Although Seoul is the third largest city in the world (1982), only one investigation has been performed on its heat island (Lee and Kim, 1979). However, they made clear the characteristics of horizontal temperature distribution on calm and clear evenings by case study.
Then, the author focused on (a) the features of the heat island under various weather conditions and times and (b) the correlation of various indices which affect the urban heat island form at ions. These indices include land-use types, the ratio of impermeable surface coverage related to them, and the concentration of air pollution (SO
2).
The study area includes Seoul (8, 400, 000 population) and its four satellite cities, Seongnam (380, 000), Kwangmyung (150, 000), Euijeongbu (140, 000) and Shindo (60, 000). Seoul is located in a basin surrounded by mountains approximately 250-830 m high. The urban area of Seoul extends radially along four main streets and the Han River flows through the southern part of the city (Fig. 1.).
Temperature observation was made by an automobile and other meteorological data such as, cloud cover, wind direction and wind speed were provided by the Seoul Central Meteorological Office. The series of observation by car were repeated along three routes A-B, C-D, and E -F (shown in Fig. 1.), eleven times during each of three time periods: early morning (04:00-05:30), day time (14:00-15:30), and night (22:00-23:30) from June to August, 1982, for a total of 99 runs. The number of observation points along the three routes totaled 134. Table 1 describes the experimental program.
For the illustration of mean horizontal temperature distribution under typical condition, limited sets of observation which were carried out on clear and calm nights were chosen, and the temperature deviation at each point from the mean route temperature (T-T
route) are calculated as shown in Fig. 2. Positive deviation approximately corresponded to the built-up areas within a 10 km-radius from the center (CBD) of Seoul and that in satellite cities. On the other hand, cool regions appeared in the rural area outside of Seoul. Especially, a cool pocket was found over an island in the Han River.
The higher temperature deviation area had the lower standard deviation and the lower temperature deviation area had the greater standard deviation in most cases. Variations of temperature deviations caused by weather conditions (cloud cover) and different times in Seoul CBD were less than those in suburban area. Gradients of the temperature deviation tended to change significantly at the boundaries between the built-up area and rural region (Figs. 3 and 4).
The heat island intensity (difference between the highest urban temperature and the lowest rural temperature,
ΔT
u-r) on clear days (cloud cover<7/10) was greater than that on cloudy and rainy days (cloud cover>8/10) and it was the greatest from night to early morning and less in daytime. The values of the maximum heat island intensity of Seoul were 7.1°C (the maximum value) and 3.7°C (mean for clear night observations) (Table 2).
Wind speed was the main determining factor for the heat island intensity and it affected the distribution of the heat island (Fig. 5). The heat island intensity was destroyed above a critical wind speed of 11.8 ms
-1 for Seoul and decreased with decreasing urban population (Table 3).
To analyze the relationship between surface coverage based on the land-use types and the urban heat island phenomenon, an analysis was done by using a multiple regression analysis. The study area was divided into a 0.5×0.5km square grid and then the areal ratio (%) of the land-use types within each grid was measured. The prevailing land-use types in the city outskirts, i.e, green belt, agricultural area and water surface played an important role in lowering the urban temperature.
View full abstract