Browsing by Author "Robinson, Andrew P."

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    ACRONYM : a hierarchical tree and forest growth model framework.
    (University of Minnesota, 1998-03) Robinson, Andrew P.; Ek, Alan R.
    ACRONYM is an extensible tree-level forest growth and dynamics model framework for the synthesis and assessment of models and submodels of forest ecosystem processes. The framework is designed for research in forest ecology and silviculture, development and testing of silvicultural guidelines, forest inventory updates, and long-term projections of forest and landscape dynamics. The framework accepts modules (submodels) of forest regeneration, tree growth, mortality, silviculture, harvesting and natural disturbance, each of which may operate at different spatial or temporal scales and levels of resolution. Component modules can range from empirical to process based, as understanding and data permit. Modules can be interchanged with a minimum of reprogramming to allow for comparison of assumptions about processes, different data input protocols, and project-specific reporting requirements.
  • Thumbnail Image
    Item
    Point and interval estimates from sequential sampling.
    (University of Minnesota, 1997-07) Robinson, Andrew P.; Burk, Thomas E.
    This paper reports a simulation-based exploration into the computation of point and interval estimates for data arising from sequential sampling. We conclude that the coverage probability of the standard frequentist confidence interval estimates is overstated. However, there are other interval estimates which do not overstate the coverage probability if the underlying population is Gaussian. Futher, the effect of non-Gaussian behavior in the underlying population upon the properties of the interval estimate varies, depending upon the severity and the flavor, that is, whether it is skewness, kurtosis, etc.
  • Thumbnail Image
    Item
    Regeneration imputation models and analysis for forests in Minnesota.
    (University of Minnesota, 1996-05) Ek, Alan R.; Robinson, Andrew P.; Radtke, Philip J.; Walters, David K.
    Tabular imputation models were developed and tested to estimate post-harvest forest stand characteristics in Minnesota. The models were based on a sorting of statewide inventory plot data into sets of tables containing estimates of number of trees and basal area per acre by covertype, diameter classes, and species for young post-harvest stands. The primary bases for sorting within the sets of tables were stand age input data to existing growth and yield models. Analysis of these tables indicated that basal area increased rapidly for young stands and then began to level off as stand ages approached 20 years; furthermore, the variability within the tables decreased as forest stands matured. Implications for alternative methods of implementing the models are also discussed.
  • Thumbnail Image
    Item
    The variagraph concept and a lack-of-fit test.
    (University of Minnesota, 1996-12) Robinson, Andrew P.; Weisberg, Sanford
    The variagraph is a graphical diagnostic technique for assessing lack of fit and estimating pure error without replication for regression modelling. The behavior of the variagraph is explored by simulation, and its use demonstrated on an example dataset previously applied in lack-of-fit testing.