HOME


Track: Database Design and Automation

Thomas Drennan
Sandia National Laboratories
POB 5800
M/S 0749
Albuquerque, NM 87185-0749


Telephone: 716-393-0221
Fax: 716-393-0223
E-mail: tedrennan@sandia.gov



John Ganter

DSM-GIS: An Application that Links MapObjects to the Powersim Dynamic Simulation Modeler  Paper Text

The rise of component software reduces the mechanical difficulty of connecting GIS and a variety of modeling tools and environments. Sandia has recently used Map Objects and Powersim to create a single application for running dynamic simulation models (DSMs) and viewing their output in geographical terms. DSMs are the tools of System Dynamics, where mental models of things and their interactions over time are formalized as stocks, flows, functions, constants, feedback loops, and other elements. The model is then run, with users observing system behavior through animated charts and other displays as model time passes. DSMs have been applied to many commercial and industrial processes (e.g., product realization, sales and marketing, human resources). They have also been used to model policy issues such as population growth, sustainable development, and resource use. DSMs in the form of Powersim and related products do not seem to have been combined with GIS, however. Sandia has performed a series of experiments to see if GIS data and possibly displays can be combined with DSM and applied to subjects like human-induced climate change, international energy flows, and global resources. Trade studies were used to identify a rich, user-friendly, and architecturally open DSM. Powersim was chosen for its capability and low-level OLE/OCX interface. Several GIS components were evaluated; MapObjects was selected over an initial client/server approach using ArcView. At this point, the two foundations were in place. We describe the ensuing integration process and discuss some lessons that were learned. Our prototype also illuminates a number of deeper, methodological issues that are fertile ground for additional research and development.



Copyright 1997 Environmental Systems Research Institute