Crop simulation models contain many different processes such as growth, leaf area expansion, assimilation, transpiration, reproductive growth, etc.  In many cases, a given process may be modeled differently by different modeling teams.  It would be desirable if the methods of modeling those processes were sufficiently modular, to allow different methods for a process to be interchanged within the same crop model, thus keeping all other “background” aspects of the model unchanged.  The CROPGRO model has several examples of this, whereby hourly leaf-level photosynthesis (current default) is substituted for daily canopy assimilation (old method).  Likewise, evapotranspiration in the model can be switched between Priestley-Taylor and FAO-56 Penman-Monteith.   This example is repeated in many other crop models, especially for ET methods.  APSIM, for example, can switch between tipping bucket soil water balance to water potential driven water flow.  The Maize Model Improvement Group is using this approach to test multiple phenology methods and multiple leaf area expansion approaches in the same base maize model.  Also important is that those intermediate processes be measurable to allow comparison to real data such as hourly or daily assimilation and transpiration.