Treesearch
Twenty five years long survival analysis of an individual shortleaf pine trees
| Authors: | Pradip Saud, Thomas B. Lynch, Jim M. Guldin |
| Year: | 2016 |
| Type: | General Technical Report |
| Station: | Southern Research Station |
| Source: | In: Schweitzer, Callie J.; Clatterbuck, Wayne K.; Oswalt, Christopher M., eds. Proceedings of the 18th biennial southern silvicultural research conference; 2015 March 2-5; Knoxville, TN. e-Gen. Tech. Rep. SRS-212. Asheville, NC: U.S. Department of Agriculture, Forest Service, Southern Research Station: 555-557 |
Download (PDF 411 KB): https://research.fs.usda.gov/download/treesearch/50836.pdf
Abstract
A semi parametric cox proportion hazard model is preferred when censored data and survival time information is available (Kleinbaum and Klein 1996; Alison 2010). Censored data are observations that have incomplete information related to survival time or event time of interest. In repeated forest measurements, usually observations are either right censored or interval censored. Interval censoring occurs if the exact year of tree death is unknown, since measurement periods are typically longer than one year. Right censoring occurs if tree has not died at the end of the study.