Global climate modeling:

GCM = general circulation model.

What are the objectives of the GCMs?

How do they work?

General characteristics of GCMs:

• All the models work by defining elements and laws for how a change in one element will effect its neighbor. Initial conditions are defined and then element interactions are computed according to the laws.
• Typically 10 vertical levels. Total number of grid elements modeling is10exp4 to 10exp6.
• Remember that there will be multiple parameters and multiple calculations for each grid element.
• Natural the finer the elements the better resolution your model will have, but the number of calculations and the computer power and speed you need increase by a power law function.

Three major grid type components: atmospheric, oceanic, land.

For an atmospheric element - what will be important properties describing its state?

• temperature
• density
• moisture and heat content
• composition
• thermal properties

For an oceanic element - what will be its important properties describing its state?

• wind stress at surface (output of atmospheric component).
• salinity.
• temperature - this in turn is a function of oceanic circulation patterns (another module).
• thermal properties.
• sea ice models - effects thermal exhange, albedo, water density, perhaps a separate element type.
• turbidity.

For a land element - what will be its important properties describing its state?

• albedo (with a vegetative component or snow/ice component that changes seasonally)
• mean elevation.
• soil moisture.
• thermal properties.
• surface roughness.

What will be the laws for how neighbors effect each other?

• Newton's laws of motion: movement in response to pressure gradients both in the atmosphere and the hydrosphere..
• heat exchange/absorbtion components, e.g. between adjacent atmospheric components and between land and water surfaces and atmospheric component (clouds and moisture very important).
• e.g laws for formation of sea-ice important.

What is input?

• Take into account orbital/seasonal factors for solar influx.
• Average heat exchange characteristics of surface elements.
• Changes in atmospheric composition.
• Changes in vegetative cover.

What will be output of computer?

• temperatures.
• wind patterns.
• precipitation?
• changes in snow cover, ice cover.
• many possibilities.

Challenges:

• Cloud parameterization: mismatch of scale, clouds produce important fluxes of mass, momentum, heat and moisture, and yet they occur at scale much smaller than grid elements of computer model.
• Linking oceanic circulation with atmospheric models.

How can you test how good the computer model is (the discussion question)?

Some thoughts on chaotic systems and projections.

• Lorenz equations - initial conditions sensitivity
• Strange attractors - and general predictability.
• Might not be able to predict specific changes in specific areas.

How accurate are the computer models?

How accurate are they in comparison to any competing methodology?

Discussion question for next time. What factors will maximize the climatic effect of a given eruption, and why are the demonstrable effects so short lived? Can they be ignored in long term climate predictions?