Turnblom, EricHawes, Bridgette2021-08-262021-08-262021-08-262021Hawes_washington_0250O_22992.pdfhttp://hdl.handle.net/1773/47549Thesis (Master's)--University of Washington, 2021Forest management is dynamic and complex, with numerous factors affecting forest development and outcomes of management decisions. In plantation forestry, forest development is typically expected to follow well understood patterns. However, these patterns may vary substantially due to unpredictable factors. One such factor that is difficult to accurately anticipate is tree mortality. To facilitate the ability of managers to make the best management decisions, mortality must be effectively estimated over the life of the stand. This is often done using forest growth and yield projections, however most growth and yield modeling techniques oversimplify the process of accounting for stand-level mortality. The objective of this project was to create a deterministic mortality model for Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco) plantations in the Pacific Northwest that can differentiate the pattern of mortality from the starting planting density. Mortality models have existed for decades, all taking different approaches to modeling and validation. Multiple new mortality modeling techniques were explored throughout this study, using non-linear sigmoidal curves to capture the trend of mortality over time. Using a parameter prediction approach and validating the process with a bootstrapping procedure, a final model was selected based on the R^2, MAE and SD values. Both absolute and relative models were used in the overall process before determining that an absolute model was capturing the behavior of mortality better over time and moved forward with. The final model will be coupled with the Stand Management Cooperatives Plantation Yield Calculator--a growth and yield model-- to improve overall estimates to aid in management decisions in Douglas-fir stands.application/pdfen-USnoneForestryForestryThesis