Hanover Forest Science Seminar Series: Erin Schliep
Date: March 28, 2017
Time: 4-5 p.m. (coffee and cookies at 3:50 p.m.)
Location: Natural Resources Building, 480 Wilson Road, Room 225
Erin Schliep presents “Process driven density-dependent diameter distribution models for forest biomass prediction.”
Prediction of aboveground biomass, particularly at large spatial scales, is necessary for estimating global-scale carbon sequestration. Since biomass can be measured only by sacrificing trees, total biomass on plots is never observed. Instead, allometric equations are used to convert individual tree diameter to individual biomass. The values for all trees on a plot are summed to obtain a derived total biomass for the plot. Regression models using environmental covariates are employed to attempt explanation and prediction of derived total biomass. Not surprisingly, when out-of-sample validation is examined, these models predict total biomass well since the holdout data is obtained using exactly the same derived approach. Apart from the somewhat circular nature of the regression model, this approach fails to employ the actual observed plot-level response data. At each plot, we observe a random number of trees, each with an associated diameter. We propose a process driven model based on the number of and observed set of tree diameters which provides understanding of how environmental covariates explain abundance of individuals and how abundance of individuals along with these covariates explain individual diameters. The motivation for density dependence stems from the distribution of tree diameters over a plot of fixed size depending upon the number of trees on the plot. The predictive distributions of individual and plot-level biomass from the density-dependent model for diameter can be more informative for capturing uncertainty than those obtained from modeling derived plot-level biomass directly. We illustrate the density-dependent diameter distribution model using data from the national Forest Inventory and Analysis (FIA) database.