HR: 1340h
AN: NG23A-0091 [Abstracts]
TI: Toward a theory for millennial-scale climate variability through application of MEP in a simple
dynamical model
AU: Rial, J
EM: jar@email.unc.edu
AF: Department of Geological Sciences, University of North Carolina, Chapel, NC 27599
United States
AU: * Noone, D
EM: dcn@colorado.edu
AF: Cooperative Institute for Research in Environmental Sciences, University of Colorado
Campus Box 216, Boulder, CO 80309
United States
AU: Nordstrom, K
EM: knordstrom@comcast.net
AF: Cooperative Institute for Research in Environmental Sciences, University of Colorado
Campus Box 216, Boulder, CO 80309
United States
AU: Chase, T
EM: tchase@cires.colorado.edu
AF: Cooperative Institute for Research in Environmental Sciences, University of Colorado
Campus Box 216, Boulder, CO 80309
United States
AU: Gupta, V
EM: gupta@cires.colorado.edu
AF: Cooperative Institute for Research in Environmental Sciences, University of Colorado
Campus Box 216, Boulder, CO 80309
United States
AB:
Paleoclimate records contain a wealth of information about the character of climate variability and change. However,
interpretation of these records is confounded by the lack of both a sufficient theory of climate system behavior and
knowledge of those parameters that govern the character of the behavior. While one might argue that these concerns can be
largely overcome by employing complex models of the climate system, interpretation of such models requires first a context
for understanding simulation results. Further, it is yet to be demonstrated that indeed complex climate models can capture
the internal and forced variability seen in proxy records. Specifically, it is not known if the ability to capture this
variability has inadvertently been removed from the models range of possible behaviors due to the desire to tune
parameterizations to closely reproduce contemporary climate.
In view of these concerns, we use a simple model of the climate system that is sustained away from thermodynamic equilibrium
by an energy flow. This model predicts the global mean temperature of the ocean from energy balance. Prediction of sea ice
extent provides a nonlinear feedback the ocean temperature, thereby introducing into the model non-trivial internal
oscillatory behavior. Such a simple model necessarily has a small number of tunable parameters. With selection of reasonable
values for these parameters it has been shown to satisfactorily produce variability reminiscent of the Dansgaard-Oeschger
variability observed in Northern Hemisphere proxy records.
We take the view that the system should organize itself such that it will attain a regime that maximizes the production of
entropy (MEP). By defining a thermodynamic entropy and maximizing it with respect to our free parameters, we determine an
optimal parameter set. Here we examine how this selection differs from those estimated from physical arguments. In so doing,
we show how the estimation of parameters can be used to argue for new understanding of the underlying physics. When the
optimal selection process is repeated for the range of solar input experienced over recent glacial cycles, the time
variations in the optimal parameters leads to the diagnosis of how the model description of climate is incomplete.
Specifically we find at least one further component is required to adequately explain the variability over recent glacial
cycles, and we argue that this is associated with strong greenhouse coupling between global temperature, state of the carbon
cycle and atmospheric hydrology. While the behavior of our model is quite satisfying given its simplicity, it is the success
of our approach for using the concept of MEP to elucidate important but poorly known or understood physics that lends itself
to developing more complete understanding of complex systems.
DE: 3309 Climatology (1616, 1620, 3305, 4215, 8408)
DE: 4430 Complex systems
DE: 4485 Self-organization
DE: 4900 PALEOCEANOGRAPHY (0473, 3344)
DE: 4901 Abrupt/rapid climate change (1605)
SC: Nonlinear Geophysics [NG]
MN: Fall Meeting 2005