Discussion/ Conclusions

2006 National Insect and Disease Risk Map

The 2006 national risk assessment employed 186 risk-agent models representing over 50 risk agents acting on 61 tree species or species groups, with all models assembled into a national composite (NIDRM) (see figure at right). Given the nature of our assignment to construct a national, 15-year assessment of forest health risk due to insects and diseases, we believe that NIDRM is successful because it is:

1. Based on an integrating technology. NIDRM represents the collection and integration of multiple risk models developed through an iterative, hands-on process by local forest health specialists. The risk assessment framework presented in this paper is able to integrate outputs from a wide range of models and is implemented through software that gives forest health specialists direct access to GIS models.
2. Transparent and repeatable. The 2006 modeling framework provides a consistent, repeatable, transparent process to conduct risk assessments. Within this framework, forest health specialists are able to determine why an area is at risk, what the source data are, and how the model(s) for that region were constructed, thus documenting any models comprising NIDRM. This type of framework also enables shortcomings in data and models to be identified and can be used to prioritize future research and data development.
3. Interactive and scalable. The framework is interactive enough to support sensitivity analysis while allowing risk assessments to be conducted at various spatial and temporal scales. Sensitivity analysis ensures that models can be adjusted according to local knowledge or as additional data and models become available or both. Scalability enables subject area experts to conduct local and regional assessments using an identical framework. This continuity ensures that national products do not conflict with local knowledge.
4. Efficacious. Efficiency, precision, accuracy, and usability must be considered when developing a framework. A national risk-map product with potentially hundreds of models behind it not only requires a highly efficient modeling process, it must be able to capture the information and variation within each individual model. With a wide range of audiences, including both subject area experts and private citizens, the risk map framework is able to produce detailed model documentation and results that are easy to interpret.
5. Comparable across geographic regions. The 2006 modeling framework has resulted in a standard modeling process that provides a level playing field for every region being examined as part of NIDRM. This ensures that regional comparisons can be made. Without standardization, NIDRM would be little more than a federation of maps with little or no consistency between them, making regional comparisons and national summaries impossible.

Although the framework described in this paper was developed around modeling potential risk of tree mortality due to insects and diseases, the process can be used for a wide range of other applications including estimating potential for wildlife and forest habitat (Krist 2001).