SSH anomalies from satellite

Observed annual mean state

Circulation creates equatorial cold tongues eastern Pacific

QuickTime™ and aTIFF (Uncompressed) decompressor

are needed to see this picture.

Trades -> Ocean upwelling along Equator -> Thermocline Tilt ->

Mean SST

East-West SST Gradient -> Walker Circulation

Positive (Bjerknes) feedback amplifies SST gradient and Walker Circulation

Normal conditions

El Nino onset conditions

El Nino conditions

ENSO Wind and SST

ENSO Ocean Temperatures

Sea surface temperature (SST) and zonal wind anomalies vary in a quasi-stationary fashion.

Thermocline anomalies along the equator show a systematic space and time evolution relative to SST anomalies.

1996

1999

1998

1997

Mean P Tahiti>P Darwin. Negative SOI is weakening of trades

SOI =10Pdiff −Pdiffav( )SD(Pdiff )

Pdiff =PTahiti −PDarwin

Precipitation Anomalies

Small changes in the distribution of sea surface temperature are coordinated with changes in atmospheric circulation and rainfall patterns;

Temperature Anomalies

Nino3.4 or CT

There is a tight coupling between the atmosphere & ocean.

Sea Surface Temperature and Sea Level Pressure

CT

SOI

r = 0.93

The spectrum of ENSO features a broad interannual (3-7 year) peak

ENSO has most variance at the end of the calendar year

Thompson and Battisti 2001

ENSO affects the global climate through atmosphere teleconnections

Upper level circulation

Anomalies during El Nino: the “warm phase” of ENSO

Zhang Battisi Wallace 1997

ENSO affects the global climate through atmosphere teleconnections

Climate Impacts of thePacific-North American Pattern (Winter)

Air Temperature Precipitation

+2°C- 2°C 0°C +1- 1 cm per month

The impacts of ENSO (cont).

• ENSO alters the Pacific storm tracks, and the probabilities of extreme weather events on a global scale.

State of the tropical Pacific today

QuickTime™ and aTIFF (Uncompressed) decompressor

are needed to see this picture.

State of the tropical Pacific today

QuickTime™ and aTIFF (Uncompressed) decompressor

are needed to see this picture.

State of the tropical Pacific today

QuickTime™ and aTIFF (Uncompressed) decompressor

are needed to see this picture.

1. Kelvin wave: travels as a first baroclinic mode gravity wave with speed sqrt( g’h) to the east, about 3 m/s to the EAST. Crosses the Pacific in about 2 months.

2. Rossby wave (first meridional mode) travels as a first baroclinic mode with phase speed equal to sqrt(g’h)/3, about 1 m/s to the WEST Cross the Pacific in 6 months. In oceanic Rossby waves, meridional advection of planetary vorticity is balanced by stretching not relative vorticity

Delayed Oscillator Theory: introduction to KW and RW

Zonal wind anomaly on the Equator (westerly)

Wind stress at 175W Wind stress curl at 175W

Delayed Oscillator Theory: a model of SSH evolution

Ocean sea surface Height anomaly(remember thermocline is opposite)

Ocean sea surface height anomaly along 140WDownwelling Kelvin Wave

Ocean sea surface height anomaly along 180E, upwellingRossby wave

Delayed Oscillator Theory: forcing of RW and KW

25 days

50 days

275 days

175 days

125 days

75 days

100 days

225 days

Delayed Oscillator Theory: evolution of RW and KW

ENSO Theories

1. El Nino is one phase of a self-sustained unstable and naturally oscillatory mode

2. El Nino is a stable (or damped) mode triggered by atmospheric random noise forcing

3. It is marginally stable, and that weather noise is needed to kick it off

ENSO Theories

1. Delayed-oscillator model2. Discharge Oscillator-pior to El Nino, warm water

volume builds up, then during El Nino is discharged to higher latitudes

3. There is a lagged feedback from the western boundary in the ocean

ENSO Theories

1. Some agreement that it is a marginally stable mode that needs noise to kick it off, most likely the Madden Julian Oscillation, an oscillatory pattern that travels the atmosphere around the globe, but can only be seen at the surface in the western Pacific. 40-50 day period and propagate to the east.

2. MJOs are not symmetric, with westerly winds bursts only occuring when the water is above 27 to 28C.

3. ENSO is not symmetric, with El Nino being understood as an event, and La Nina as “normal”, it is not sinuisoidal, although it is in simple models and in many coupled GCMs

Challenges

1. El Nino cannot be predicted before it starts, once it starts, then useful predictions can be made (9 months)

2. El Nino starting up seems to rely on atmospheric phenomenon like MJOs that are difficult to model in GCMs

Decadal ENSO-like variability (aka PDO)• After filtering to remove the interannual variability, the leading pattern of

variability in the Pacific has an ENSO-like pattern in SST and atmosphere circulation.

Decadal ENSO-like variability (aka PDO)

• The PDO spectrum is red• Due to stochastic forcing (a) the Aleutian Low (white) and (b) ENSO

teleconnections• Time scale is determined by surface ocean ocean heat capacity

(including re-emergence).