Improving growth and yield estimates with a process model derived growth index

Abstract

Forest managers have long made use of the regular and predictable nature of tree growth by using empirical growth and yield models to update forest inventories. Updated inventories support better decision making without requiring on the ground reassessment of the forest resource. Growth and yield model predictions can suffer from inaccuracies due to the influence of climate and environmental variability on the growth of trees. Researchers have been attempting to assess and predict the effect of this variation by developing mechanistic process models that often do not generate outputs applicable to inventory update. Here we create a growth index dependent on process model outputs to improve growth and yield estimates. Estimate accuracy was modestly improved over the basic growth and yield estimates and was comparable to previous efforts to account for environmental variability in growth and yield estimates. Using a process model we are nominally considering the entire environment, and by adjusting the growth and yield estimates external to both model types we have avoided difficulties involved with refitting or recreating either model. These are key differences from previous efforts to include environmental variability in growth and yield estimates.