Transcript The Role of the Basic State in Determining the Predictability of

The Role of the Basic State in
Determining the Predictability of
Tropical Rainfall
Andrew Turner, Pete Inness and Julia Slingo.
Talk Outline
•
•
•
•
•
Motivation.
Systematic errors in the UKMO climate model.
Flux adjustments used to correct the mean state.
Does the warmer mean state influence variability?
Future work.
Motivation
• South east Asian monsoon affects the lives
of more than 2 billion people.
• Could correcting systematic mean-state
errors in coupled GCMs improve the
simulated behaviour of the monsoon on all
timescales?
• Does this improve the prospects for
seasonal and climate change prediction?
UK Met Office Hadley Centre
model: HadCM3 L30
• HadCM3
- atmosphere: 3.75° x 2.5° x 30 levels,
- ocean: 1.25° x 1.25° x 20 levels,
- integrated for 100 years.
• ECMWF reanalyses ERA40 (1958-1997)
• CMAP (1979-1997) rainfall data (Xie &
Arkin 1997).
Simulation of summer mean climate
• HadCM3 simulates southwesterly monsoon flow but it is
too strong, adversely affected by
biases in the model.
• west Pacific warm pool confined
to Maritime continent and
equatorial central Pacific too cool
excessive Pacific trades
• Maritime Continent too warm
excessive westerly inflow
to the region from the Indian
Ocean via a Gill response (Gill,
1980)
Flux adjustment applied over 10°N-10°S in HadCM3
• First implemented
by Inness et al.
(2003) to study
MJO.
• Annual cycle of
ocean-surface heat
flux adjustments is
applied in the
tropical Pacific and
Indian oceans.
Improvements to the summer (JJAS) mean state
• Central equatorial
Pacific warmed.
• Flux adjustments
reduce westward bias
in the Pacific.
• Excessive trades are
reduced.
• Much better rainfall
picture over Indian
Ocean, Bay of Bengal
and Maritime
Continent; much more
like CMAP.
What effect on the Variability?
Stan. Dev.
HadCM3
HadCM3FA
ERA40
DMI Nino-3
1.22 0.94
2.06 1.21
1.60 0.85
Regression coefficients
(m/s per °C)
0.740
Response of Nino4 region HadCM3
10m zonal winds to Nino3 HadCM3FA 0.876
SSTs
ERA40
1.084
Stochastic Forcing
• El Nino excited by stochastic forcing
(Lengaigne et al. 2004, J. Climate,
submitted).
• More and stronger WWEs found in the flux
adjusted model, consistent with improved
MJO (Inness et al, 2003).
What effect on the monsoon-ENSO
teleconnection?
Larger ENSO magnitude
with flux adjustments,
coupled with stronger
trade wind response to
SSTs
Stronger monsoon-ENSO
teleconnection.
Flux adjustments also improve the timing of the Indian rainfallENSO teleconnection.
Summary
• Flux adjustments partially correct systematic biases
in HadCM3, giving monsoon and Pacific systems a
better mean.
• Monsoon much more variable due to stronger
Pacific variability and better wind response.
• Monsoon-ENSO teleconnection timed better.
• Greater stochastic forcing on intraseasonal
timescale contributes to broader ENSO periodicity.
• GCMs must have the correct basic state and the
right level of stochastic forcing in the coupled
system in order to accurately represent global
teleconnections.
Future Work
• Warmer mean state has consequences for
both climate and variability of monsoon
systems.
• Set-up and tune SPEEDY model in
Reading.
• Apply knowledge to results of greenhouse
gas climate change integrations of SPEEDY
and HadCM3.
References
• Codron et al. (2001) “Monsoon Dynamics: Predictability
of Monsoons” Journal of Climate 14.
• Gill (1980) “Some simple solutions for heat-induced
tropical circulation” Q.J. Roy. Met. Soc. 106.
• Inness et al. (2003) “Simulation of the Madden-Julien
Oscillation in a coupled GCM part II: the Role of the Basic
State” Journal of Climate 16.
• Lengaigne et al. (2003) “Coupled mechanisms involved in
the triggering of El Nino by a Westerly Wind Event”
submitted, Journal of Climate.
• Webster & Yang (1992) “Monsoon and ENSO: selectively
interactive systems” Q.J. Roy. Met. Soc. 118.
• Webster et al. (1998) “Monsoons: processes, predictability,
and prospects for prediction” J. Geophys. Res 103.