GePSI: A generic plant simulator based on object-oriented principles

TitleGePSI: A generic plant simulator based on object-oriented principles
Publication TypeJournal Article
Year of Publication1997
AuthorsChen J., Reynolds J.F
JournalEcological Modelling
Volume94
Pagination53-66
Date Published1997
Accession NumberJRN00239
Keywordsarticle, articles, journal, journals, model, GePSI, model, plant, plant, model
Abstract

The Generic Plant Simulator (GePSi) is a physiologically-based model that combines modules for canopy, root environment, water relations, and potential growth to generate whole-plant carbon, nitrogen, and water balances. The version presented here is coded in the object-oriented programming (OOP) language, C + +, to enhance the implementation of modularity. In the aboveground aerial environment, the Weather module defines the weather conditions above a canopy, and Micro Weather defines the vertical profiles of micro-meterological variables in a canopy. The belowground soil environment contains the SoilProperty modules, which define vertical profiles of physical and chemical variables in a soil column. The ‘part-of’ hierarchy in GePSi follows the structure of a real plant: the Plant module calls canopy and root system modules; the Canopy module, in turn, calls leaf, stem and fruit modules; and the RootSystem module calls coarse and fine root modules, etc. Our long-term goal is for GePSi to serve as a template for building a plant growth simulator by simply selecting appropriate modules for the question being asked. We are building a suite of plant modules (and their interfaces) based on general principles that are fundamentally similar for different kinds of plants. This includes photosynthesis, growth, nutrient and carbon allocation, water uptake, etc. These modules can be parameterized for specific species, related groups of species, life-forms, or broader groups depending on how variable the processes are across the groupings and the amount of unexplained variability that is acceptable for the question being investigated. Our modular-based approach has numerous advantages, including improving the understanding of the model, reducing duplication of effort, and facilitating the adaptation of the model for different sites and ecosystems.

URLfiles/bibliography/JRN00239.pdf
DOI10.1016/S0304-3800(96)01928-X