As increasingly more ecological models are developed, it becomes clear that in many cases we are 'reinventing-the-wheel'. The problem is that models are often developed as non-reusable 'one-of-a-kind' items, or worse, that models are 'black boxes' that contain code that by its very nature contains no suitable level of abstraction by which phenomena can be properly understood. One way to avoid that is to build a general ecosystem model, which in theory could eliminate the need for continuous remaking of models for different systems and/or sites. One such model, the General Ecosystem Model (GEM), has been designed to simulate a variety of ecosystem types using a fixed model structure. Such 'generality' logically leads to one of the broader objectives in ecosystem research: with a standard structure for developing a (model) synthesis of a system, comparisons among systems may be facilitated.
The model was to be generally applied to ecosystems that range from wetlands to upland forests. It was to provide at least two useful functions in synthesizing our broader understanding of ecosystem properties. One involves using the model as a quantitative template for comparisons of the different controls on each ecosystem, including the process-related parameters to which the systems are most sensitive. Second, a simulation model, which is general in process, orientation, and structure, could provide a tool to analyze the influence of scale on actual and perceived ecosystem structure. Object orientation provides one example of such a structure with ecosystems being natural phenomena for this type of design. Other models, such as CENTURY for example, can claim to be of the same kind of functionality, providing for a wide range of processes that can be parametrized for very different locations and ecosystem types.
However, the general approach turned out to be somewhat insufficient to cover all the possible variety in ecosystem processes and attributes that come into play when going from one ecosystem type to another, and from one scale to another. Modeling is a goal-driven process, and different goals in most cases will require different models. There is too much ecological variability to be represented efficiently within the framework of one general model. Either something important gets missed, or the model becomes too redundant to be handled efficiently, especially within the framework of larger spatially explicit models. Similarly, when changing scale and resolution, different sets of variables and processes come into focus. Certain processes that could be considered at equilibrium at a weekly timescale need be disintegrated and considered in dynamic at an hourly timescale. For example, ponding of surface water after a rainfall event is an important process at fine temporal resolution, but may become redundant if the time step is large enough to make all the surface water either removed by overland flows, or infiltrated. Daily net primary productivity fluctuations, that are important in a model of crop growth, may be less important in a forest model that is to be run over decades with only average annual climatic data available. Once again the general approach may result in either insufficiency or considerable redundancy.
The modular approach is a logical extension of the general approach. In this case instead of creating a model general enough to represent all the variety of ecological systems under different environmental conditions, we develop a library of modules simulating various components of ecosystems or entire ecosystems under various assumptions and resolutions. In this case the challenge is to put the modules together, using consistent and appropriate scales ofprocess complexity, and make them talk to each other within a framework of a full model. We avoid the 'reinventing-the-wheel' by keeping much of the model structure and replacing only the parts that need to be modified under the particular goals of model implementation.
Was this article helpful?