Projecting COSAGE output in discrete time.

• Main Author: Schweighofer, Marc C.
• Other Authors: Gaver, Donald P.
• Language: en_US
• Published: Monterey, California: Naval Postgraduate School 2012
• Online Access: http://hdl.handle.net/10945/13472

The Army's Combat Sample Generator (COSAGE) is a two-sided, symmetrical, high-resolution stochastic simulation model that projects the outcome of ground combat between two forces. Blue force is typically a division; Red force size may be scaled from a fraction of a division to a combined arms army. Because COSAGE is high-resolution (many asset types), it requires extensive data preparation time, and because output is the result of 16-20 replications, substantial simulation run-time. The analytical model implementation of this thesis is developed to economically project ground combat attrition and munitions expenditures beyond the 48-hour period currently modeled in COSAGE. The implementation evaluates Bayesian estimators of time-period survivorship to estimate expected numbers of kills, both friendly and enemy, during the first 48 hours of combat, then extrapolates those estimates in discrete time StepS (here 24 hours) beyond 48 hours. The implementation can be used to project COSAGE output for all combat postures in Northeast and Southwest Asia (NEA and SWA respectively). An application of the current implementation is to support the warfighting Commanders in Chief (CinC) need to create a Phased Threat Distribution (PTD) in accordance with the Capabilties-Based Munition Requirement Process introduced in June 1997.