A framework for regional-scale forest production modeling using process-based models and GIS

R. W. Kirk, T. E. Burk, and P. V. Bolstad

Department of Forest Resources, University of Minnesota

While research scientists have used process-based models of forest growth for several decades, forest managers have only recently begun to adopt them in production environments. This lag is accredited to the nature of process-based models, which are often difficult to parameterize, challenging to validate, and built around limited technical implementations. This research addresses these limitations by incorporating standard information system and GIS concepts into the modeling framework. As a sample implementation, the PnET-II and 3-PG models were run within a GIS for the Arrowhead region of northeastern Minnesota and compared against growth estimates from other studies in the region and from the U.S. Forest Service Forest Inventory and Analysis (FIA) database. Based on the experiences of this modeling study and a review of the literature, a framework for implementing process-based models within a GIS is introduced. Primary components of the framework include ecological modeling considerations and data and technological processing requirements. Several GIS-based modeling strategies, including raster and point process approaches, are evaluated. In addition, current technological and programming trends, research priorities and implementation challenges are discussed.

Keywords: process-based models, GIS models, Minnesota

(presentation)