Développement de modèles numériques pour la simulation de croissances organo-minérales: application aux microbialites
Responsable du projet Eric Verrecchia
Pascal Felber
Collaborateur Damien Chappatte
Résumé This project is an interdisciplinary approach involving geology, biology, and computer sciences, in order to develop pertinent numerical models to simulate organo-mineral growth structures. It is a continuation of former studies performed by the applicants which will be carried out more in depth.
This project will be twofold: (i) a theoretical approach, that will allow the development of algorithms and simulation softwares and (ii) a model validation step through application of theoretical results to natural objects (microbialites). The development of computer simulations will lead to the definition of new criteria such as statistic and simulation parameters which will be compared with natural variables.
The objective is to simulate an ecosystem essentially composed of bacteria and minerals, and to investigate the nature of their growth. More precisely, various combination of intrinsic and extrinsic factors that leads to calcium carbonate precipitation must be tried and the morphology of emerging structures must be described and visualized as well as compared with natural shapes.
Therefore, the goal is to develop a tool, a virtual laboratory, able to experiment with given factors and to identify the configurations that lead to the frontier between dissolution and precipitation of CaCO3. The challenge comes from the fact that calcium carbonate precipitation is a rather rare phenomenon compared to dissolution in microbial mats. Under almost identical conditions, precipitation can take place or not. The main technical challenge of this proposal is to investigate new simulation methods, and to design and build this virtual laboratory. This virtual laboratory will allow easy visualization of ecosystem dynamics and testing of various ecological hypotheses, outside of the influence of any uncontrolled external factors. A choice of paradigms for modeling and simulating will be proposed. Those have to fit as best as possible the representation of colonies of bacteria involved in the ecosystem, taking into account the computer scientist's constraints of computing time and memory space.
In conclusion, this multi-disciplinary study is based on an intense collaboration between geobiology and computer sciences. This interdisciplinary research emerges from an important analysis and implementation work, with the goal being to study the behavior of a selected ecosystem. In particular, it forces the geobiologist to identify and thoroughly explain all the components and mechanisms of the simulation, before testing hypotheses.
Mots-clés Stromatolites, Microbialites, Artificial intelligence, Modelization, Laminar Structures, Multi-Agent Systems, Stochastic 3D Models
Page internet http://p3.snf.ch/Project-120207
Type de projet Recherche fondamentale
Domaine de recherche Autres secteurs des sciences de la terre
Source de financement FNS - Encouragement de projets (Div. I-III)
Etat Terminé
Début de projet 1-4-2008
Fin du projet 31-3-2010
Budget alloué 108'672.00
Contact Eric Verrecchia