Analyzing thermal characteristics of urban streets using a thermal imaging camera : A case study on commercial streets in Seoul

With continuing city growth and global warming over the past decades, the urban heat island (UHI) effect, referring to the phenomenon that ambient air temperatures in cities are observed higher than in rural areas, has become a serious threat to urban populations. Impervious surfaces, buildings with low-albedo materials and the lack of vegetated areas are the major causes of poor urban thermal environments especially during the summer. Previous research has focused primarily on thermal behavior of an individual building unit. Only few studies consider the impact of street-scale thermal environments on pedestrian thermal comfort. The purpose of this study is to analysis thermal characteristics of various physical elements on urban streets using an thermal imaging camera and to present policy implications for improving pedestrian thermal comfort. The present study estimates street-scale thermal environments of three major commercial streets, Garosu-gil, Serosu-gil and Narosu-gil, found in Seoul, South Korea. The field measurements were conducted during both daytime (13:00 – 14:30 h) and nighttime (19:00 – 20:30 h) in June in order to investigate changes in urban surface temperature across time. The results show that street trees are the most effective mitigation strategy to reduce surface temperature. When it comes to building use types, the highest surface temperature is measured generally at restaurants. Building façades which are dark-colored or partially covered with metal contribute to high surface temperature, similarly, the temperature of artificial turf or wooden decks on urban street is significantly high.


Introduction
With continuing urbanization and global warming, thermal environments in cities have become serious issues over the last decades.In recent years, the summer air temperature has been constantly increasing, moreover, more and more citizens are threatened by Urban Heat Island (UHI) effects and the tropical night phenomenon.These phenomena negatively affect not only thermal comfort but also life satisfaction, social life, or mobility of people, eventually.In relation to this, many scholars from different parts of the world have tried to solve the issues mentioned above by using UHI mitigation strategies to lower urban air temperature directly or by using urban elements (e.g., street trees) to achieve shading effects and to create better walking environments.Even though a number of studies on UHI effects in densely built-up areas have been carried out, most studies conducted in South Korea primarily focused on buildings and used thermal simulation tools.In other words, there are only few studies which measure thermal characteristics of the urban built environment and analyze them.The present study investigates effects of street trees, pavement materials, building use types, and building materials on road surface temperature in Seoul, the capital of South Korea, using thermal imaging cameras.The temperature differences mainly caused by the amount of solar radiation reaching the ground surface across time is also addressed by analyzing data during daytime and nighttime.Thus, the aim of this study is not only to suggest better urban forms and materials for buildings and pavements in respect of thermal environments in cities but also to propose implications for improving outdoor thermal comfort at pedestrian level by analyzing thermal characteristics of urban elements.

Literature review
The effects of urban morphology or fundamental elements of urban form (e.g., buildings, streets, etc.) on thermal environment in cities have been analyzed in previous studies.However, only few studies analyze urban surface temperature using a thermal imaging camera and there is almost no study which explores the thermal environment especially at pedestrian level, which is a relatively more microscale analysis.Hwang et al. (2014) investigated urban elements which can negatively affect thermal environments using a thermal imaging camera.According to their results, urban surfaces, for example, road pavements, green areas, shading effects, building façade, and materials of signboard, play an important role in UHI effects.In most cases, dark pavements, such as asphalt, concrete, or granolith paved surfaces, contributed to high temperature, whereas paved roads with vegetation were significantly cooler by 10℃ than other paved roads or structures nearby because of trees which shield the direct solar radiation.Construction materials were also addressed.The surface temperature of building façade made of artificial construction materials is generally higher than that of green areas since artificial construction materials has lower albedo values and the building façade is exposed to direct solar radiation during daytime.They found that several major factors causing UHIs, for example paved roads, signboards, and building walls, etc., however, the study addressed thermal behavior of only a few construction materials.Cho and Lim's (2011) study performed in Seoul, South Korea analyzed the mitigation effects of urban road vegetation on the road surface temperature.The study found that the roadside land use affects the road surface temperature.For example, if the roadside land is used as green areas or open spaces, the road temperature tends to be lower than other land-use patterns.Similarly, the surface temperature of artificial pavements (e.g., concrete, interlocking blocks, etc.) is higher than that of natural pavements (e.g., green spaces, bare soil, etc.) during not only daytime but also nighttime.The study, however, have limitations that various types of pavements were not addressed.Nam (2007) explored thermal environments in environment-friendly apartment complex to analyze mitigation effects of vegetation, water bodies, and albedo values based on surface temperature measurements obtained from a thermal imaging camera.Measurements were performed at 14:00 h, the hottest time of a day.The results showed that concrete blocks had the highest surface temperature (37.4℃) followed by vegetation (27.9℃) and water bodies (27℃).Nam, however, also had his focus on the mitigation strategies of only three-limited urban surfaces, concrete blocks, vegetation, and waterbodies.Yoon (2009) investigated thermal environments in Asan, a medium-sized city in South Korea, across two different time periods which were autumn and winter season.In accordance with previous studies on urban surface temperature, the results of the study showed that the surface temperature of asphalt pavements was higher than that of vegetated areas or sidewalk blocks around an artificial fountain during both seasons.The study also found that there is a strong relation between urban surface temperature and ambient air temperature, for example, high surface temperature causes warmer air temperature in most cases.Despite such findings, the study still had limitations that it only considered limited urban surfaces found in the study area.Ahn et al. (2010) monitored UHI effects and explored the relationship between building surface temperature and ambient air temperature using thermal imaging cameras installed at 5 different locations in Seoul metropolitan areas.Analyzing climate data collected from a representative sample of the hottest time of a summer day, August 14, 2009 at 14:00 h, air temperature in Seoul was 32.4℃, while, surface temperature observed at the 5 locations varied from 46.0 to 62.1℃.The results showed that building surface temperature was significantly higher (by 22℃) than air temperature during a hot summer day.The scholars concluded that low albedo values of building materials, such as concrete, and anthropogenic heats generated by the use of air conditioning increased the building surface temperature.Even though this study analyzed surface temperature at 5 locations in Seoul, they addressed only few factors which affected urban surface temperature.Therefore, more various factors should be considered in further study.Elhinnawy's study (2004), which was conducted at a building of Arizona State University, evaluated the thermal behavior of building envelope.The study found that surface temperature of east façade varied depending on different albedo values.For example, the surface temperature reduction of 3℃ was observed for every 0.05 rise in albedo at 9 am when the highest temperature values were measured.Therefore, surface albedo

Proceedings of 2017 International Conference of Asian-Pacific Planning Societies
values can be effectively applied to urban planning process as a powerful UHI mitigation strategy.Despite of their finding, this study has limitations that it only focuses on one specific building which makes it impossible to explore the changes in the temperature of building façade made of the same material at different locations.
Similarly, Elhinnawy (2004) analyzed the contribution of building surface temperature to urban thermal environments using a thermal imaging camera.The study found that surface temperature of a concrete balcony was the highest among all the other building elements and green areas were observed relatively cooler.Further discussions of the effect of building orientation on surface temperature were also addressed.In most cases, west and south walls were observed warmer than east and north walls since west and south walls were exposed to direct solar radiation.Yamazaki et al. (2009) carried out research on UHI effects in Tokyo, Japan using thermal images.The results showed that surface temperature of asphalt road pavements was significantly higher.The study also found that road-side trees and road-center plants were cooler than road pavements or building walls.They also used airborne thermal sensors to investigate mitigation effects of urban rooftop greening as well.The results indicated that the highest temperature of tiled roof was 46.4℃, whereas, the vegetated roof surface was significantly cooler.Furthermore, mitigation effects of different vegetation types were addressed in the study.For example, vegetated rooftops covered with the cactus were observed with high temperature even though the rooftop had a high normalized vegetation index (NDVI).They concluded that it was because the dry surface of cactus affected thermal environments negatively unlike other vegetation.
Reviewing the previous studies, we develop the current study focusing on the followings: (1) We consider more various construction materials to investigate their effects on the urban thermal environment.Moreover, thermal behavior or urban elements with and without shading is also addressed.( 2) The current study analyzes thermal environments of three adjacent urban streets in Seoul in order to investigate how thermal behavior of urban elements differs across varying urban street patterns under a controlled climate condition.(3) The majority of previous studies gives only a very simple understanding of thermal images because thermal imaging cameras are generally used to prove the result reliability of a certain simulation they developed.However, we analyze thermal images of urban streets in a more detailed way to suggest a better urban street form to improve thermal comfort at pedestrian level.

Description of study areas
The three major commercial streets, Garosu-gil, Narosu-gil, and Serosu-gil, located in Seoul, the capital of South Korea, are selected as study areas.These streets are adjacent so they are considered under the same climate condition.Various characteristics are found on each street even though they are all commercial streets.More specifically, various building use types (e.g., restaurant, coffee shop, cosmetic/clothing store, residential building, etc.), density of street trees, road width, and building shape make the distinctive characteristic of street (Fig. 1).

Proceedings of 2017 International Conference of Asian-Pacific Planning Societies
Garosu-gil, a main commercial street in Sinsa-dong, Gangnam-gu, Seoul, has sidewalks along a two-lane road and is famous for its characteristics.Street trees and a lot of stores, such as art gallery, designer boutique or small coffee shop, give the very unique atmosphere to the street.Compared to other study areas, more art galleries and retail shops are found on Garosu-gil and the two-lane road causes driving at relatively higher speed.Regarding to building configurations, different construction materials and building heights offer interesting variations on building design.There are almost no building setbacks, therefore, a simple building line is found around the study area and the width of sidewalk is also constant.
With the expansion of Garosu-gil, Serosu-gil where relatively less stores are located than Garosu-gil has been developed over the past years.Small clothing stores, restaurant, and offices are mainly found on the street.Unlike Garosu-gil mentioned above, there are almost no street trees seen on the street and people drive a car with much lower speed because of the narrow road, one-lane road.In most cases, multipurpose buildings with offices at upper stories and retail shops at lower stories, are found in the study area.Some of the stores leave public open space for pedestrians using building setbacks.
Similarly, Narosu-gil has also gotten popular as a result of a huge success of Garosu-gil.Newly opened small shops and existing buildings, which are mostly residential and educational buildings, coexist on Narosu-gil and form the distinctive characteristic of the street.Even though trees are not seen along the road, there is a small amount of vegetation near residential or educational buildings.Regarding to car driving, the narrow width of the road in Narosu-gil, one-lane road, makes driving speed lower.In addition, only a few car trips are found in the study area because of its residential characteristics.To analyze the surface temperature of urban elements, such as building use type, construction materials, width of road, street trees, etc., under a controlled climate condition, three adjacent commercial streets are selected as study areas.Even though the streets are approximately 700m long, however, we define 350m of each street as study areas in order to avoid having any influence from eight-lane roads, Dosan-daero and Apgujeong-ro, which are connected to the streets.

Methodology
High resolution thermal imaging cameras, FLIR T430sc and T440, are used to measure urban surface temperature.Using analysis software, ResearchIR Max+HSDR, collected thermal images from each study area are analyzed.Climate data from Korea Meteorological Administration and Kweather data base are used as input data for the thermal imaging cameras.The pilot test is conducted from 13:00-14:00 h on June 5, 2017, in order to see if there are temperature differences between two cameras.Based on the results, the surface temperature from T430sc, a latest model with better data reliability, is used as a standard to calibrate results from T440.Thermal images are taken at the frontal direction since it offers more accurate thermal data and is less blocked by pedestrians walking around.Field measurements are performed on June 16, 2017 and June 22, 2017.In addition, it is carried out during daytime and nighttime, defined as 13:00-14:30 h and 19:00-20:30 h, respectively.The characteristics of urban surface temperature during both time periods, therefore, is also addressed by analyzing average values at each time.

Proceedings of 2017 International Conference of Asian-Pacific Planning Societies
Two thermal imaging cameras are used for each street.Every building on the street is taken including the sidewalks adjacent to the building.To obtain more accurate surface temperature of sidewalk, the other camera is used to take thermal images of sidewalk in a short distance, simultaneously.(Fig. 2) We analyze the surface temperature of 5 spots, at the middle height of the exterior wall of the first floor, bottom part of building façade, side walk adjacent to building, sidewalk adjacent to road.The spots are selected since they are considered as the major factors affecting pedestrian thermal comfort.Therefore, the thermal characteristics of urban elements at 5 different spots are investigated.

Comparative Analysis of Garosu-gil, Narosu-gil, Serosu-gil temperatures
The photographing was performed twice.However, 6.16 was cloudy and not used for analysis, only 6.22 was used for final analysis.The basic climate for shooting at 1:00 pm is temperature: 29.1 ℃, wind direction: south wind, wind speed: 0.8m / s, humidity: 44% Wind speed: 1.9m / s, Humidity: 42%, Cloudy: clear.Figure 3 shows the mean values of the day and night temperatures of the three roads, which were measured using an infrared camera.It shows the temperature of five parts such as building, foundation, sidewalk(inside) , sidewalk(outside), road.Basically, the road is the highest temperature and the temperature of the building is low.In the case of the building, it is thought that the exterior of the first floor shop is often made of glass and there is a factor of lowering the temperature such as air conditioner.When comparing the daytime temperature of the three roads, Garosu-gil is the lowest and the Serosu-gil is the highest.When comparing the characteristics of three roads, Compared characteristics of three roads.In the case of Garosu-gil, the roadway is broad and produces natural shadows on the sidewalk through the trees.In the case of Serosu-gil, the roadway is narrow and there is no way to block heat because there is no avenue.In addition, the use of buildings is often a restaurant, the artificial heat emission of the building is high.In the case of Narosu-gil, the middle temperature is shown, but the shape of the road is similar to that of Serosu-gil.Trees are planted in some areas, so the median value is raised.In the case of night, the temperature of the Garosu-gil is the highest.The reason for this is that the movement of the vehicle and the person is concentrated and the artificial heat is high and the roadside tree is preventing the heat from being discharged to the outside.[5] In the case of road temperatures, Serosu-gil averaged 50 ° C and Garosu-gil averaged 40 ° C in the noon, but at night they converged to 33 ° C. In the case of building temperatures, Serosu-gil averaged 37 ° C and Garosu-gil averaged 32 ° C in the noon, but at night they converged to 29 ° C. In the case of the sidewalk, Garosu-gil and Narosu-gil tend to have a median value of the temperature of the road and the temperature of the building, but the Serosu-gil tends to be higher than the road.because Serosu-gil has a lot of materials that are thermally weak.

Analysis of Analysis of Garosu-gil temperature
In the case of Garosu-gil, the temperature is the lowest among three target roads.Characteristically two lanes and street trees can be cited, it is a road made only commercially.The shadow inside the sidewalk caused by the street trees was 76.44% on average.The number of photographs classified according to the use of the first floor belonging to the 350 m of the photographing range was 45, and analysis was carried out divided into 5 divisions in order to display the temperature difference for each block.In the case of Garosu-gil, the block with the highest temperature is the 1st block of the 5 th block , and its configuration is as shown in FIG. 4. The shadow occupied the lowest percentage at 60% in Garosu-gil.Usage is commercial, the building which used 60% or more of glass in the first floor part is the main material, black steel, brick, paint was poisonous.The average sidewalk temperature in the block was 37.47 ° C.
The block with the lowest temperature was the 2nd block and its configuration is shown in Fig 5 .Shadows were not the highest at 82.22% in the Garosu-gil, but showed higher shadows than the average.(The highest block is the 5th block, showing a shadow ratio of 88.88%.)Commercial use is in the first floor, and 80% or more glass is used in the first floor.The main material is white paint, block, and media facade.The average sidewalk temperature was 33.35 ℃, which was 4.12 ℃ lower than the highest block.
Thermal images classified by usage were extracted and analyzed for five low -temperature applications and five high -temperature applications based on the sidewalk.The temperature table is sampled three times and summarized in Tables 1 and 2. The samples with the highest temperature were those with a shadow percentage of less than 40%, commercial or open space, and high percentage of glass but with iron and panel points.In the open space, artificial turf is covered in the open space, and it is thought that the temperature is increased without decreasing the temperature.[6] The samples with the lowest temperature were 100% shaded, and commercial floors were glass floors with glass facades and digital facades and bricks.
At night, the temperature was unusual.The lowest temperature in the sample was higher than daytime.In the case of the highest sample, the temperature decrease was evident when the effect of sunlight was removed after sunset.As mentioned in the temperature analysis in the previous Section 4-1, at night, the sidewalk temperature of the Garosu-gil was higher than other two target roads.

Analysis of Narosu-gil temperature
In the case of Narosu-gil, Indicates the intermediate temperature in the three target roads.Characteristically, there are few lanes of vehicle donation, landscaping planting, you can mention that there are several locations, which is a mixture of commercial and residential.There is no street tree, there is no shadow of crying in general by reporting, but in some reports shadows are generated by landscaping trees.The number of photographs classified according to the use of the first floor belonging to the 350 m of the photographing range was 33, and analysis was carried out divided into 5 divisions in order to display the temperature difference for each block.In the case of Narosu-gil, the block with the highest temperature was the 5th block of the 5th block, and its configuration is shown in Fig 6 .It is a commercial, educational, and athletic facility.In the case of commercial, it was a building that used more than 80% of glass in the first floor.In the case of education and sports facilities, it was a front space or parking space.Commercial buildings were made of wood and stucco, and education and sports facilities were made of bricks and ceramics.The average sidewalk temperature in the block was 43.94 ° C.
The block with the lowest temperature was the 3rd block and its configuration is shown in Fig 7 .Usage is commercial and residential, in the case of commercial, it is a building using more than 80% of glass in the first floor part, and in the case of residential landscaping planting was planted.Bricks, stucco, trees were the main materials.The average pressing temperature was 6.19 ° C lower than the highest block at 37.75 ° C. Thermal images classified by usage were extracted and analyzed for the lowest temperature and the highest use based on the report.The three temperature tables are sampled and listed in Tables 3 and 4. The samples with the highest temperature were commercial, with high percentage of glass but with iron and panel points and wooden decks on the front.In the case of wooden decks, the temperature was abnormally high during the day in all roads.The samples showing the lowest temperature were the sidewalks covered with landscaping and the white sidewalk bricks.
At night, the temperature abnormally dropped from the wooden deck to below the existing sidewalk.The samples with the highest temperature during the day were the coolest at night.The difference between daytime and nighttime temperature of wood deck was 32.2 ℃.The lowest temperature difference in Narosu-gil was only 1.2 ℃ difference from the sidewalk under the landscape gardening material.W hite brick sidewalks showed a small temperature difference of 6.4 ℃.In the case of Serosu-gil, it showed the highest temperature among the three target roads.Characteristically, there are fewer lanes and the number of vehicles moving than Garosu-gil, but parking garages are gathered and there are a lot of stopped vehicles.There are many restaurants different from the Garosu-gil, and it is a commercial road.It is also characterized by no street trees and no shadows that cry out in reports.The number of photographs classified according to the use of the first floor belonging to the 350 m of the photographing range was 45, and analysis was carried out divided into 5 divisions in order to display the temperature difference for each block.In the case of a Serosu-gil, the block with the highest temperature was the 4th block of the 5th block, and its configuration is as shown in FIG 8.The application was commercial and the proportion of restaurants was high.We used 40 to 100% of glass according to the characteristics of the shop, the materials diversified, such as bricks, paint, punch steel etc.The average pressing temperature in the block was 55.

Proceedings of 2017 International Conference of Asian-Pacific Planning Societies
The block with the lowest temperature was the third block and its configuration is shown in Fig 9 .There are commercial, parking lot and stairs of commercial and piloti.In the case of commerce, it is buildings that used more than 80% of glass in the first floor part.In the case of the parking lot or the staircase, it was empty space and the shadow was made on the piloti part, and it did not affect the artificial heat in the front sidewalk.Ceramics and stucco were the main materials, and some buildings made shadows through the shade.The average sidewalk temperature was 38.48 ℃, which was 16.94 ℃ lower than the highest block.
Thermal images classified by usage were extracted and analyzed for the lowest temperature and the highest use based on the report.The three temperature tables are sampled and summarized in Tables 5 and 6.The highest temperature samples were restaurants and shops, with high percentage of glass but with iron and panel points, with wooden decks and artificial turf in the front.In the case of wooden decks, as noted earlier in the Narosu-gil, the temperature was abnormally high during the day on all roads.In addition, the artificial turf showed higher temperature than general sidewalks and was not thermally efficient.The samples with the lowest temperature were shaded through the shade, where the percentage of glass in the store was 100%.
At night, the temperature abnormally dropped from the wooden deck to below the existing sidewalks.The artificial turf area also appeared similar to the general sidewalk temperature.The daytime and nighttime differences of artificial turf and timber decks in Serosu-gil were 26.1 ° C and 34.9 ° C respectively.The area with the lowest temperature difference was 10.3 ° C below the sunshade.7.In the case of Garosu-gil, the presence or absence of trees tends to influence the temperature change and the other factors tend to be buried, except for Narosu-gil and Serosu-gil examined.In the case of buildings, in the case of commercial buildings, we used glasses most often to showcase our products.Not only was the glass susceptible to the influence of the room temperature lowered to the air conditioner, it passed heat and became the lowest temperature factor.In the case of media coverage followed by white bricks and paint, in the case where the shadows were bordered, the lowest continued white lineage report was low.The factor which appeared at the highest temperature was where the building emphasized the railroad, a panel of black lines and bricks were used, and in the case of the press was made of trees and artificial turf.In the case of trees and artificial turf, If you absorb the heat and see the temperature rise abnormally in the daytime, it can be regarded as the most thermally insufficient material.

Conclusion
People's thermal comfort in the urban environment, the most obvious part is the report.We examined the analysis results of three target roads investigated in order to investigate the press temperature which changes according to the surrounding environment.The avenue has many effects such as landscape and flood prevention, but the shadow effect is the best among them.It was found that when compared with Serosu-gil without street trees as well as landscape planting, it shows an effect of a difference of 14 ° C. on average.Also, even in the case of a use, it was possible to know that the temperature rises and affects the temperature of the press in the case of a space using fire like a restaurant.In the case of buildings that looked at buildings and sidewalk factors that affected the temperature, the glass and white family exterior materials lowered the temperature.
In the case of sidewalk, the shadow and the white line of the sidewalk lowered the temperature.As a factor to raise the temperature of the press, the building was a black type exterior material, steel, and the report was an artificial turf and a tree.It turns out that Albedo, which is an indicator that the material that looked at absorbs heat and reflects, directly or indirectly affects the temperature of the report.Summarized in this research is as follows.
-The shadow of the roadside tree lowers the temperature of the sidewalk during the summer.
-The sidewalk in front of a building that uses fire such as restaurants is higher than other places.
-The sidewalk changes temperature according to the surrounding building and the albedo of the sidewalk.
-In the case of buildings, glass does not absorb heat when used, so do not increase the sidewalk temperature.-In the case of sidewalk, when using artificial turf and wood, heat was absorbed and the temperature was high.
This study largely derives five conclusions.However, for accurate measurement, it is necessary to increase the amount of sample through more measurements and pay attention to the weather change or cloudiness in the middle of shooting.The presence of a person at the time of shooting and the movement of the vehicle are also considered as variables.Also, the photographing equipment is a thermal image camera which the person directly measures, and its analysis program may cause many errors in setting the temperature measurement range arbitrarily.In the case of shadows and artificial lawns and trees in the sidewalk, the results are clear, but other results need to be analyzed more accurately, and it is necessary to retest at a different site than the street.The reason for this is that in the case of the three target sites this time, the first floor of most buildings on commercial roads was used as glass and other materials were used only for the building structure and decoration factors.In terms of the materials of buildings, it should be handled through comparative press temperature analysis on the materials of buildings and residential fences where less glass is used in the future.Through this study, it is thought that the temperature controlled by factors controlling the factors of the thermal environment appearing higher than the average temperature can be greatly reduced when walking in the city.Currently, it is thought that if we control the thermal environment factors to distance development projects implemented in other cities in Korea and encourage street trees, we can create a more effective city and street environment.

Figure 3 .
Figure 3. Surface temperature analysis of study areas.

6 )
Proceedings of 2017 International Conference of Asian-Pacific Planning Societies