Creating homogenous landfire vegetation classes for forest inventory applications in the Interior West

Abstract

This project involved collapsing the existing vegetation type (EVT) classes created by the LANDFIRE Program into a smaller set of structure metrics based on four key forest attributes: (1) aboveground biomass, (2) net volume of trees, (3) count of trees per acre, and (4) basal area. Our goal was to create categories that were internally homogenous with respect to these four variables and heterogeneous between categories so that the resulting categories could be effectively used in estimation, mapping, and analysis applications for forest inventory. We used k-means clustering as the main classification method and made refinements to the algorithm in order to make the final categories as intuitive and useful as possible. The result was two classification schemes (1) one with five categories, and (2) one with 10 categories. We used data from one cycle of the Interior West region of the Forest Inventory and Analysis Program (FIA) but intend our method to be applicable for all FIA administered regions.

Parent Publication

• Celebrating progress, possibilities, and partnerships: Proceedings of the 2019 Forest Inventory and Analysis (FIA) Science Stakeholder Meeting.

Citation

Basil, Madelon R. K.; Huque, Salma; McConville, Kelly S.; Moisen, Gretchen G.; Frescino, Tracey S. 2020. Creating homogenous landfire vegetation classes for forest inventory applications in the Interior West. In: Brandeis, Thomas J., comp. Celebrating progress, possibilities, and partnerships: Proceedings of the 2019 Forest Inventory and Analysis (FIA) Science Stakeholder Meeting; November 19-21, 2019; Knoxville, TN. e-Gen. Tech. Rep. SRS-256. Asheville, NC: U.S. Department of Agriculture Forest Service, Southern Research Station. p. 254-267.