We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Two-Step Regression Process for Whole Stand Loblolly Pine Survival Projection and Quantifying Uncertainty.
- Authors
Gallagher, Derrick A; Montes, Cristian R; Bullock, Bronson P; Kane, Michael B
- Abstract
The two-step regression process for whole stand survival modeling has been widely used for multiple species. However, estimating the parameters of a probability model for the discrete event of mortality and a whole stand survival model separately can lead to biased projections and poor estimates of the variability associated with future projections. This study derives a mixture probability density function for the combined equation used in the two-step survival regression method that can be used to estimate stand survival and projection uncertainty. Our new density function was used in the maximum likelihood framework to find optimal parameters for a whole stand survival function developed for loblolly pine (Pinus taeda) in the Western Gulf region of the US. The mixture probability density function results in accurate and unbiased projections of future stand density and provides a means to evaluate long-term model prediction uncertainty.
- Subjects
UNITED States; LOBLOLLY pine; PROBABILITY density function; TREE mortality; SPECIES
- Publication
Forest Science, 2019, Vol 65, Issue 3, p265
- ISSN
0015-749X
- Publication type
Article
- DOI
10.1093/forsci/fxy055