Broad-Scale Classification and Mapping of Tree Size and Density Attributes in Productive Old-Growth Forests in Southeast Alaska's Tongass National Forest
Abstract The forest classification and mapping system currently used in managing the Tongass National Forest (NF) is based largely on an economic forest measure, net board foot volume per acre. Although useful for timber economic modeling, this forest measure poorly differentiates old-growth forest types in a way that is meaningful to ecological and social concerns. In 2005, we published an article presenting a proposed tree size and tree density mapping model for the Tongass NF. We claimed the model would provide better information on the structural patterns in old-growth forests than did the current mapping models based on net board foot volume per acre. We also stated that further testing of our proposed model is required before it can be fully integrated into forest management plans and landscape analysis. In this article, we used independent field data to evaluate our proposed tree size and density model and better define its accuracy. Results showed differences among mapping classes similar to differences observed in the development stages of the model. Results also showed mapping accuracy estimates between 60 and 80%. We used the model in a forest management application by comparing the representation of old-growth forest types within a landscape to the representation within a management-defined subset of that landscape.