Abstract
The present study considers the regional characteristics of truck movements within Hokkaido by using quantitative .analyses, and attempts to clarify the regional structure of Hokkaido from the viewpoint of truck movements as an economic activity. The basic material for this study is the Automobile O-D Survey conducted by the Hokkaido Development Bureau during the autumn of 1971. For the present study the following information was tabulated from the material:
1) total number of daily truck trips between paired unit areas (cities, towns and vil lages),
2) total number of daily trip ends in each unit area, by type of land use, type of loaded freight and trucks.
Consideration is focussed on the description of both inter-areal trip patterns and originating/terminating trip patterns. The study results are summarized as follows.
The 209 (number of unit areas) × 209 O-D truck trip matrix was subjected to factor analysis, and the first 44 factors each with more than 0.54 percent of the total variance were thought to reveal the inter-areal trip patterns (Fig. 1). The inter-areal trips are focussed on such major cities as Sapporo, Asahikawa, Kushiro, Kitami, Obihiro and Hakodate (including Kameda-cho) cities. Other major cities Otaru and Muroran, however, , are not fully functioning as traffic centers because of the relative situations to other larger cities. By examining the distribution of factor loadings and factor scores on each factor, . the writer delimited 21 traffic regions (Fig. 3), each of which has unity and self-sufficiency in terms of inter and intra-areal truck trips. The boundaries of traffic regions coincide with mountains and administrative divisions, i.e. sub-prefectural Shicho's divisions.
Subsequently, 32 variables were selected to indicate originating/terminating aspects of truck trips for unit areas (Table 2). The 32 × 209 data matrix was subjected to factor analysis, and 7 factors with eigenvalues of more than 1. 0 were extracted. These factors are, in their order of occurrence and importance:
1) urban traffic density based on manufacturing and commerce,
2) rural traffic density based on primary activities,
3) traffic density of heavy trucks based on public works,
4) traffic size,
5) external contact of traffic,
6) and 7) remain unexplained.
Unit areas within the same traffic region being treated as a swarm of points in the fivedimensional space, the centroids of each of 21 swarms were located. Ward's grouping=algorithm was used to classify 21 traffic regions on the basis of the coordinates of the centroids. Judging from the increased loss of information derived from the grouping, the writer cut off the dendrogram at the stage of 5 groups (Fig. 4). The characteristics of each of 5 groups were described by using the average score of the group.
The traffic regions in the first group are primarily characterized by higher traffic density based on manufacturing and commerce, and smaller traffic size (A Type in Fig. 5). Southern Dooh (southern part of Central Hokkaido) forms a group of its own, reflecting its dominance in the industrial activities which are developed in and around Sapporo city, the regional capital of Hokkaido (B Type). The traffic regions in the third group have higher traffic density of heavy trucks involved in public works, and lower external contact of traffic (C Type). The fourth group has no particular features, except that it has. lower traffic densities (D Type). Yoteisanroku, Northern Shiribeshi and Eastern Tenpoku formm the fifth group. They are characterized by higher traffic density based on primary activities, higher external contact of traffic, lower urban traffic density and lower traffic density of heavy trucks (E Type).
