HR: 1340h
AN: NG23E-0123 [Abstracts]
TI: An Approach for Understanding and Comparing the Dynamics of the Climate System and Climate
Models
AU: * Bhatt, U S
EM: bhatt@gi.alaska.edu
AF: Geophysical Institute, University of Alaska Fairbanks, 903 Koyukuk Dr., Fairbanks, AK 99775-7320
United States
AU: Newman, D E
EM: ffden@uaf.edu
AF: Geophysical Institute/Physics Department, University of Alaska Fairbanks, Physics Department
P.O. Box 755920, Fairbanks, AK 99775-5920
United States
AU: Polyakov, I V
EM: igor@iarc.uaf.edu
AF: IARC, University of Alaska Fairbanks, 900 Koyukuk Dr., Fairbanks, AK 99775
United States
AU: Wackerbauer, R
EM: ffraw1@uaf.edu
AF: Physics Department, University of Alaska Fairbanks, Physics Department
P.O. Box 755920, Fairbanks, AK 99775-5920
United States
AB:
The underlying goal of much of the work in the geosciences is to develop predictive models of the fundamentally complex
behaviors in the system of interest. Due to concerns over the impact of climate change (anthropogenic or natural), it is
critical to validate and verify both the physics going into the climate models and the dynamics coming out of these models.
However, it is difficult to do so with the increasingly complex models (see Roache,1998) of these systems (e.g., Global
Climate Models or GCMs). This verification and validation is of fundamental importance if one is to confidently use such
models for prediction. It is a daunting task to simply evaluate complex climate models, but it is even more difficult to
determine the cause of a particular model behavior. In order to further our understanding of the important underlying
physical mechanisms needed to model these complex natural systems, we apply non-linear dynamical analysis techniques (Hurst,
1951; Alekseev and Yakobson, 1981; Kantz and Schreiber, 1997) to investigate characteristics of both the real climate system
and existing climate models.
In this poster, we will present preliminary analysis of the dynamical characteristics of real climate data, utilizing a
variety of mathematical non-linear dynamical techniques from statistical physics, scaling theory, and information theory.
Here we use dynamics to denote the temporal evolution and the physics as the mathematical description of the physical
mechanisms underlying the system. These terms are defined in order to avoid any confusion because, in climate modeling
parlance, `physics' refers to sub-grid scale processes that are parameterized. These analysis techniques allow us to
characterize the dynamics of these systems with various measures. These same techniques are then used to analyze the results
from both simple models and the full primitive equation GCMs with various levels of completeness (ie fixed SSTs vs dynamical
oceans) to determine whether they have similar dynamical characteristics. This will allow us to quantify the strengths and
weaknesses of the model in a way which is complementary to, and as important as, the standard statistical analysis. In many
cases, dynamical differences are found even when many standard statistical measures are the same. The goal of this work is
finding techniques that characterize the dynamics of observed climate data, elucidate the key underlying physical mechanisms
and identify strengths and shortcomings in GCMs, and provide suggestions for reducing the limitations of GCMs.
DE: 1620 Climate dynamics (0429, 3309)
DE: 1626 Global climate models (3337, 4928)
DE: 3270 Time series analysis (1872, 4277, 4475)
DE: 3333 Model calibration (1846)
DE: 4430 Complex systems
SC: Nonlinear Geophysics [NG]
MN: Fall Meeting 2005