Title: The history and key results of the Indian Ocean Climate Initiative (IOCI)
1The history and key results of the Indian Ocean
Climate Initiative (IOCI)
Pandora Hope and IOCI Research PartnersBureau
of Meteorology Research Centre
8ICSHMO, Foz do Iguaçu, Brazil April, 2006
2Contributions From
IOCI Research Partners Bureau of Meteorology and
CSIRO Jorgen Frederiksen, Lynda Chambers,
Bertrand Timbal, Wasyl Drosdowsky, Scott Power,
Fuqin Li, Zhihong Li, Brian Ryan, Ian Smith,
Wenju Cai, Steve Charles, Bryson Bates Brian
Sadler and IOCI Partners particularly Ian
Foster and Colin Terry Acknowledgements National
Climate CentreA host of others
3IOCI Partners
Department of Agriculture WA Conservation
and Land Management Environment Planning and
Infrastructure Premier and Cabinet State
Greenhouse Strategy State Water Strategy Fire
and Emergency Services AuthorityForest Products
CommissionWater Corporation of WA Bureau of
Meteorology, WA
4Outline
IOCI Background Motivation Brief history and
possible future Synoptic Types Method Links with
rainfall Shifts through time IOCI Key messages
5The most reliable rainfall occurs in the extreme
south-west between Perth and Leeuwin. Here the
rains rarely vary 10 per cent. from their average
amount, and the lot of the farmer should be a
happy one.Griffith Taylor, Australian
Meteorology 1920
http//www.bom.gov.au/cgi-bin/silo/reg/cli_chg/ave
ragemaps.cgi
6http//www.bom.gov.au/silo/products/cli_chg/index.
shtml
7Average monthly rainfall SW WA
8May-July SW WA Rainfall
9IOCI History and possible future
- Stage 1 commenced January 1998
- Understand climate variability
- Seasonal forecasting targeted forecasts
- Reports, 1999, 2000, 2001 Final report 2002
- Stage 2 commenced July 20033 phases, the first
two have been completed - 3 Themes
- Current Climate Regimes
- Climate Change
- Short-term Climate Prediction
- 2004 Report, Flyers, Workshop
- Stage 3 Negotiating to commence July 2006
- www.ioci.org.au
10May-July MSL Pressure Anomaly(1976-2003 minus
1958-1975)
11Vertical cross-section of the Southern Hemisphere
July zonal wind (ms-1) averaged over 100E-130E
1949-1968 1975-1994 difference
12Development of Self-Organising Map
- Select size and topology of SOM
- Select size of data region
- Select neighbourhood function and radius
- Start with SOM of random numbers
- Present data to SOM
- Vector similarity is found via Euclidean distance
- SOM is trained
- Data is presented again until error is low
From Kohonen 2001
13Self Organising Map NCEP/NCAR 1948-2003 June
July 0 12Z MSLP
14Transitions
Forward transition
15Self Organising Maps associated rainfall
anomalies
Wet types
Trough to west types
Pandora Hope BMRC
16Frequency of wet synoptic patterns ( rainfall)
1976-2003 average 27 synoptic trough types, 199
mm rainfall
1958-1975 average 33 synoptic trough types, 241
mm rainfall
Decrease in frequency of wet types accounts
for 50 of rainfall decrease
Decrease in rainfall associated with troughs
to west types accounts for 30
17Results from 20th Century synoptic study
- Wet synoptic types have decreased
- Dry types have increased, to a lesser extent
- Rain linked with northerly flow types has
decreased - Increase in rainfall linked to southerly flow
types - Average south west rainfall links well with wet
synoptic types
18 Difference 2046-2065 cf 1961-2000
MRI-CGCM2.3.2
19SRES B1 2081-2100 Difference
CSIRO-Mk3
GFDL-CM2.0
GISS-ER
MIROC3.2(med)
20SRES A2 2081-2100 Difference
CSIRO-Mk3
GFDL-CM2.0
GISS-ER
MIROC3.2(med)
21Synoptic types in future scenarios
- Increasing levels of greenhouse gases lead to a
growing reduction in the occurrence of deep
troughs in the SWWA region in most GCMs. - All GCMs examined here show a decline in direct
model output rainfall with increasing greenhouse
gas concentrations. - These results suggest that, as greenhouse gas
concentrations rise in the future, it is likely
that rainfall in SWWA will continue to be low.
However, decadal variability must also be
considered.
22Key Messages from IOCI, Phase 1
- IOCI (2002) concluded that Most likely, both
natural variability and the enhanced greenhouse
effect have contributed to the rainfall decrease
and that Other local factors, such as land-use
changes in the southwestseem unlikely to be
major factors in the rainfall decrease, but may
be secondary contributors.
23Key Messages from IOCI, Phase 2
- The large-scale synoptic changes are consistent
with the drying in SWWA, providing further
evidence that the rainfall decline is linked with
large-scale global circulation changes - It is feasible that the drying trend could have
been the result of unforced climate variability - However, at least part of the observed decline in
rainfall is consistent with the modelled effect
of anthropogenic forcing - Changes in land cover may also have contributed
to the rainfall decline. - The latest set of CSIRO control and transient
simulations confirm both natural variability and
greenhouse gas radiative forcing are
24Key Messages from IOCI, Phase 2
- The latest set of CSIRO control and transient
simulations confirm both natural variability and
greenhouse gas radiative forcings contributed to
the rainfall decline - 20th century simulations generally underestimate
the observed rainfall decline - The latest suite of climate change simulations
show that even with the most optimistic emission
scenarios, SWWA is projected to be drier and
warmer later this century.
25Thank You
For further details about IOCI and its research,
please speak to Pandora Hope, or visit the
website www.ioci.org.au
Pandora Hope and IOCI Research PartnersBureau of
Meteorology Research Centre
8ICSHMO, Foz do Iguaçu, Brazil April, 2006