Estimating Stand Tables from Aerial Attributes: a Comparison of Parametric Prediction and Most Similar Neighbour Methods

Parameter prediction and the most similar neighbour (MSN) approaches were compared to estimate stand tables from aerial information. The study was based on 50 stands in the south-eastern interior of British Columbia, Canada. In the parametric prediction approach, stand tables were estimated from aerial attributes and three percentile points (16.7, 63 and 97%) of the diameter distribution. In the MSN analysis, stand tables were estimated from the MSN stand that was selected using 13 ground and 22 aerial variables. The accuracy of these approaches was evaluated by comparing the observed and estimated species composition, stand tables and volume per hectare. While the parametric prediction approach is easier and flexible to apply, the MSN approach provided reasonable projections, lower bias and lower root mean square error.