You are here:
A Conceptual Framework for Evaluating the Domains of Applicability of Ecological Models and its Implementation in the Ecological Production Function Library - International Society for Ecological Modelling Conference
Citation:
Moon, J., Ted DeWitt, Randy Bruins, AND M. Sheng. A Conceptual Framework for Evaluating the Domains of Applicability of Ecological Models and its Implementation in the Ecological Production Function Library - International Society for Ecological Modelling Conference. Presented at International Society for Ecological Modelling 19th Biennial Conference, Toulouse, FRANCE, October 28 - 31, 2013.
Impact/Purpose:
EPA scientists and an ORISE postdoctoral fellow at NHEERL/WED have developed a conceptual framework for evaluating the transferability of ecological models to locations, across landscapes, or different time periods than the models had been applied to previously. This framework provides a logical and consistent procedure for making that evaluation of transferability or “suitability”. Furthermore, in collaboration with EPA scientists and collaborators at NERL/EERD, they are developing tools to implement this framework using the Ecological Production Function Library. The use of ecological models to inform environmental management and policy has proliferated in the past 25 years. A key attraction of ecological models is the ability to apply them in new ‘contexts’ (i.e., locations, ecosystems, spatial and temporal extents) without having to collect extensive new datasets. The combination of the transferability framework and the Ecological Production Function Library provide environmental managers and modelers with a guided assessment of the applicability of existing ecological models to the user’s context.
Description:
The use of computational ecological models to inform environmental management and policy has proliferated in the past 25 years. These models have become essential tools as linkages and feedbacks between human actions and ecological responses can be complex, and as funds for sampling have become increasingly limited. A key attraction of ecological models is the ability to apply them in new ‘contexts’ (i.e., locations, ecosystems, spatial and temporal extents) without having to collect extensive new datasets. There are, however, recognized risks when an ecological model developed in one context is applied in another. In order to better assess this risk we have developed a general conceptual framework that aids in evaluating the potential contextual range of a given model’s application, and we have implemented this framework within a database that allows end users to browse, compare and select among ecological models. The framework draws on the “transferability” literature and the importance of a model’s “life cycle”, as described in the 2007 National Research Council Report, Models in Environmental Regulatory Decision Making. The framework, in the form of a decision tree, assesses a model’s conceptual validity (i.e., model type, complexity and structure) during development as well as its evolution and its use in the literature by evaluating its performance (i.e., validation, model comparison and uncertainty) and its similarity in four domains of context: scale, geography, ecology (e.g., geophysical, biotic, etc.), and parameter range or hyperspace. We will showcase the implementation of this framework in the Ecological Production Function Library (EPF-L), a database currently being developed by the U.S. Environmental Protection Agency. Its goal is to provide end users with an array of ecological models and a guided assessment of their applicability to the user’s context.