|
Abstract:
|
Prediction of above ground biomass at large spatial scales is necessary for estimating global-scale carbon sequestration. Since biomass can only be measured by sacrificing trees, allometry models are used to derive individual biomasses which are then summed over trees on a plot to obtain a derived total biomass for the plot. Regression models using environmental covariates are employed to try to explain this total biomass and make predictions. We argue that we should model what we observe. That is, a random number of trees on a plot, each with an associated diameter, producing a point pattern of diameters. In building a process model, we must provide a point pattern model. Environmental covariates will directly explain the point patterns rather than indirectly connecting them to derived biomasses. Density dependence is incorporated since the distribution of tree diameters over a plot of fixed size depends upon the number of trees on the plot. We show that predictive distributions of plot level biomass obtained from density dependent models for diameters can be more informative for capturing uncertainty than those obtained from modeling derived plot-level biomass directly.
|
