Madison Chaos and Complex Systems Seminar

Fall 2004 Seminars

All seminars are Tuesday at 12:05 pm in 4274 Chamberlin except as noted.

Short List


September 7, 2004

Chaos microwave transmitters for advanced communications and radar

John H. Booske, UW Department of Electrical and Computer Engineering

Chaos has been a subject of significant scientific investigation for several decades.  Chaotic systems are characterized by sensitive dependence on initial conditions and random-like behavior.  However, chaotic systems are governed by deterministic relationships and therefore order is hidden within the “randomness”.  Two consequences of this determinism are: (1) the statistics of chaotic systems can be different than the statistics describing random events, and (2) two similar chaotic systems can be synchronized to exhibit the same chaos.  These features have inspired recent initiatives to extract benefits from chaos (rather than suppressing it) in electrical engineering applications such as communications and radar. The vision perceives possibilities for improved system performance along with immunity from eavesdropping or jamming. The power requirements of these applications motivate investigations into high power, wideband, microwave amplifiers with delayed feedback (e.g., traveling wave tubes, TWTs) as candidate chaotic transmitters.  This talk will describe preliminary results of an exploratory research investigation of chaotic TWTs.  The initial goal is to develop a basic understanding of the dynamics of a driven, nonlinear TWT “oscillator”.  Theoretical models suggest that it should behave similarly to a driven van der Pols oscillator system, exhibiting both quasi-periodic and period doubling routes to chaos.  Experimental observations qualitatively confirm a number of the predictions although, to date, we have only observed the period doubling route to chaos.  Further improvements to our basic understanding of the system are required.  The second, longer-term, goal is to quantitatively assess the prospects of chaotic TWT transmitters for applications in advanced communications and radar.

September 14, 2004

Modelling human brain development and disease using stem cells: a role for chaos?

Clive Svendsen, UW Department of Anatomy

Arguably, the human brain is one of the most complex systems in the universe.  Recently, individual human stem cells have been isolated and grown in the culture dish.  They seem to be able to hold on to memories of where they came from even after expansion - although the mechanism for this remains a mystery.  They have also shown to form complex networks of brain cells in the culture dish.   Expressing specific genes can modify their development and simulate specific disorders such as Down syndrome and Parkinson's disease.  With no knowledge of chaos theory whatsoever (perhaps a distinct advantage) - I would like to propose that chaos may play some role in this process and would like the audience to explain how this may happen.

September 21, 2004

Human vision as a complex dynamical system: A case study using magnetic encephalography

Amir H. Assadi, UW Department of Mathematics & The Wisconsin Genome Center

The human brain is one of the most studied complex dynamical systems. Visual perception, among the vast array of functions of the brain, has played a prominent role with an illustrious history of contributions by scientists, mathematicians, and philosophers. While applications of dynamical systems theory to research in neuroscience have enjoyed great appeal with many wonderful contributions, the subject matter appears merely to be scratched on the surface! In this lecture, I will present progress report in my research collaborations on modeling visual object recognition (which is also a well-studied topic in vision!) My emphasis, however, will be on the dynamics of the neurophysiologic processes while the brain performs object recognition and categorization. I will try to follow lessons that are learned from investigation of the physics of complex dynamical systems. That is, observations of the dynamics of the system are recorded through precise physical measurements, and the problems are formulated in terms of extraction of regularities in observable quantities. The main mathematical tools are taken from analysis of massive data sets that could help us in probabilistic and statistical formulation of general physical principles and predictions of the system’s behavior. Recent interdisciplinary research by neuroscientists, physicists, and engineers has been tackling the challenges of this approach, with remarkable success stories.  I will report on a joint ongoing collaboration with the Vision Lab at Harvard in applying Magnetic Encephalography (MEG) to problems of higher visual processing, such as recognition of human faces and other objects.

September 28, 2004

Whence the climate of our Earth,

C. S. Clay, UW Department of Geology and Geophysics

Our climate is not simple. Our Madison winters are warmer. Our Madison summers are cooler. The Arctic Ocean is melting and has half as much ice as it did 50 years ago. Droughts cover areas of our west. People say that we are in "Global warming". About 50 years ago, a pair of crusty geologists said that the Arctic ocean may have been ice free during the last ice age.

Chunks of data are in front of us:

1) The duration of a frozen Lake Mendota is a fairly robust measure of Madison's winter climate (12). The freeze dates and ice breakup dates have been recorded since 1851. Since 1970, the freeze dates have gotten later and the breakup dates are earlier.

2) Jon Berkson, Tze-Kong Kan and Clay took their ice imaging sonars to Alaska. The research was to image the underside of sea ice in the Arctic. From Point Barrow, they were flown to the old ice island, T3, about 300 nm from the North Pole. The temperature was a steady -30 C. The one year old leads in the ice were 2 to 3 m thick. Now, 3 decades later, the Arctic sea ice is less than a meter thick and there is open water. Something is happening here.

3) Long temperature history is the series of the oxygen isotope ratios for 800 thousand years, Fig. 1. The lighter isotope (16) evaporates faster than the heaver (18) and the ratios indicate the amount of ice on land. The peaks are spaced at roughly 100 thousand years. Now, the ratio and temperature are at the top and warm of the fluctuations. Simple extrapolations or extensions of the current trend predict "Global Warming".

4) These numbers are approximate. The eccentricity of the Earth's orbit has a 100 thousand year period. The Earth's spin axis has a combination of 19-21 and 41 thousand year periods (Milankovitch).

5) In 1956, the oceanographers M. Ewing and W. Donn described a theory of Ice ages. They used cores of the ocean sediments from Arctic to argue that the Arctic Ocean was open during the "Wisconsin Ice Age", 110,000 to 10,000 years ago.

A few references:

October 5, 2004

The politics and psychology of media violence: How research confronts rhetoric

Joanne Cantor, UW Department of Communication Arts

This seminar gives an overview of the research findings on the effects of media violence on the attitudes, behaviors, and emotional responses of youth. These findings are contrasted with the way the issue has been covered in media reports of the controversy.  The seminar also looks at the relationship between research findings and projected solutions to the problem, including legislative initiatives and media literacy campaigns.

Joanne Cantor, Professor Emerita of Communication Arts, has authored more than 80 scholarly publications on the impact of the mass media on youth. She is regularly quoted in the media and has testified repeatedly before Congress. She has also written a parenting book, Mommy, I'm Scared: How TV and Movies Frighten Children and What We Can Do to Protect Them (Harcourt, 1998).  Her new children's book, Teddy's TV Troubles (Goblin Fern Press, 2004), helps young children cope with our scary media environment.

October 12, 2004

A chaos theory of brain function

Greg Schmidt, UW Department of Psychiatry

October 19, 2004

Self-organization of power relationships among adults in informal groups

John Martin, UW Department of Sociology

Dominance orders have been posited to be a recurrent form of organization for humans as well as other animals.  The presence of such orders in a form similar to other animals is, however, difficult to detect, since other forms of social organization (e.g. institutions) have their own principles of stratification which may supplant, cut-across, or redefine dominance orders, should they exist.  Even where other forms of organization are minimal, we may only see tendencies towards such ordering.  If, however, humans do tend to create linear dominance orders when other forms of institutional order are absent, there should be a trend toward increasing linearization over time in settings without other organization.  This talk examines longitudinal network data from two sets of groups where dominance orders might be expected to arise -- camp cabins of young adolescents and communes of young adults.  Application of stochastic models finds no evidence of a tendency towards increasing linearization over time, but reveals insights as to the social processes through which hierarchies form and evolve.

October 26, 2004

CO2 air-sea flux variability: Ocean models & atmospheric inversions

Galen McKinley, UW Department of Atmospheric and Oceanic Sciences

<>At present, one-third of the CO2 emitted by anthropogenic activities is absorbed into the ocean. Understanding the year-to-year variability of this large flux helps us to understand the mechanisms controlling ocean CO2 concentrations and to predict future climate change. A global ocean general circulation model is used to estimate the magnitude of interannual variability in air-sea fluxes of CO2 and O2 from 1980 to ­1998 and to examine the controlling mechanisms. The global variability in the air-sea flux of carbon (±0.5 x 1015 grams Carbon yr-1 (PgC yr-1)) is forced by changes of pCO2 and wind speeds related to the El Nino/Southern Oscillation (ENSO) cycle in the equatorial Pacific. In contrast, the air-sea O2 flux is controlled by two regions: the equatorial Pacific and North Atlantic. Carbon flux results are shown to be in broad agreement with independent estimates from a new atmospheric CO2 inversion.

November 2, 2004

Prenatal stress and fetal alcohol exposure: Effects on complex systems in a primate model

Mary Schneider, UW Department of Kinesiology

The increased prevalence of substance abuse and violence in American families has resulted in a growing concern about the effects of psychological stress and fetal alcohol exposure on child and adult outcomes. Our studies with monkeys suggest that how an individual copes with the myriad of environmental events and stimuli can be influenced by a trajectory laid down in fetal life. The broader issue includes fetal brain plasticity or how the fetal brain may encode information about the environment through changes in organization. We are particularly interested in the striatal dopamine system because of the role of fronto-striatal dopamine circuits in integrative functions in the brain, including motor control, attention, working memory, and inhibitory control. We also study how environmental risk factors interact with genetic variation to influence psychobiological development so that a certain genetic variation might be a risk factor for the expression of certain behavioral disorders in some environments or a protective factor in others.

November 9, 2004

Ecosystem consequences of forest fragmentation in the Pacific Northwest: Biogeochemical edge effects within oldgrowth forest remnants

Tom D. Hayes, UW Department of Forest Ecology and Management

My research addresses the long term impact of clearcut edges on biogeochemical processes affecting carbon (C) and nitrogen (N) retention within fragmented oldgrowth Douglas fir/western hemlock forests in the Cascade Range of southern Washington. Relative to distance into remnant forest, I test the general hypothesis that edge-altered microclimate and structure initiate positive feedbacks linking abiotic and biotic responses. Using a mechanistic approach, I seek broader application to conservation and experimental design. Along 360-m gradients spanning clearcut to forest at 9 sites, long-term monitoring of edge effects integrates microclimate, above ground structure, litter fall, decomposition, and N dynamics. Mixed-effects statistical models quantify depth of influence for more than 100 microclimatic, structural, and biogeochemical variables.

Abrupt changes in height and structure at edges increase microclimatic variability in adjacent forest. I find reductions in cover and live-crown volume of plant canopies to be the most significant structural changes near edges. As evidenced by distance effects on basal area and relative importance, altered competition among dominant tree species may affect succession in edge environments. Extreme variability in temperature and reductions in both floor moisture and snow retention are the most notable microclimatic effects near forest edges. Tree live-crown volume and microclimatic variability are negatively correlated. Relative to other forest types, the longer persistence of abiotic edge effects in north temperate coniferous forest may result from the tall stature and slow side-canopy closure of the long-lived tree species.

Structural effects, by modulating microclimatic variability, change the complex biotic interactions involved in nutrient cycling in forest within 120 m of edges. Due to such interactions, N and C pool sizes in soil and vegetation, and net ecosystem production, vary in a nonlinear manner with distance into forest from edge. Field and laboratory assays reveal increases in litter decomposition and N availability in near edge (0 30 m from edge) forest, and higher rates of litter fall and soil organic matter storage within far edge (30 120 m) forest, relative to interior forest (more than 120 m). The decoupling of basic ecosystem processes near edges, including litter fall and decomposition, predicate the polymodal distance effects observed for biologically available N, N uptake, and forest productivity.

The overall results indicate an accurate quantification of edge effects is required within fragmented forest landscapes, in order to assess the cumulative impact of land use on the processing of organic matter and nutrients important to ecosystem sustainability and conservation planning.

November 16, 2004

Genetic reprogramming of the brain by the proteolytic fragments of the amyloid precursor protein: Someplace between chaos and clarity

Jeff Johnson, UW School of Pharmacy

Alzheimer's disease (AD) may be caused by the abnormal processing of the amyloid precursor protein (APP) and the accumulation of b-amyloid (Ab). The amyloid precursor protein can be proteolytically cleaved into multiple fragments, many of which have distinct biological actions. While a high level of Ab can be toxic, the a-secretase cleaved APP (sAPPa) is neuroprotective. However, the mechanism of sAPPa protection is unknown. Here we show that sAPPa increases the expression levels of several neuroprotective genes and protects organotypic hippocampal cultures from Ab-induced tau phosphorylation and neuronal death. Antibody interference and siRNA knock-down demonstrate that the sAPPa-driven expression of transthyretin and insulin-like growth factor 2 is necessary for protection against Ab-induced neuronal death. Mice overexpressing mutant APP possess high levels of sAPPa and transthyretin and do not develop the tau phosphorylation or neuronal loss characteristic of human AD. Chronic infusion of an antibody against transthyretin into the hippocampus of mice overexpressing APPSw leads to increased Ab, tau phosphorylation, and neuronal loss and apoptosis within the CA1 neuronal field. Therefore, the elevated expression of transthyretin is mediated by sAPPa and protects APPSw mice from developing many of the neuropathologies observed in AD.

November 23, 2004

Dynamics of high-dimensional systems

Clint Sprott, UW Department of Physics

Complex systems generally involve the nonlinear interaction of many variables. This talk will describe some simple general mathematical models including polynomial maps and flows, artificial neural networks, and Lotka-Volterra models, that are capable of replicating a wide range of complex phenomena in the biological, social, and physical sciences. Certain generic behaviors are observed, such as a high probability of weak chaos, a unique route to chaos, spontaneous symmetry breaking, and robustness with respect to parameter variation.

This talk is available in PowerPoint format.

November 30, 2004

The application of chaotic dynamics, synchronization, and parameter estimation in an asymmetric vocal fold system

Yu Zhang, UW Department of Surgery

Methods from nonlinear dynamics, including Poincaré map, Lyapunov exponent and dimension, are applied to a vocal fold model with tension, stiffness, and mass imbalances simulating vocal disorders from laryngeal paralysis and vocal mass. Bifurcation diagrams illustrate the effects of these imbalance parameters.  When tension, stiffness, and mass imbalance parameters deviate from the normal value of 1, chaotic vibrations associated with disordered voices occur. Furthermore, the technique of chaos synchronization allows us to manipulate a simulator to approach the vibratory patterns of an original vocal fold system. The minimal glottal area is applied as a feedback variable connecting between two systems. The simulator and the original system are considered synchronized when their state differences asymptotically converge to zero. The effects of noise and parameter mismatches on synchronization are investigated.   Finally, a parameter estimation scheme based on chaos synchronization is applied. Despite noise perturbations and large initial parameter mismatches, the original system parameters can be reproduced in the simulator with parameter controls, and two chaotic vocal fold systems can be synchronized.  Parameter estimation scheme shows the potential application to extract asymmetric biomechanical parameters of the vocal folds from a time series of the glottal area.

December 7, 2004

Stephen Gammie, UW Department of Zoology

The evolutionary and biological basis of protective behaviors
In a wide range of mammals, including primates, bears, cats, dogs and mice, mothers are highly protective when their offspring are young and vulnerable. As part of this protective behavior, lactating females will often express a fierce attack against a larger male intruder in a behavior termed maternal aggression. Although maternal aggression plays a critical role in the perpetuation of species and offspring, it has received relatively little research attention. I will highlight some recent work in mice that suggests one of the reasons females can be so bold and aggressive is that they are ‘fearless’. How the brain regulates fear and anxiety will be discussed. I will also describe some of the general approaches we use in the lab to examine the neural basis of maternal aggression that include: direct injections of peptides into the brain, examinations of brain activity with behavior, and analysis of gene expression changes using arrays. An evolutionary perspective will be used to frame the current and future research questions.

December 14, 2004

3D computer graphics : A fusion of science and art

David Camp, UW Memorial Library

This will primarily be an informal overview of the kind of 3D graphics that can now be done on home computers.  Working in 3D can involve both visual and mathematical skills.  3D space is numerical.  My own work is unique in that I deal with theme galleries structured in symbolic ways.  An example of science in my art is the way I have embedded galleries based on six visible colors within a much broader spectrum of themes in one project intended to portray the vibrational basis of different realities.  This is the "Mystical Planes" project on my web site: