Effects of Measurement Errors on Individual Tree Stem Volume Estimates for the Austrian National Forest Inventory

National forest inventories typically estimate individual tree volumes using models that rely on measurements of predictor variables such as tree height and diameter, both of which are subject to measurement error. The aim of this study was to quantify the impacts of these measurement errors on the uncertainty of the model-based tree stem volume estimates. The impacts were investigated using two approaches: the law of propagation of error and Monte Carlo simulation. Estimates of total uncertainty also included variability associated with the model itself. Results for both approaches indicate that the relative standard deviation over plots of the volume estimates for all tree species is approximately 11%. A partition of the total uncertainty by sources indicates that error in measurement of the upper diameter makes the greatest contribution. Thus, the greatest potential for improvement in the precision of overall estimates lies in increasing the accuracy of upper diameter measurements. Although the uncertainty of individual tree stem volume estimates may be considered negligible for nationwide assessments of growing stock volume, it is relevant for small-scale and plot-level estimates used as training or accuracy assessment data for remote sensing applications that rely on emerging technologies such as airborne laser scanning.