Transcript Climate Modelling in Australia Michael Manton Bureau of

Climate Modelling in Australia
Michael Manton
Bureau of Meteorology Research Centre
APN Symposium, 23 March 2004
Climate models capture the
complexity of the climate system
Why is climate modelling
important?
●
World Climate Research Programme
(WCRP) in 1980 recognised the climate
model as the tool to
–
Simulate climate system and its components
–
Test understanding of climate system
–
Combine observations in a consistent manner
–
Simulate past climate variations and changes
–
Predict future climate variations and changes
●
●
Australia has a long history of
involvement in climate
modelling
Universities
–
Macquarie – Land surface modelling
–
Melbourne – Southern hemisphere phenomena
–
Monash – Detection of climate change
–
NSW – Ocean modelling
–
Tasmania – Sea ice modelling
Government agencies
–
ANSTO – Isotopes & land surface
–
CSIRO – Weather and climate
–
BMRC – Weather and climate
There is substantial
collaboration between groups
●
Cooperative Research Centres
–
●
●
Collaborative projects
–
CSIRO & BMRC with AGO
–
CSIRO & BMRC with WA Government
Australian Academy of Science
–
●
Antarctic Climate & Ecosystems
NCESS workshop
Australian Research Council
–
Network on ESM
Some Examples
●
Model validation of land surface schemes
–
●
Use of isotopic data to validate models
–
●
ANSTO
Coupled modelling for inter-annual prediction
–
●
Macquarie University and BMRC
BMRC and CSIRO Marine Research
Coupled modelling for climate change
–
CSIRO Atmospheric Research
Surface Energy Complexity
Does it matter in climate models?
●
●
●
●
Macquarie University and BMRC
AMIP-2 result analysis
Using CHASM (captures various levels of
surface energy balance [SEB] complexity)
See Pitman et al., GRL, 2004, in press
Zonal differences in simulated temperature variance
No systematic differences:
SEB does not explain AMIP-2
differences
Colour = various modes of CHASM
Thick black line = observed
Thin black lines = AMIP-2 model results
Results give confidence in
climate model projections
of basic values
Maximum temperature variance
Most complex mode –
Includes tiling …
Tiling leads to significantly higher maximum temperatures
Results imply SEB complexity affects extreme values
AMIP2 Analysis
●
●
●
●
Prediction of
land surface
climate evolved
over time.
‘SiB’lings
others
No canopy
Not always
forwards
Schemes
capture a wide
range of
behaviours.
Not all schemes
equally good.
Henderson-Sellers et al. 2003
(Geophys. Res. Lett. 30,1777 )
Isotope model studies
●
●
Emerging area for model studies
Independent validation tool
ARC Linkage & other funding agencies
●
• Weakened signal at Manaus
means more water-recycling.
• Other indicators say more
non-fractionating sources.
18O
in 1960s
& 1980s
H-S, McGuffie & Zhang,
J. Clim., 2002, 15, 2664
ANSTO
POAMA
Predictive Ocean Atmosphere Model for Australia
●
●
●
●
●
●
Global coupled model GCM seasonal forecasting system
Joint project between BMRC and CSIRO Marine
Research
Partly funded by the Climate Variability in Agriculture
Program (CVAP)
Run in real-time by Bureau operational section since 1
October 2002
Operational products issued by the Bureau National
Climate Centre (NCC)
Experimental products available on the POAMA web site
www.bom.gov.au/bmrc/ocean/JAFOOS/POAMA
Introduction- POAMA operational system
Observing network
Obs/data Assimilation
Daily NWP
Atmos. IC
Model
Atmos. Model
T47 BAM (unified)
Real-time
ocean assimilation
latest ocean/
atmos obs
Coupler: OASIS
9-month forecast
once per day
Atmospheric
observations
Ocean
observations
Ocean assimilation
- Temp. OI every 3 days
+ current corrections
Forecast/products
Ocean Model
ACOM2 (~MOM2)
Ensemble
forecasts
Skill of SST Predictions
Hind-casts: one forecast per month, 1987-2001 (180 cases)
Anomaly correlation
Green - model,
red - anomaly persistence
2 months
Anomaly
Correlation
4 months
6 months
Decay of 2002 El Nino
POAMA Real-time forecasts correctly predict decay
Prepared P. Reid NCC
Sample OLR intra-seasonal forecast from POAMA-1
5-member ensemble starting 10 Dec 2003
Days 1-5
Days 6-10
MJO
Days 11-15
Days 16-20
Days 20-30
Days 30-40
CSIRO Atmospheric Research
CSIRO Mark 3 model
•
•
•
•
•
•
3-dimensional global model
18 levels in atmosphere
31 levels in ocean including sea-ice
6 soil levels, 9 soil types, 13 vegetation types
3 snow levels
180 km between grid-points (100 km in tropics to better simulate El
Nino)
• Data for 100 climate variables computed in 30-minute time-steps
for a series of months, years decades or centuries
• Models adequately simulate observed daily weather and average
climate patterns
• A one-year simulation takes 1 day of computer time
Improved simulation of El Nino
Southern Oscillation
Observed sea
surface
temperature
anomaly
CSIRO
Mark 2
model
CSIRO
Mark 3
model
CSIRO Mark 3 simulation 1870-2020+
Global surface air temperature change
Model hierarchy
Complex
Simple
Global climate model
(grid: 180 km by 180 km)
PC software, e.g.
MAGICC, OzClim
Regional climate model
(grid: e.g. 70 km by 70 km)
Regional climate model
Statistical downscaling
(grid: e.g. 14 km by 14 km)
(local sites: e.g. Perth)
Modelled and Observed Mean Winter
and Spring Rainfall, years 1961-1975
CSIRO Cubic Conformal Atmospheric Model – stretched grid
OzClim PC software
Database includes:
Observed and simulated
monthly-average data on 25
km grid
10 climate models
6 IPCC emission scenarios
3 climate sensitivities
9 climate variables
Functions:
Plot maps and global
warming curves
Save regional average data
Run simple impact models
Package is used for impact studies and education
Land
surface
Ocean
Ocean
New components developed
and tested separately, then
coupled in the model and
tested again
IPCC 2001
Future Directions
●
●
●
●
●
●
●
Enhanced complexity
Improved parameterisations
Improved representation of external forcings
Improved understanding of predictability
Analysis of extreme events
Use of ensembles to represent uncertainty
Coupling of economic and climate models