Development and application of the generic Plant growth Modeling Framework (PMF)

Abstract

The use of crop models as part of scientific research models or economic farm tools leads to a wide range of applications. On the one hand they need to be simple; on the other hand they should be complex enough to simulate a variety of growth mechanisms. The development of entirely new models for different questions requires a lot of coding and work such as changes in the model structure, the inclusion of alternative process descriptions or the implementation of additional functionality. Often, added model components do not really fit to the model philosophy of the originally developed base model. We therefore developed a flexible (modular, generic and mixed procedural object oriented) and integrative (replaceable, expandable, independent and interactive) software tool for the setup of adapted crop models. The Plant growth Modeling Framework (PMF) is based on the Unified Modeling Language and implemented in Python, a high level object-oriented programming language. PMF provides the code flexibility to rapidly exchange and compare different process mechanisms. An interface facilitates a straightforward coupling with other models. Two virtual case studies are presented to show the general functionality of PMF. In the first application PMF is coupled with a simple but widely used water balance model presented in the Food and Agricultural Organization FAO56 crop water guidelines. In contrast, the second case study focuses on the coupling of PMF with a complex Richard's equation based water balance model type created using the recently developed Catchment Modeling Framework (CMF). CMF follows the same philosophy as PMF, being flexible and integrative at the same time. In both case studies we show the tight connection between the water balance models and PMF and highlight the framework's ability to function as a model integrator while still being independent from the coupled models. This independency allows, for example, a further development of the different models (here the plant growth and the hydrological model components) by different research communities, which is often required in today's cross-disciplinary research consortia. Further, the modular and generic structure of PMF enables the use of process modules, which fit the level of complexity of the water balance models. Besides aspects of architectural software development, PMF is equipped with an innovative interactive root growth mechanism. Root growth and branching is linked to soil depth dependent environmental conditions such as nitrogen supply or soil moisture. This reflects allows modelers to reflect the high interactivity of root (and plant) growth with soil environmental conditions. Thus, root elongation and root biomass allocation in PMF directly respond to changes of resource availability along the soil profile. This behavior is in close agreement with reality as plants grow where water, nutrients or other resources are available. We conclude that such a dynamic reaction on changes in resource availability improves the overall model credibility.