Common Modeling Framework
The NC CSC is committed to using open access and community driven protocols for climate, impact, and decision support analysis in its research initiatives. This collaborative and community-based model has been successfully utilized by the National Center for Atmospheric Research through their Community Earth System Model and sub-models, which were designed with input from various research communities and rapid inclusion of state-of-the-art science coming from the Intergovernmental Panel on Climate Change and other sources. The NC CSC will look to follow this successful example in building a common modeling framework for the north central region by following the below framework:
- To the extent possible, modeling will be done through a collaborative process between decision makers, the climate modeling community, and researchers familiar with the relevant ecological, energy, cultural, or management disciplines;
- Models will be as transparent as possible so that their outputs can be mapped;
- Software will take advantage of high-performance computing as available
- Modeling frameworks will allow for community input.
Provenance and Portability using VisTrails
Recently developed workflow and visualization tools will be integrated into this common modeling framework. VisTrails is an open source software program that can help researchers and managers visualize workflows and display analysis results while capturing the provenance of each model run - including input data and modeling parameters. VisTrails provides a documented, repeatable audit of model analysis. NC CSC is using this software to establish protocols for implementing a common framework for ecological response modeling. By integrating modeling workflows in VisTrails, the NC CSC will start to design a system where the outputs from one modeling effort can easily become inputs to subsequent analysis.
This software has a number of benefits, including that: it allows for the formalization and tractable recording of the entire analysis process; it allows for easier collaboration by enabling one group to understand what another has done; it lets teams iteration on necessary adjustments and changes; it reduces the burden on researchers by offering disparate tolls and custom processing steps in a user-friendly interface; and it formalizes components in a modular setting and allows for the incorporation of future modeling routines and tools (Morisette et al).