An organizational model for multi-scale and multi-formalism simulation: Application in carbon dynamics simulation in West- African Savanna Mahamadou Belem a,⇑ , Jean Pierre Müller b a Centre Régional Agrhymet, Département Formation et Recherche, BP 11011 Niamey, Niger b CIRAD Agricultural Research Centre for International Development, UR Green, TA C-47/F, 34398 Montpelier cedex 5, France article info Article history: Received 2 April 2012 Received in revised form 2 November 2012 Accepted 9 December 2012 Keywords: Complex systems Multi-scale Multi-formalism Simulation Multi-agents systems Organization Discrete event system specification abstract This paper presents Organization-Role-Entity-Aspect (OREA) model, an organizational model for multi-scale and multi-formalism description of complex systems. OREA is based on the assumption that one issue in complex system simulation is to integrate multi-scale and multi-formalism representation. To achieve this issue, we use an approach based on organization-centered multi-agents systems and Discrete Event System Specification (DEVS) formalism. While the organizational approach allows to deal with an explicit rep- resentation of global and local levels, DEVS formalism allows integration of models of dif- ferent types to describe a system perceived at different scales. Integration of OREA formalism within DEVS allows multi-formalism specification of a model both at global and local levels. In addition, this allows specification of the social structure of a complex system following DEVS formalism. In OREA, the organizational structure is specified with- out any assumption on entities structure. The roles description in OREA concerns only the detailed description of interactions within organization. The way that individuals conceive their system and make decision is defined through the concepts of aspect. OREA is applied in implementation of a generic model for carbon dynamics simulation in West-African Savanna. Future works would concern the integration of organizational dynamics and holonic representation in OREA. Ó 2013 Elsevier B.V. All rights reserved. 1. Introduction Complex systems simulation takes more and more an important place in many domains: ecosystem management [1,2], social science [3], economy [4], transport [5], etc. A complex system (1) is as a set of individuals with non-linear behavior, interacting with each other and evolving at different scales of time and space and (2) such that the behavior at the global level cannot be reduced to the composition of the local behaviors [6]. In general, the analysis and the management of such systems are based on the articulation of at least three levels of description: the individuals (local) level, the global level, and underlying environment [6]. The representation of such systems requires a multi-scale and a multi-formalism approach. The individuals’ behavior can be described using utilities optimization process as behavior rules. The interactions among individ- uals and between individuals and their environment are easily represented by multi-agents systems (MAS) [7]. Finally, the global level can emerge from the individuals interactions [8] or be described by compartment models [9]. From what pre- cede, one issue in complex system simulation is to integrate the multi-scale and multi-formalism simulation since we con- 1569-190X/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.simpat.2012.12.006 ⇑ Corresponding author. Tel.: +227 97 93 84 29; fax: +227 20 31 54 35. E-mail addresses: mahamadou.belem@gmail.com, m.belem@agrhymet.ne (M. Belem). Simulation Modelling Practice and Theory 32 (2013) 83–98 Contents lists available at SciVerse ScienceDirect Simulation Modelling Practice and Theory journal homepage: www.elsevier.com/locate/simpat