In this study topographic measurements of channels and drainage basins were performed to clarify the quantitative relationships between channel slope and drainage basin topography, characterized by two distinct parameters, drainage area (A) and mean drainage height (H). The latter was defined as the mountain volume of a catchment divided by the former. Digital elevation data prepared by the Geographical Survey Institute of Japan were utilized for automated and nationwide measurement. Before calculating the drainage area and the mean drainage height at an arbitrary grid point of the digital elevation data, grid drainage networks composed of flow direction lines, which point to the lowest grid point among the eight neighbors, were made. The channel slope can be obtained as the fall of a flow direction line; however, it is not expected to show the actual channel slope because the grid interval is too short to estimate gradient. Thus the channel reach, composed of several steps in the flow direction lines, was defined according to the condition that in each reach the integer values of the logarithm of the associated drainage area are equal to each other. The base of the logarithm must be 2 for the consistent ordering of the channel reaches. The relationship among the drainage area (A), the mean drainage height (H), and the slope of the channel reach (I) was defined by the following equation: I=kHAm, where k and m are constants. The values of m were computed by least-square regression for 84 basins in Japan whose drainage areas were ranked between 1, 024 km2 and 2, 047 km2. The values of m ranged from -0.6 to -0.7 with very high correlation coefficients, and the average was -0.66. Although the investigated river basins have various climatic and geologic conditions, the deviations of m did not indicate any regional trends. Presuming m as -0.5 in the previous equation, the following equation: I=βH/√A, was also examined, where, β is a dimensionless constant. The variable H/-√A, which represents the drainage relief number, indicates the aspect ratio or relief ratio of the drainage basin. The average value of β derived from the 84 basins mentioned above was 0.38, and the standard deviation was 0.045. The constancy of β among various basins shows that the channel slope is more strongly regulated by relief characteristics than by other conditions such as climate and lithology.
Many cities in Japan have only a few large public open spaces with woods, but they have many small patches with garden trees scattered in private land. Among various types of garden trees, tall tree is the most momentous amenity resource in urban areas and as an important item in composed townscapes. The aim of this study is to make clear the distribution and species component of preserved tall trees which have been planted in private land in the Ward area (23 wards) of Tokyo Metropolis, Japan's capital city. In this study the author used data on preserved trees in 20 wards (refer to Table 2) as an index for the degree of accumulation of these trees. The result of the data analysis can be summarized as follows: 1. The 12, 224 tall trees located in 20 wards can be classified into 122 species. 2. The density (n/km2) of preserved trees is generally high in wards located in upland and low in those located in lowlands (Fig. 2). 3. The main component species, which account for over 1 % of all trees, are the following 11 species : Zelkova serrata, Ginkgo biloba, Castanopsis (main species is Castanopsis cupidata), Prunus (main spe-cies is Prunus X yedoensis), Pinus (main species are Pinus densiflora and Pinus thunbergii), Cinnamomum camphora, Quercus (main species is Quercus mysinaefolia), Cedrus deodara, Celtis sinensis, Aphananthe aspera, and Machilus (main species is Machilus thunbergii) (Table 2). Among these, Zelkova serrata, Ginkgo biloba, and Castanopsis occupy over 10% of the all trees, and are generally distributed in all 20 wards (Table 2). 4. Exotic species include 45 species, which correspond to 40% of all 122 recorded species and to 25% of all trees. Seven of these, Zelkova serrata, Castanopsis, Pinus, Quercus, Celtis sinensis, Apha-nanthe aspera, and Machilus, are local species. 5. The distribution patterns of these main component species can be divided into two types : widely distributed pattern (Zelkova serrata, Ginkgo biloba, Castanopsis, Cinnamomum camphora, Celtis sinensis, and Aphananthe aspera), and localized pattern (Prunus, Pinus, Quercus, Cedrus deodara, and Machilus) (Fig. 4). Among the former, the distribution patterns of Zelkova serrata, Ginkgo biloba, and Castanopsis are different in each high-density zone: Zelkova serrate is distributed in the outer city zone, especially the western part ; Ginkgo biloba in the inner area, and Castanopsis in the centralwestern part of the upland (Yamanote area). The latter type is divided into three patterns : 1) Prunus, Qercus, and Cedrus deodara occupy the upland, 2) Machilus occupy the lowland, 3) Pinus distribution is divided between a part of the upland and into a part of the lowland. 6. From the viewpoint of species diversity of the preserved trees, the 20 studied wards are divided into two zones according to the values calculated by diversity indexes (Fig. 5). Eight wards-Suginami-ku, Nakano-ku, Toshima-ku, Shinjuku-ku, Shibuya-ku, Meguro-ku, Minato-ku, and Shinagawa-ku -show high rates, and they are located on a part of the upland area (bordering the Yamanote area and along the JR Chuo Rail Line). The remaining 12 wards have a low rate, and all of them except Katsushika-ku are divided into two sub-types: Zelkova serrata type and Ginkgo biloba type. The former are located in the outer city zone: Nerima-ku, Setagaya-ku, Itabashi-ku, Kita-ku, Adachi-ku, and Edogawa-ku; the latter -Bunkyo-ku, Arakawa-ku, Taito-ku, Sumida-ku, and Koto-ku-are in or near the Shitamachi area (Fig. 5, Table 2).