PAPER  1997 Year

　

　

　

　


　

Evaluation of the Urban Landscape in Pusan by multivariate analysis

　

　

1.ABSTRACT

　

　　The main purpose of this paper is to develop a method to clarify characteristics of urban landscape using a picture database system and to evaluate the urban landscape of Pusan by means of multivariate analysis. In this study, the following items were analyzed to evaluate the relation between landscape attribute and landscape composition . As the first step, we construct the Landscape Picture Database System which can combine a picture and its attribute data in order to synthetically analyze a large quantity of image data and its characteristics. Second step, urban landscape was analyzed by the " Hayashi's quantification theory Ⅲ" a multivariate analysis method. As a result, the urban landscape was classified into 5 types: (1)town view landscape including seacoast, (2)residential area landscape densely composed by multi-elements, (3)residential and commercial area view landscape, (4)commercial area street landscape, (5)landscape including river and field.

　

　

2.INTRODUCTION

　

Pusan, the biggest international port city in Korea, lies on the southeast coast of the Korean peninsula. It has achieved a rapid growth over the last 50 years, with its population increasing from 280,000 to 4 million. Mountainous topographical conditions, rapid urbanization, and green-belt districts of original Korean urban planning system created a dense urban space with many high-rise buildings. At present, there is increasing interest the urban landscape which is part of the living environment, and so it is important to grasp the present state of the urban landscape. In order to make effective landscape control regulations which maintain the original landscape, this study examines the characteristics of the landscape composition element in Pusan city using the landscape picture database.

　

　

3.BUILDING PICTURE DATABASE FOR LANDSCAPE EVALUATION

　

1)Setting the view point
When preparing a picture database, first we set 35 view-points that cover a balanced range of various view-objects, next we made a matrix graph of use districts of urban planning and constituent element of landscape. Then, approximately 1,100 photographs were taken at the view-points. Finally, 300 of the photographs were chosen for the database.

　

2)Organization of landscape picture database
Each photograph is classified according to the following five characteristics : 1) viewpoint attribute , 2) view-object attribute, 3) classification by objective point, 4) landscape composition elements by short-distance, medium-distance and long-distance in view area, 5) factor of landscape obstacle

　

　

4.CLASSIFICATION OF URBAN LANDSCAPE IN PUSAN CITY

By using the following data, the urban landscape in Pusan city was classified according to the landscape composition elements and by distance. We established landscape distances as follows: short-distance landscape is 0-200 meters, meddle-distance landscape is 200-1,000 meters, and long-distance landscape is beyond 1,000 meters. Landscape composition elements consist of 11 natural composition elements and 25 artificial composition elements were divided into 3 distance categories. This study analyzes the characteristics of the landscape composition element in Pusan city using the " Hayashi's quantification theory Ⅲ".

　

1)Basic concept of urban landscape in Pusan
In factor 1, item's category scores which are ' advertisement tower in short-distance', 'sign board in short-distance and medium-distance' and 'signal lamp in short-distance' shows high toward positive. And item's category scores which are 'river in long-distance', 'factory in long-distance' and 'seacoast in long-distance' show high toward negative. In factor 2, item's category scores which are 'river in short-distance and long-distance' and 'field in medium-distance and long-distance' shows high toward positive. And item's category scores which are 'mountain in short-distance', 'seacoast in short-distance and long-distance', 'port in long-distance', 'advertisement tower in medium-distance' and 'high-rise building in long-distance' show high toward negative.
Judging from the above, a positive value for factor 1 shows that items are composition elements of street landscape elements and indicates that there are many small scale elements, and a negative value for factor 1 indicates many large scale elements. And so, factor 1 is called " the large element-small element composition axis " and it reflects various forms of urban landscape in Pusan city that have various landscape compositions. And, a positive value for factor 2 shows the topographical character of the west suburban area in Pusan city, which is surrounded by designated green-belt districts. A negative value for factor 2 shows the topographical character of the downtown area and describes the town landscape. And so, factor 2 is called " the town high density-suburban low density axis ", and it reflects the unique land system of Pusan city.
( i.e. designated green-belt districts )

　

　

　

2)Classification landscape in Pusan
In this section, a classification was made by the multivariate analysis which were explained by two factors composing the Pusan urban landscape. By means of an item's category scores and sample score by the " Hayashi quantification theory Ⅲ", the urban landscape was classified ( figure 2 ). The result of analysis shows the following five type extracted.

　

Type 1 ; Town view landscape including seacoast The elements of sea, mountain and building group in the long-distance are common with the obtained landscapes. Especially, the appearance frequency distribution of sea is higher than any other type.
The landscape highly rated are divided to two types with a mountain of the background. One is panoramic view landscape of the city across water, and the other is the town view landscape from the edge of the seacoast.

　

Type 2 ; Residential area landscape densely composed by multi-elements 　　The short- distance landscape elements is vehicle, road and electric pole in high ratio of appearance frequency. Middle-distance landscape is house, apartment building, and low-rise building.
Composition of this type landscape is from the street landscape perspective to the landscape that views to the wider residential area.

　

Type 3 ; Residential and commercial area view landscape　 The elements of mountain in middle-distance and long-distance, and building in long-distance are common to obtained landscapes. Representative photographs are the view landscape overlooking the combined residential and commercial area.

　

Type 4 ; Commercial area street landscape 　The elements of road, sign board, electric pole, vehicle and low-rise building are common to obtained landscapes. Representative photographs are the commercial area's street landscape. Composition of this type landscape is from the street landscape perspective.

　

Type 5 ; Landscape including river and field 　 The elements of field, river and mountain are common with the obtained landscapes. Especially, the appearance frequency distribution of field, river, is higher than any other type. And concerning elements of the long-distance, the appearance frequency of building group, high-rise apartment building and factory is slightly high. Composition of this type is the open landscape.

　

　

5.CONCLUSION

It is clear that this landscape picture database system which is a method clarified characteristics of urban landscape is effective. We obtained that the urban landscape consists of two factors as follows: the one is large element-small element composition axis which represents various forms of urban landscape in Pusan city, and the other is the town high density-suburban low density axis which represents unique land using system and topographical character of Pusan city. The urban landscape is classified by these factors into 5 types. Using a multivariate analysis for landscape evaluation, the comprehensive information on the characteristics of landscape in Pusan are proposed.

　

　

REFERENCE

HAGISHIMA, S. and OHGAI, A. (1990), "A study of the urban landscape in the 19C european landscape paintings", Journal of Archit. Plann. Environ. Engng, ALJ, No 413
KAWASAKI, K. and HIRAO, K.H. (1994), "Settlement of the view-points and classification of the mountain-views in Kyoto", Journal of Archit. Plann. Environ. Engng, ALJ, No 462
Pusan Metropolitan City, (1996), Municipal white paper, Pusan Metropolitan government
Seoul Development Institute, (1994), Proposed Master Plan for the Urban Landscape Management for Seoul , 94-R-2