Influence of urban growth on urban heat island phenomenon around Jakarta, Indonesia: Insight from Depok and South Tangerang city

Economic activities in developing countries play a crucial role in national development. Jakarta, as one of the growing hubs in Indonesia, is influencing the urban planning of neighboring cities. Depok and South Tangerang City are among the cities that are most impacted by Jakarta’s growth. One of the impacts is that those cities are experiencing accelerated urbanization. This condition has caused a rise in the surface temperature of metropolitan areas, known as the Urban Heat Island (UHI) effect. This study used multi-temporal Landsat 8 OLI data to determine the land cover and Land Surface Temperature (LST) in the two cities. The relation between those variables was then evaluated using simple regression. The results show that common conditions occurred in those two cities.


Introduction
City has common social characteristics such as a high population of heterogeneous and materialistic socio-economic strata [1].Economic acitivities in developing countries have an important role for development and fulfilment of daily needs.The development of a city has both good and bad impacts not only on the city itself but also its surrounding area.Urban areas are famous for urbanization flows, this flow is a process of moving people from villages to cities with various purposes.According to Antrop [2], urbanization is not limited to the movement of people from rural areas to urban areas, but it also includes the increase in urban population and expansion of urban areas.
Depok and South Tangerang are both cities located near to Jakarta.As a satellite city, they have a strategic located results in a significant increase of population and area of built-up.According to [3], fast growth of built-up resulted in higher value of surface temperature in urban areas, called as Urban Heat Island (UHI).When UHI occur, the surface temperature in area urban is relatively higher compared to the sub-urban city [4].Therefore, based on the mentioned motivation above, the aim of this study is to investigate the relation between urban growth and UHI phenomenon.

Method
Our study areas are located in Depok and South Tangerang City.According to regional statictis (BPS), the total population in Depok was reported 1,813,612 people in 2010 and 2,056,335 people in 2020.For the same year, the population density was 9,055 people/km² and 10,267 people/km 2 in 2010 and 2020, respectively.South Tangerang is also a satellite city of Jakarta, located in the southwestern part of Jakarta.As a result of expansion from Tangerang Regency, this city becomes one of targets for urbanization.The population density of Tangerang Selatan was 8,694 people/km 2 and the number increased to 11,614 people/km 2 in 2020 (BPS).
The main data used in this study is Landsat 7 ETM+ and Landsat 8 OLI data.The time-series Landsat datasets of 2010 and 2020 were collected and pre-processed.Specifically, Landsat 7 data was corrected using a gap and fill technique.Those data were used to extract data of land cover and Land Surface Temperature (LST).Land cover information was obtained from supervised classification technique.LST was calculated using the mono-window Brightness Temperature algorithm.In the next step, UHI distribution was identified adopting the method from [5].Finally, the correlation between urban land cover and LST was conducted using a regression test.Figure 1.shows the overview of stages taken in the study.The distribution of average temperature of different land cover in Depok and South Tangerang is provided in Table 1.The results show that there was an increasing trend of temperature from 2010 to 2020 in both study areas.

Influence of urban growth on LST
The increasing area of built-up in Depok was 33.89 Km 2 with the increasing temperature of 4.2℃.Whereas in South Tangerang, the increment of built-up was 58.19 Km 2 and a difference in average temperature of 2.05℃.The correlation of this urban growth on LST was tested using ANOVA (Figure 4).The correlation (R) values were 0.636 and 0.551 for Depok and South Tangerang, respectively.This means that there is a positive correlation between area of urban and surface temperature.Figure 5 shows the results of the simple linear regression.For Depok, the resulted formula was y = 3.865 + 0.0007 x, and for South Tangerang the formula was y = 1.542 + 0.0005 x.This implies the positive relationship between the urban growth with the higher temperature in both study areas.

Urban Heat Island
Urban Heat Island is something phenomenon in urban areas where temperature surface warmer compared to the surrounding area [6].Table 2 provides information on the UHI threshold values.The spatial distribution of UHI is presented in Figure 6.According to Figure 6, UHI phenomenon in Depok was dominantly located in the central area for both year of 2010 and 2020.A different pattern was shown in South Tangerang that UHI appeared in the eastern part of the region.This indicates that the UHI distribution follows the spatial pattern of built-up area.

Conclusion
Depok and South Tangerang are both the satellite city of Jakarta.In this study, we analysed the correlation between the urban growth and land surface temperature in both regions.Our findings show that there was an increasing trend in temperature in Depok and South Tangerang from 2010 to 2020.It was also found that the larger urban area positively correlated with the higher temperature.It is resulted that the UHI pattern in Depok was dominantly in central area, while in South Tangerang the UHI was located in the eastern part near to Jakarta.This condition implies the similar spatial pattern between UHI and built-up area distribution.

Figure 1 .
Figure 1.Flow chart of the study.

Figure 2
Figure 2 presents the LST distribution in Depok and South Tangerang.LST in the study area was classified into five levels.According to the results, the temperature surface of Depok and South Tangerang experienced increasing trend.In 2010, Depok was dominated by class II (22.6-25.1℃),but the temperature increased to the class IV in 2020.Similarly, in South Tangerang the surface temperature increased from class III to class IV from 2010 to 2020.

Figure 3
Figure 3 shows the distribution of LST in 2010 and 2020 in Depok and South Tangerang.As shown in the figure, the surface temperature in Depok experienced a significant change from 2010 to 2020.In 2010, the high temperature class was only appeared in the central area.However, the high class was spread to almost all area of the region in 2020.In South Tangerang, we can see a quite different spatial pattern of the temperature class.In 2010, the high class of temperature was identified in the eastern part which is near to Jakarta.In 2020, the high class of temperature covered most region in Tangerang Selatan.

Figure 3 .
Figure 3. Land Surface Temperature distribution of Depok and South Tangerang in 2010 and 2020.

Figure 4 .
Figure 4. ANOVA results between built-up area and land surface temperature.

Figure 5 .
Figure 5. Regression results between built-up area and land surface temperature.

Table 1 .
Average land surface temperatures of different land cover in 2010 and 2020.

Table 2 .
UHI values of Depok and South Tangerang in 2010 and 2020.
Figure 6.UHI distribution in Depok and South Tangerang.