MEDIA RELEASE
Climate change scientists meet in Melbourne
Computer models of Greenhouse the hot topic
Scientists from Europe, the United States and Japan will join Bureau staff to discuss how to improve the accuracy of "coupled" climate models in preparation for the next phase of deliberations on global climate change. The next report of the Intergovernmental Panel on Climate Change (IPCC) is due in 2001.
The coupled climate models have over the past 10 years played a significant role in shaping the outcomes of United Nations climate change negotiations, such as at the Earth Summit at Rio de Janeiro and last year's Session of the Conference of the Parties to the Framework Convention on Climate Change in Kyoto, Japan.
The modelling scenarios were particularly influential in the lead-up to Kyoto, when the IPCC concluded: "the balance of evidence suggests a discernible human influence on global climate". Coupled models have also been important in simulating the response of the climate to various measures aimed at reducing greenhouse gas emissions.
Coupled climate models, including the one developed by the Bureau of Meteorology Research Centre (BMRC), combine separate models of the atmosphere, the ocean, sea ice and the land surface to predict climate change.
Over the 10 days of the conference, three groups will meet in Melbourne:
The Working Group on Coupled Modelling chaired by Dr Lennart Bengtsson of
the Max Planck Institute for Meteorology (Germany) will meet from Friday
October 16. The Group comprises 15 leading researchers in the field,
including the BMRC's Dr Bryant McAvaney.
The Coupled Modelling Intercomparison Project will also meet for two days
to compare models; to weigh one against another and against measured
observations. Led by Dr Jerry Meehl (USA) from Wednesday October 14, it
will look at how climate models are performing.
The two-day Coupled Modelling Workshop run by the BMRC from Monday
October 12 will involve conference participants, BMRC staff, and other
scientists who will brainstorm on overcoming the challenges of combining
the four components of a coupled climate model.
For more information, contact
Dr Bryant McAvaney, BMRC: telephone (03) 9669 4134 or email
b.mcavaney@bom.gov.au
Phil Meighen, BMRC: telephone (03) 9669 4374 or email p.meighen@bom.gov.au
Notes:
The Intergovernmental Panel on Climate Change provides objective
assessments of scientific and technical information on climate change for
use by policy makers. It was established by the UN Environment Programme
(UNEP) and the World Meteorological Organization in 1988.
Background:
"The climate is a very, very complex beast, and certainly if you try to
model climate you are working with one of the most complex systems that
exists." - Dr Bryant McAvaney, BMRC
Modelling the earth's climate involves feeding weather observations into the various components of a coupled climate model to produce mathematical representations of the atmosphere, the oceans, sea ice and the land surface. The components are then coupled, and from this come predictions on changes to climate.
The BMRC model in use today is made up of 30,000 to 40,000 lines of code and is fed with data as diverse as atmospheric and ocean temperatures, wind speed, atmospheric pressure, radiation, humidity, cloud cover and soil temperature. To build a mathematical picture of the atmosphere, for instance, information from grids of 80 square kilometres in 29 levels up to 30 kilometres into the stratosphere are taken.
The BMRC coupled model is started or "initiated" by the ocean model which has data from 1000 years ago - up to 8000 years ago for the deep ocean - until it reaches a state close to equilibrium. The atmospheric, sea-ice and land surface models are started with data from about five years ago. The four systems are then coupled and allowed to run out for 20 to 30 model years before the climate change is determined.
Challenges for coupled climate models:
"Even if you had a model which was perfect against the current climate, it
doesn't mean it will work in the future." - Dr Bryant McAvaney, BMRC
How coupled climate models represent clouds and take account of the way they block radiant energy from the sun will be one of the subjects for discussion at the Melbourne conference.
Also on the agenda will be how to improve ocean and sea-ice models which rely on a data base that is not as refined as that of the atmosphere.
Of particular interest will be the various methods by which models are
initiated and the way an accurate representation of the ocean/atmosphere
interface is maintained as the model runs.
"The initial state of the ocean is problematic," Dr McAvaney says. "Very
small changes can have a large impact because of the interaction between
systems. Sometimes a small part amplifies up very rapidly; sometimes it
doesn't."