Non-combatants have become an important aspect of U.S. military operations in urban areas. Recent experience shows how non-combatants can affect the United States’ ability to meet tactical and strategic objectives in engagements across the spectrum of warfare. However, there is currently little systematic research on civilian behavior within the defense community, including the military modeling community. As the policy questions about dealing with civilians continue to gain in importance, further research on non-combatants would be beneficial. This dissertation reviews recent urban operation campaigns and attempts to provide background research that will assist in incorporating non-combatants into models, simulations, training scenarios, and other analytic tools in a more formal way. It identifies non-combatant behavior from recent urban operations that have affected U.S. military activities. It recommends a layered approach to civilian behavior, beginning with basic population density and other demographic characteristics. To this, it adds simple and then complex behaviors. This dissertation also assesses methods for modeling large numbers of non-combatants and proposes using agent-based modeling (ABM). Introducing agent-based non-combatants into existing models and simulations also has the potential to extend current force-on-force models and allow them to be used in examining urban operations. This is an important practical consideration and an alternative to waiting, possibly for years, until new urban combat models are built, tested, and formally accredited.
