or transition between
El Nino and La Nina though
ocean dynamics
Klaus Wyrtki
Decrease of sea level = Thermocline rise
A dynamical response NOT surface heating
Wind Anomaly
applied for
30 days
Warm SSTa
Dynes/cm**2
Response of
upper-ocean
structure
-
+ Warm SSTa
(Courtesy: Dave DeWitt, IRI)
Evolution of
upper-ocean structure
(or thermocline)
anomalies
Perturbations move
eastward on the equator;
westward off the equator
Perturbations move slower
as latitude increases
(Courtesy: Dave DeWitt)
Continuing Evolution of
upper-ocean structure
(or thermocline)
anomalies
Warm SSTa
At western boundary,
waves are reflected and
channeled onto equator
 Delayed negative
feedback
Warm SSTa
(Courtesy: Dave DeWitt)
Wind Stress
a
Thermocline
Anomalies
Near peak
El Nino
b
c
Transition
(neutral)
d
Near peak
La Nina
e
Main Points:
* The tropical Pacific air-sea system is coupled, with the
pattern of SSTs, the low-level winds and the thermocline
slope all dynamically connected
* El Nino & La Nina events result from coupled
instability of the atmosphere/ocean system in the
Tropical Pacific
Bjerknes Hypothesis of coupled growth + equatorial ocean dynamics
* Among the fruits of the Bjerknes hypothesis, with
Wyrtki’s contribution…
ENSO events can be predicted
ENSO events have been predicted
The essence* of ENSO is understood
*The “linear essence” at least
Zebiak-Cane
Intermediate Coupled
Ocean-Atmosphere
Model
Atmosphere Part –
Low-level winds converge
towards warmest SSTa,
so atmospheric heating
(SH & LH fluxes) are
proportional to SSTa.
This effect is amplified in
regions where the mean
SST is warm (mean
convergence).
Zebiak-Cane
Intermediate Coupled
Ocean-Atmosphere
Model
Ocean Part –
Very simplified ocean model
(kind of like 2-layer fluid toy).
Ekman transport in surface
layer. Convergence or
divergence in surface layer
leads to changes in the depth
of the thermocline, which sits
at base of upper layer.
Temperature anomaly in the
sub-surface is determined
by depth of the thermocline.
First Successful [Documented] El Niño Prediction
After Cane, Zebiak and Dolan - Nature 1986 and see
Barnett, Graham, Cane, Zebiak, Dolan, O’Brien and Legler, Science 1988
Contours at 0.5°C
Going back, they were able to get 1982/83:
Going forward, they were able to get 1990/91 (neutral):
Going forward, they were able to get 1991/92 (El Niño):
Going forward, they were NOT able to get 1993:
Factors limiting the current skill of forecasts:
• Model flaws
• Flaws in the way the data is used
(data assimilation and initialization)
• Gaps in the observing system
• Inherent limits to predictability
Some periods appear to be
more predictable than others
Prediction accuracy decreases
at longer lead-times.
Chen, et al 2004 Nature
Example of
Inherent Limit to
Predictability
 Sensitivity to
Initial Conditions
°
C
°
C
°
C
Another Example
regarding inherent limits to predictability
(and somewhat model flaws also)
Evolution of the 2002-03 El Nino event
compared to the 1997-98 El Nino event
“Signal” versus “Noise” issues
ENSO is a slowly varying coupled ocean-atmosphere
phenomenon with a timescale of a year or longer.
Sub-seasonal weather acts rapidly on the coupled ocean-atmosphere
system with a time scale of weeks to months.
Eastward-propagating
convective anomaly
related to the MJO
(Madden-Julian Oscillation)
creates strong low-level
winds.
1996-1998 : Low Frequency
1996-1998 : High Frequency
2001-2003 : High Frequency
Even if a model has skill
ENSO Prediction is Not a Guarantee
El Niño
IMPACTS
“Expectations” of
climate anomalies
during El Niño events
Source: Ropelewski & Halpert, 1987 J. Climate
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/impacts/warm.gif
Monsoon Rainfall Index
Red = warm NINO3 SSTA - El Niño
Blue = cold NINO3 SSTA - La Niña
“Relative Frequency”
of Climate Impacts
(rainfall) due to
El Niño Events
Data & maps available through
IRI Data Library:
http://iridl.ldeo.columbia.edu/SOURCES/.IRI/.Analyses/.ENSO-RP
Drought & El Nino
Spatial Extent of Tropical Drought Correlated with El Niño
Note: 5 month lag between
max. NINO3.4 SSTA
and extent peaks
Source: B. Lyon, 2004, GRL
What is the indirect (i.e. through climate
teleconnections) impact of El Niño events
on CO2 variability?
Summary
• The basic ENSO mechanism is understood, and
can be predicted, but gaps remain
Role of MJO/WWBs, different “flavors” of ENSO,
decadal differences in predictability
• Prediction skill is limited by
Model flaws, data assimilation methods, limited data,
inherent limits to predictability
• ENSO events have global impacts
Many occur reliably,
but most are just more likely with an El Niño or La Niña event
• ENSO events impact CO2 variability
Extra Slides…
1. Climate Mean State (focus on tropics):
Annual Mean Solar Radiation
Annual Mean Heat Flux into the Ocean
Steric Height
relative to 2000m
From T, S data
at 1500m
at 0m
Temperature along the equator
Equatorial Undercurrent
SST Anomalies: Dec 1997
1997/98 El Niño
Economic “Cost” of El Nino
1982-83
Economic “Cost” of El Nino
1997-98
$14b USD : World Meteorological Organization
$36b USD : NOAA OGP (excluding ’98 China floods)
$45b USD : OFDA/CRED Int’l Disaster Database
Caution Must be Exercised when
Attributing a “Cost” to an “Event”
In the case of ENSO…
• Would the what is the baseline of ‘cost’?
Or, What
is the economic cost of
disasters ENSO-neutral years??
• Could the impact (or cost) have occurred
in the absence of the event?
DROUGHTS
Southern Africa
FLOODS
Peru
Southern California