Crop modelling and validation : integration of IRENE_DLL in the WARM environment

The growing importance of biophysical models in research and application-oriented projects is driving a growing interest in developing suitable approaches to evaluate model performance. Valuable validation techniques should assess the performance of complex models under a variety of conditions, and should include a wide range of validation measures. After discussing validation issues and methods currently used to assess the quality of simulation models, the integration of the software component for model output evaluation IRENE_DLL within the rice modelling system WARM is illustrated. The purpose is to demonstrate, via a case study, that great utility in validation can be gained through the implementation and use of object-oriented software tools targeting at modularity and reusability inside a modelling environment. This facilitates model validation sessions and extensibility of tools towards new approaches possibly coming out of research. These challenges can be met by using a wide range of approaches and by expanding horizons in validation whilst tailoring the evaluation requirements to the specific objectives of the model application. The availability of appropriate software tools allow actions that are not frequently executed within the context of project-based modelling activities, thus helping the dissemination of validation experiences and preventing future modelling projects from repetition of validation efforts.