Keith C. Clarke
Department of Geography/NCGIA
University of California, Santa Barbara Santa Barbara CA 93106 e-mail: email@example.com Phone:805-893-7961 FAX:805-893-3146 Home Page: <http://www.geog.ucsb.edu/~kclarke>
My interest in land use change comes from a background in land use mapping, enriched by a visit to the National Mapping Division of the USGS in 1992. Before concentrating in work on Analytical Cartography, GIS and simulation modeling, I also researched and taught urban geography. In 1992, Len Gaydos and I worked together with others to develop a small land transition modeling group within the USGS global change research program. This project became the Human-Induced Land Transformations project. Primarily centered on the transition to urban land use in the San Francisco Bay area, we realized that to understand urban growth, one needed to know the path of historical land changes. We built a historical land transitions data base for San Francisco, and used the data to build and then calibrate a cellular automaton model of urban growth.
The success of the HILT project led to several sequels. At the USGS and elsewhere, a second data set was constructed for the Washington/Baltimore metropolitan region. The data were also animated successfully, to generate some very revealing visualizations, under a project called temporal urban mapping. As student of mine in New York, Jacob Kramer, applied the Cellular model to a small area in the Stirling Forest area. From my perspective, as the modeling team leader, I extended and refined the model developed in the Bay Area, and began a new project called Gigalopolis, funded out of EDC. This project is extending the model to cover land cover transitions.
The land use change work is that I hope to cover in the paper for this workshop. The work is based on an extended cellular model which (1) takes the cue for land use change from the amount of urbanization in the previous time period (2) uses artificial agents called "Deltatrons" to both diffuse and enforce change and (3) includes additional requirements for calibration and validation.
Issues I am interested in at the meeting are:
Can innovations in visualization and representation be used for understanding
the results of model outputs?
Looking forward to a great meeting.