Good morning,

Below is another upcoming seminar at Krasnow.


Begin forwarded message:

Please join us for the next Krasnow Monday Seminar on 10/22/12.
Refreshments will be served at 3:30pm.  Come chat with colleagues and like-minded researchers and students prior to the talk at 4pm.

Modeling geographies of human behavior in critical scenarios

Paul Torrens
Geosimulation Research Laboratory
Department of Geographical Sciences, University of Maryland
College Park, MD

DATE:  Monday, Oct 22, 2012
TIME:  4:00 p.m.
LOCATION:  Lecture Room (Room 229)
           Krasnow Institute Building
           George Mason University, Fairfax, VA

Putting people in the right place, at the right time, in the right context, and in the right company is often a priority in critical scenarios. These dimensions of human geography can be important in determining people’s exposure to risk; in planning efficient coverage for emergency response; in delivering services and resources to populations during times of need; in assessing potential demand and surges on system capacity; and in exploring likely actions, reactions, and interactions of the population during critical events. While significant, estimating and evaluating realistic geographies of human behavior for critical events is a difficult task—by their very nature, many facets of critical situations may be novel, surprising, or unknown. Often, we may have little ground truth or prior experience upon which to base our decisions. In such cases, we often turn to computer modeling and simulation, “in silico”, to explore plans, policies, hypotheses, and ideas that are too difficult (or dangerous) to experiment with on the ground, that have large elements of uncertainty, or that have simply yet to happen. But, to be effective in supporting decisions or plans, computer modeling needs to provide realistic representations of human behavior, and it needs to do so in a manner that is useful for a variety of scenarios and users. The current state-of-the-art in computer modeling is mostly lacking in its ability to provide these functionalities, focusing instead on computational tractability and convenience, at the expense of behavioral realism, with the result that the set of questions that can be posed and answered in simulation is often artificially constrained. In this talk, I will discuss how geosimulation has been pioneered as an alternative scheme for representing human behavior in simulation, with a focus on realism, extensibility, and exploration. I will demonstrate the usefulness of the approach in a variety of quotidian and critical scenarios, including crowd evacuation dynamics, anti-social behavior during riot and protest, and urban dynamics during earthquakes.
For additional directions or information call 703-993-4333 or browse to .
The full semester seminar schedule is at .