This dataset portrays the "current" (2010) viability score (scale of 0 - 1.0) for Blue spruce (Picea pungens) in western North America. It serves as the base condition for future species-climate profiles. (From Crookston et al. 2010): To develop the climate profile, we used a data from permanent sample plots largely from Forest Inventory and Analysis (FIA, Bechtold and Patterson, 2005) but supplemented with research plot data to provide about 117,000 observations (see Rehfeldt et al., 2006, 2009) describing the presence and absence of numerous species. The Random Forests classification tree of Breiman (2001), implemented in R by Liaw and Wiener (2002), was then used to predict the presence or absence of species from climate variables. [...]
Summary
This dataset portrays the "current" (2010) viability score (scale of 0 - 1.0) for Blue spruce (Picea pungens) in western North America. It serves as the base condition for future species-climate profiles.
(From Crookston et al. 2010): To develop the climate profile, we used a data from permanent sample plots largely from Forest Inventory and Analysis (FIA, Bechtold and Patterson, 2005) but supplemented with research plot data to provide about 117,000 observations (see Rehfeldt et al., 2006, 2009) describing the presence and absence of numerous species. The Random Forests classification tree of Breiman (2001), implemented in R by Liaw and Wiener (2002), was then used to predict the presence or absence of species from climate variables. The Random Forests algorithm outputs statistics (i.e., vote counts) that reflect the likelihood (proportion of the total votes cast) that the climate at a location would be suitable for a species. We interpret this likelihood as a viability score: values near zero indicate a low suitability while those near 1.0 indicate a suitability so high that the species is nearly always present in that climate.
