This paper deals with the selection of representative unit block and the estimation of its stand volume (m3/ha) for the purpose of the planning of forest road network in the palm form woking system of Kyushu University Forest in Hokkaido.Based on form patterns and terrain classes, all of the 50 unit blocks were classified into six types, and a representative unit block was selected from each type.Furthermore, the stand volume in m3 per hectare of each unit block was estimated by a linear regression equation which was constructed using the field and photo data from the sample plots.The stand volume (m3/ha) estimated in applying the regression equation was distributed from 50 to 250 with a large variation in the same unit block.The results obtained showed the effective measure for planning of the forest road network.

The palm form working system has been applying to a part of Kyushu University Forest in Hokkaido.It is necessary to classify the natural broadleaved stands corresponding to the planning of timber cut and forest road network to maximize the efficiency of the system.This paper analyzes the classification of forest types based on aerial photographs and multivariate analysis.Data used in this analysis are 4 attributes obtained by photointerpretation.For the methods of multivariate analysis, the procedures are as follows : 1)The grouping of unit blocks is shown by the dendrogram resulting from a cluster analysis.2) The unit block is presented by the scatter diagram produced by using principal component analysis.3) After the mutual relationships between the unit blocks were discussed, each unit block was classified into 8 forest types.The results showed the effective measure for planning of timber cut and forest road network.