Why we go where when we do

-A A +A
By Roger Snodgrass

Knowing where people are likely to be at a given time may be nearly as valuable and far easier to figure than knowing where they are.

Geographer Kriste Henson, working on a doctorate from the University of California, Santa Barbara, has been awarded three prestigious grants and fellowships worth more than $46,000 to create a template for the movements of individuals according to common profiles and patterns.

“Once you have their travel patterns, you can look at all sorts of different problems,” she said in an interview Tuesday.

The grants will enable her to develop her dissertation while continuing to work part time in the Decision Applications Division at Los Alamos National Laboratory, which has provided significant support for her advanced studies and her current project.

“A lot of people think, ‘My city is unique, and no one can match it,’” Henson said in a telephone interview, noting that others saw relationships that were similar across many populations.

The question she posed was, “If so, what influences their travel?”

Describing individual movements within populations has been addressed mainly by household travel surveys.

Real people keep detailed diaries that describe their daily wanderings – how they get to work, where they stop along the way, when they walk and when they take the bus, for example.

Henson’s research proposes to identify general characteristics based on such factors as land-use and land-form data, population characteristics, and access to transportation routes, shopping opportunities and other amenities.

With the model she is developing, one person’s recorded activities can be “transferred” to another individual living somewhere else. The “synthetic” population can go about their programmed business and their movements can be compared to the patterns of real people with a range of corresponding “lifestyle” characteristics.

Towns or cities that haven’t updated full-scale travel studies, which are expensive and time-consuming, can still have a theoretical basis for policy decisions, Henson said.

Are some of these characteristics more relevant in determining travel behavior than others? And, if so, which of the measures most effectively encapsulate nationally available data? These are among the questions to be explored.

Henson began her computer modeling 11 years ago while getting her master’s degree from the University of New Mexico.

She worked on the lab’s TRANSIMS projects, modeling moment-to-moment transportation scenarios for cities like Portland, Ore.

TRANSIMS is a short version of the name Transportation Analysis SIMulation System.

The LANL TRANSIMS program models traffic flow and enables analysis of the transportation system and how it will perform under a variety of situations.

These activity patterns are related to a number of transportation and population questions of interest to the Department of Homeland Security, among others. DHS has been tasked with protecting the nation’s critical infrastucture with a strategy based on a rational assessment of threats and risks.

Models like the one Henson is developing may help prioritize key resources and potential targets, like a bridge or a shopping center, and how their value or risk may vary during the course of a day.

These types of simulations, referred to as agent-based systems have become increasingly useful for studying the complexities of behaviors involving living entities. TRANSIMS derivatives are also used to explore problems like the spread of a contagious disease.

For her research, Henson was awarded an Eisenhower Graduate Fellowship from the United States Department of Transportation, a Geography and Regional Science Dissertation Improvement Grant from the National Science Foundation and a doctoral Dissertation Grant from the University of California Transportation Center.

Henson attributed her good fortune in part to her faculty advisor, Kostas Goulias, a UCSB professor, who has a knack for raising funds, she said.