Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

The Madden-Julian Oscillation (MJO) : Michelle L'Heureux NOAA Climate Prediction Center (CPC)

Download as pdf or txt
Download as pdf or txt
You are on page 1of 41

The Madden-Julian Oscillation (MJO)

Michelle LHeureux NOAA Climate Prediction Center (CPC)

Adapted from the presentation by Jon Gottschalck, NOAA CPC

Outline
What is the MJO?: General characteristics and structure What causes the MJO?: Theories How do we monitor the MJO?: Tools, methods and data

MJO Impacts: Tropics and mid-latitudes


MJO Prediction: Forecast tools and skill MJO Products at NOAA CPC: CPC weekly MJO update, Global Tropics Hazards/Benefits Assessment

PART 1: What is the MJO?

MJO Characteristics
The MJO is an intraseasonal or subseasonal fluctuation originating in the tropics

The MJO results in changes in atmospheric and oceanic variables


Lower- and upper-level wind Cloudiness and tropical rainfall Sea level Pressure Sea surface temperature (SST) Ocean surface evaporation Ocean chlorophyll

Typical period of the MJO cycle is approximately 30-60 days Acts on a global spatial scale Coherent eastward propagation (Eastern Hemisphere: 5 m/s and Western Hemisphere 15 m/s)
Madden and Julian, 1971; 1972; Zhang, 2005; Hendon and Salby, 1994

MJO Characteristics
The MJO is most active during the boreal late fall, winter and early spring periods
Hendon et al. 1999; Zhang and Gottschalck 2002; Zhang 2005

The MJO tends to be most active during ENSO neutral years


Often absent during strong El Nio events

Outgoing Longwave Radiation (OLR)

Matthews, 2000

MJO Structure
8 phases altering patterns of:
Tropical Rainfall (cloud) Winds (green arrows) Sea level pressure (red/blue shading)

Convective signal strongest across the Indian Ocean, Indonesia and Western Pacific
Circulation signal transverses the globe

Schematic of Three-Dimensional MJO Structure

The MJO in the Real World


Previous figure is an idealized depiction
In reality, the characteristics differ considerably among different MJO events The strength, duration and propagation speed vary from event to event

Not all elements of the MJO -- convection, zonal wind, moisture convergence, and SST anomalies -- are clearly visible for each event Only when averaging many MJO events together does a less noisy, idealized picture of the MJO emerge

Composite of the MJO Lifecycle


Tropical Rainfall Green areas
Increased rainfall Enhanced phase of the MJO

Brown areas
Decreased rainfall, Suppressed phase of the MJO

Eastward propagation of each phase (dashed lines)

mm/day

PART 2: What causes the MJO?

MJO Mechanisms
The mechanisms for MJO development, maintenance and decay are not well understood

Focus of considerable ongoing research (diagnostics and modeling)


Initiation is likely a result of complex interactions (1) between atmosphere and ocean (2) across multiple temporal and spatial scales (3) across regional boundaries

No one theory completely explains the mechanisms for the period, zonal scale and eastward propagation of the MJO

Zhang, 2005

MJO Mechanisms
Two main hypotheses:
(1) Independent external forcing
Stationary intraseasonal oscillations in the Asian-Australian monsoon region excite the MJO Stochastic forcing Extratropical disturbances interact with the Tropics to force the MJO

(2) Internal MJO mechanisms


Complex air-sea interactions initiate and maintain propagation Varying temporal and spatial scales of convection interact and organize convection on a larger spatial scale (i.e., the MJO)
Zhang, 2005

PART 3: Monitoring the MJO

Monitoring the MJO -- OLR Anomalies


MJO activity December 2009 January 2010
Blue shades Enhanced convection Increased rainfall

Red/Yellow shades Suppressed convection Decreased rainfall

Monitoring the MJO -- Time-Longitude Sections


Eastward Propagation

Indian Indonesia Western Ocean Pacific

MJO Propagation as an Envelope


MJO enhanced phase:
Composed of convective elements acting on multiple spatial and temporal scales Both eastward (red) and westward (green) moving components

Blue ovals highlight the MJO intraseasonal convection through time-space filtering

Envelope of all of these convective elements (large and small, fast or slow) is what travels eastward with time (black dashed line)

Monitoring the MJO 850-hPa Zonal Wind

MJO event during late 2007 Boxes indicate enhanced westerly (red) and easterly (blue) anomalies and low-level convergence (black circle) when the enhanced convective phase of the MJO is centered over the eastern Indian Ocean

Monitoring the MJO 200-hPa Zonal Wind


A

C
C A

Anticyclonic (A) and cyclonic (C) circulations straddling anomalous tropical convection Westerly anomalies across the eastern equatorial Pacific Ocean Enhanced phase of the MJO (black circle) centered across Indonesia

200-hPa Velocity Potential Animation


Spring 2005 MJO Event

Wheeler-Hendon MJO Index


The axes (RMM1 and RMM2) represent daily

values of the PCs from two leading modes The triangular areas indicate the location of the MJO enhanced convective phase Counter-clockwise motion is indicative of eastward propagation. Large dot is the most recent day.

Weak MJO

Distance from center indicates strength

Strong MJO

--based on combined EOF analysis (OLR, 850 and 200 hPa zonal wind)
Wheeler and Hendon, 2004

Interannual Variability of the MJO


There is strong year to year variability in MJO activity.
Grey areas are active MJO periods.

Courtesy: Australia CAWCR

http://cawcr.gov.au/staff/mwheeler/maproom/RMM/ts.PCvar91drm.gif

Seasonal Cycle in the MJO

The MJO variability is greatest during the Northern Hemisphere winter (Jan.- Mar.) and is weakest during the summer (June-Aug.)

Courtesy: Australia CAWCR

http://cawcr.gov.au/staff/mwheeler/maproom/RMM/ts.PCvar.seasonal cycle.gif

PART 4: MJO Impacts

MJO Impacts are Global


The MJO affects both tropical and extratropical areas The MJO substantially impacts tropical rainfall patterns

While not shown here, keep in mind that the MJO also influences tropical cyclone activity in all ocean basins, monsoon systems, and can affect the evolution of the ENSO cycle.

MJO Impacts Tropical Rainfall Hotspots


Boreal Winter

Boreal Summer

Rainfall during Northern Hemisphere Summer


Phase 2

Over Central/ S. America/Caribbean: Enhanced rainfall during Phases 7, 8, & 1. Suppressed rainfall during Phases 3-6.

Over Africa: Enhanced rainfall during Phase 2. Suppressed rainfall during Phases 6-8.
Phase 3

Phase 4

Phase 5 Phase 6

Over SE Asia/Philippines: Enhanced rainfall during Phase 4-6. Suppressed rainfall during Phases 8, 1-2.

Phase 7 Tropical anomalies are strongest just north of the equator during the Northern Hemisphere Summer (wrapped up in monsoon systems) Phase 8

Phase 1

Rainfall during Northern Hemisphere Winter


Phase 2

Over South America: Enhanced rainfall during Phases 8 & 1-2 Suppressed rainfall during Phases 4-6. Over SE Asia/Philippines: Enhanced rainfall during Phase 4-5. Suppressed rainfall during Phases 7-8, 1-2.

Over Africa: Enhanced rainfall during Phases 2-3. Suppressed rainfall during Phases 5-6.

Phase 3

Phase 4

Phase 5 Phase 6

Phase 7 Tropical anomalies are strongest just south of the equator during the Northern Hemisphere Winter (wrapped up in monsoon systems) Phase 8

Phase 1

MJO Impacts Extratropical Circulation


200 mb winds and OLR Deep persistent
tropical convection can excite Rossby Wave Trains (RWTs)
RWTs can enhance or dampen ongoing circulation patterns

H L L H H

Since the overall weather pattern the MJO event is occurring in is


different, the tangible synoptic results will be different in each case
The RWT forcing moves eastward as the MJO shifts eastward
Kiladis and Weickmann, 1992; Renwick and Revell, 1999

MJO Impacts Pacific/North American Region


Negative PNA & Positive AO/NAO Negative AO/NAO

Riddle et al. (in press, Climate Dynamics)

Average 500-hPa geopotential height anomalies based on two cluster patterns (out of 7- not shown)
These patterns have a close relationship with the propagation of the MJO during Dec.-Mar.
AO/NAO result is consistent with analysis by Cassou (2008), LHeureux and Higgins (2008), Lin et al. (2009)

Initial condition:

Phase 1 Phase 2 Phase 3 Phase 4 Phase 5 Phase 6 Phase 7 Orange color shows % frequency of the pattern above climatology Green shows % frequency of the pattern less than climatology
Climo is all DecemberMarch days from 1979-2010
Colors are 95% significant based on Monte Carlo Markov Chain test

Phase 8
x-axis: Days (0-40) after MJO Phase X

PART 5: MJO Prediction

MJO Prediction -- Overview


The MJO is predictable approximately out to 2-3 weeks depending on the model.
Much of this skill comes from continuation of established ongoing events Transitions (MJO onset/demise) are not captured well at present Both statistical and dynamical model predictions are available Only recently have dynamical predictions reached and in several cases surpassed statistical predictions

MJO Forecast Examples

NCPE

CANM

UKME

Differences in: (1) Ensemble spread (2) Propagation speed (3) Amplitude ECMF BOME
Gottschalck et al. 2010

Dynamical + Statistical Model Forecasts


Initial conditions based on the 1st of each month from 1980-2006

Useful skill (r > 0.5): Persistence: 8 days Simple statistical: 15 days POAMA model: 16/19 days One of the first studies to show comparable skill between dynamical and statistical models

model statistical persistence

Statistical MJO forecasts from several different methods (lagged regression, constructed analog, etc.) are generally comparable Rashid et al. 2010

Dynamical Model Forecasts


Forecasts initialized from Jan. 1st, 2008 through Dec. 31st, 2010 (3 years) Using operational medium range ensemble forecasts provided though TIGGE/THORPEX. Correlation skill in excess of 0.5 is present in top models out to 9+ days. Top models are ECMWF, UKMO, NCEP, JMA, CMC.

Matsueda and Endo (2011)

MJO Prediction Dynamical Models


Model forecasts of MJO index are best when the MJO shifts from the Pacific into the western hemisphere (Phases 5-8) Model forecasts of the MJO index have problems when enhanced convection shifts across the Indian Ocean into the Maritime continent (Phases 2-4) Verification is best when wind component of MJO signal is strong

Panel on left shows anomaly correlation based on GEFS ensemble mean as a function of MJO phase in the initial condition

PART 6: MJO products at NOAA CPC

CPC MJO Weekly Update


http://www.cpc.ncep.noaa.gov/products/precip/CWlink/MJO/mjoupdate.pdf

--Review of weekly changes in the MJO


--Anticipated evolution and impacts of the MJO during the next 1-2 weeks --Released every Monday ~4 PM --Global Tropical Hazard Outlook released on Tuesday ~4PM
Comments/Questions: Jon.Gottschalck@noaa.gov

Global Tropical Hazards Outlook

Dynamical Model Forecasts


http://www.cpc.ncep.noaa.gov/products/precip/CWlink/MJO/CLIVAR/clivar_wh.shtml

Gottschalck et al. 2010

Primary Information Sources


CPC MJO webpage (monitoring and predictions):
http://www.cpc.ncep.noaa.gov/products/precip/CWlink/MJO/mjo.shtml

CPC weekly MJO update:


http://www.cpc.ncep.noaa.gov/products/precip/CWlink/MJO/mjoupdate.pdf

CPC Global Tropics Hazards and Benefits Assessment:


http://www.cpc.ncep.noaa.gov/products/precip/CWlink/ghazards/index.php

Australian Bureau of Meteorology MJO webpage:


http://cawcr.gov.au/staff/mwheeler/maproom/RMM/

Questions?

You might also like