Research Progress Report 156 November, 1964
Time-dependent Correlation Coefficients from Remeasured Forest Plots
James N. Hool, and Thomas W. Beers Department of Forestry & Conservation
Introduction
Many industrial and public organizations in recent years have installed on their lands an inventory system known as continuous forest inventory (CFI). This system dealing with the establishment and remeasurement of permanent forest plots is a means of obtaining current and periodic information basic to the control and management of the forest. The periodic remeasurement of permanent plots provides certain information which cannot be obtained satisfactorily in any other way. The alert CFI forester will be remiss if all such information is not gleaned from his inventory. This paper discusses one type of analysis to which permanent installations are uniquely amenable, with the long-term goal of improving the efficiency of the forest inventory itself. Needless to say such efficiency improvement should be the constant concern of the forest manager.
The purpose of this report is twofold: to show by example how past and future CFI remeasurements can supply information needed by the inventory designer and to present empirical correlation coefficients and trends for discretionary use in specific inventory design situations. It is
believed that these empirical data are the most complete currently available.
Considerable interest has been shown in the method of forest inventory most recently described by Ware and Cunia (1962) as continuous forest inventory with partial replacement of samples. While a complete description of this method is beyond the scope of this paper, the primary reason for considering such an inventory scheme is to improve the efficiency of estimating current forest parameters with some recognized sacrifice in the estimation of parameters which depend on the passage of time, i.e., forest growth. By this approach the overall efficiency of the inventory is meant to be improved. It is appropriate to quote the succinct statements of Ware and Cunia (1962) in this regard. "The merit of remeasured permanent sample plots for augmenting (via partial replacement) the efficiency of a sampling plan for estimating current volume of the forest has been even less generally recognized [than the value of remeasured permanent plots over successive temporary plots in estimating growth]. But on examination it is intuitively apparent that the correlation coefficient observed in the remeasured portion of the initial sample can be applied to the initial plots that are not re measured thus giving an estimate of current
PURDUE UNIVERSITY • Agricultural Experiment Station • Lafayette, Indiana