RS & GIS FOR DOSASTER MANAGEMENT

Assessment of Surface
Urban Heat Island
Monitoring the impact of rapid urbanization on
Lahore, Pakistan
Unza Ahmed 190601015
Introduction:

Global population is increasing at a rapid pace and people are migrating to urban areas (Mitchell
and Moss, 2012), which leads to extensive. Fifty percent of the world’s population are urban
residents, which is expected to increase by 70% (2.5 billion) between 2010 and 2050 with 90%
contribution from Asia and Africa), and it is projected that more than 50% of world population
will live in Asian Cities serve as a significant source of economic growth and contribute about
60% of global GDP. But, they use more than 60% of energy and humans produce almost 70% of
greenhouse gases (GHGs) (Un-Habitat, 2018). Therefore, the process of urbanization is affecting
the natural environment, which influences the effects of GHGs, leading to global warming and
climate change which is one of the major issues in the current situation. Increasing population in
cities expands the boundary and exerts pressure on rural areas and replaces natural resources
(i.e., agriculture, vegetation, forest). Thus, urbanization is a major reason to alter the changes in
Land Use Land Cover (LULC) and also a primary source of predictor to identify the LULC
change which imbalances the surface energy and takes a way to higher surface temperature. In
the past few decades, urban green spaces significantly reduced in the world regardless of its
significance to balance the LST in the urban environment The loss of vegetation at the cost of
urbanization causes extensive changes in UTE, which increases the LST in urban buffer as
compared to rural buffer, leading to SUHI phenomena and bring different environmental issues,
that is, land degradation, ecosystem problem, heat waves, urban flash floods, increasing
precipitation, and temperature in urban . Roughness and surface reflectance properties of
different LULC vary from each other and contribute differently to LST because of the
uniqueness in the absorption and radiation energy of LULC. Increase in urbanization at the cost
of vegetation loss has been experienced by many regions around the globe. Urban climatology
was first studied and observed by Luke Howard in 1833 (Howard, 1833). The phenomena of
UHI or SUHI indicate the positive difference between temperature of rural and urban land due to
the results of urbanization. Usually, it is calculated with the difference of air temperature of
urban met-station to air temperature of rural met-station, although due to scarcity of
meteorological stations in developing countries, it is very challenging to accurately estimate the
UHI. Therefore, advancements in remote sensing (RS) technology have overcome this problem,
allowing us to calculate the SUHI with surface temperature difference between urban buffer and
rural buffer. Now, researchers are carrying out research to quantify the impact of urbanization on
LST and effects of UHI on UTE at the regional and global scale .RS technology offers the
opportunity to observe the LULC changes and study the influence of urbanization on LST with
high, moderate, and low spatial and temporal resolution satellite images. In this study, we have
utilized the moderate spatial and temporal resolution satellite (Landsat) which was first launched
50 years ago to provide free-of-cost data to study various applications, that is, land monitoring.
The long available dataset can offer a better insight of the past and present and predict future
changes on land. To evaluate the LULC change using satellite data, there are various models,
that is, supervised and unsupervised image classification, machine learning algorithms,
objectbased image classification algorithm (OBIA), and many more. In this study, we have used
the supervised image classification method assisted with the maximum likelihood classifier
algorithm which has been successfully used in many studies before in Lahore/Pakistan (Imran
and Mehmood, 2020), Ahmedabad and Jaipur/India (Mathew et al., 2018), Guangzhou/China
(Xiong et al., 2012), Bangkok/Thailand (Srivanit et al., 2012), and Alborz/Iran (Solaimani et al.,
2010). All aforementioned studies prove the usefulness of the research related to LULC changes
with reference to regional climate change in terms of UTE, LST, and SUHI. In the last few years,
urban area in Lahore expanded exceptionally and many housing societies were formed which has
resulted in a loss of vegetation cover substantially; ultimately, it has changed the urban
environment (Nespak, 2004; Shah and Ghauri, 2015), and this brings lots of health issues for the
local residents, that is, discomfort, difficulties in respiration, and heat stroke (Kotharkar et al.,
2018; Zuhra et al., 2019). As mentioned above that Asian cities will have 50% of world
population; therefore, it is necessary to study the impact of rapid urbanization on LST and SUHI
and assess how cities are performing to influence the climate change, and it is also necessary to
highlight the significance of sustainable cities and communities and contribute to UN sustainable
development goals. The purpose of this study is to 1) evaluate the LULC change for the years
1990, 2000, 2010, and 2020; 2) examine the impact of urbanization on LST; and 3) analyze the
surface urban heat island intensity (SUHI).

Study Area:

Lahore, Pakistan's second-largest city, serves as the study area for this research. Located in the
province of Punjab, Lahore has experienced exceptional urban expansion in recent years. The
city has witnessed rapid population growth and extensive urbanization, resulting in significant
changes to its land use and land cover (LULC) patterns. The expansion of urban areas in Lahore
has led to the formation of numerous housing societies, which, in turn, has resulted in the
substantial loss of vegetation cover. This loss of green spaces has had a profound impact on the
city's urban environment. The alteration of LULC patterns and the subsequent reduction in
vegetation cover have brought about various health issues for local residents, including
discomfort, difficulties in respiration, and an increased risk of heat stroke.

Lahore's prominence as a study area is further emphasized by its significance in the context of
Asia's urbanization trends. With the projection that more than 50% of the world's population will
reside in Asian cities, understanding the impact of rapid urbanization on LST (land surface
temperature) and the surface urban heat island (SUHI) phenomenon in Lahore becomes crucial.
Such research contributes to the broader goal of highlighting the importance of sustainable cities
and communities in achieving the United Nations' sustainable development goals.

Given the availability of long-term satellite data from the Landsat program, this study utilizes
remote sensing technology to evaluate the LULC changes in Lahore over multiple decades. By
analyzing the impact of urbanization on LST and SUHI, the research aims to provide valuable
insights into the urban climate dynamics of Lahore and its implications for regional climate
change.

Data Acquisition and Processing

We have investigated the land use land cover (LULC) and land surface temperature (LST) of
Lahore and acquired four Landsat satellite images for the years 1990, 2000, 2010 (Landsat 5,
Thematic Mapper), and 2020 (Landsat 8 Operational Land Imager). The Landsat images of
Lahore with path/rows of 149/ 38 were downloaded from United States of Geological
SurveyCenter for Earth Resources Observation and Science (USGSEROS)
(https://earthexplorer.usgs.gov/), with less than 10% cloud cover for the month of March (Table
1) because usually winter season (fog) and summer season (heavy clouds due to monsoon) are a
big hindrance to accurately classify the images. We have used almost cloud-free images in this
study, and details of the images are given below in Table 1. Landsat data are freely and easily
available to classify the LULC and calculate LST and SUHI (Tariq and Shu, 2020). The Landsat
satellite image has 30 m spatial resolution and completes the revolution of the earth in 16 days
on average. Before using the Landsat satellite images for LULC classification, the images
underwent radiometric, atmospheric, and topographic correction in the semiautomatic
classification (SCA) plugin tool in QGIS software to avoid the erroneous classification of the
land. Later, the Landsat bands were stacked together with necessary band combination, that is,
red, green blue, near infrared, and short-wave infrared. The thermal band was used for the
estimation of LST and SUHI. The administrative boundary of Lahore was obtained from Nature
Earth website.

Supervised Image Classification and Accuracy Assessment:

LULC information can be obtained from multiband satellite images using image classification
techniques (Alam et al., 2020). There are various methods to classify the image, that is, machine
learning, object-based image classification algorithm (OBIA), and supervised and unsupervised
classification. In this study, we have used the supervised classification method with the
maximum likelihood classifier algorithm which require training sites from the user (Zubair Iqbal
and Iqbal, 2018). Thus, we have trained the classifier with enclosed polygon training sites to
capture the spectral signature for each LULC class. Using training sites, the maximum likelihood
classifier algorithm was run to classify the image into four different classes: built up land, water,
barren land, and vegetation cover. LULC maps were obtained for the years 1990, 2000, 2010,
and 2020 from Landsat 5 TM and Landsat 8 OLI using the maximum likelihood classifier
algorithm. The advantageous property of maximum likelihood classifier is parameter
interpretation based on prior knowledge of the expert. This method is very easy to use and
implement (Cheng et al.). Accuracy of the LULC classification is one of the significant parts
which can be determined by user and producer accuracy, overall accuracy, and kappa co-
efficient in confusion matrix (Congalton, 1991; Van Oort, 2007). If the errors are uninvestigated,
the LULC map will lose its reliability for decision making (Shao and Wu, 2008). Producer and
user accuracy are the number of accurately classified pixels over the total number of pixels in a
particular class, while the number of correctly classified pixels divided by the total number of
reference pixels is known as overall accuracy. A total of 200 random points were taken using a
random point tool in ArcGIS, with 50 points in each class to assess classified images with high-
resolution Google Earth image.
 Classes 1990 2000 2010 2020
Built up 411.9 5981 615.2 835.3
Vegetation 1,089.3 1,021.6 1,015.6 858.4
Barren Land 289.2 171.8 166.2 104.6

4
Land Surface Temperature Calculation:

LST has been calculated from the thermal band of Landsat 5 TM (1990, 2000, and 2010) and
Landsat 8 OLI (2020). Landsat 8 has two thermal bands (band 10 and 11), while Landsat 5 has
only one thermal band (band 6), though we have used band 6 and band 10 of Landsat 5 and 8,
respectively, in this study. Band 11 of Landsat 8 was avoided due to the substantial calibration
issues identified by USGS (Avdan and Jovanovska, 2016). The month of March is a good fit to
calculate the LST of the study area as we have observed very less intervention of clouds and fog.
Digital Number to Spectral Radiance (Lλ) Landsat 5 and 8 thermal bands are in the form of
digital numbers (DNs); therefore, we have converted the digital number (DN) of the thermal
band to spectral radiance (Lλ) using Eq. 1 and Eq. 2.. The Rozenstein et al. (2014) method has
been followed in this study to convert the DN value to Lλ for Landsat 8 thermal band using Eq. 2
below. L ML p QCAL + AL. (2) Multiplicative rescaling (ML = 0.0003342), digital number
(QCAL = band 10), and additive rescaling band 10 (AL = 0.1). Spectral Radiance Conversion to
Satellite Brightness Temperature We have followed Eq. 3 for the conversion of the spectral
radiance to satellite brightness temperature. T is the satellite brightness temperature in degree
Celsius (o C), and the value of Lλ has been obtained in the previous step. K1 and K2 are the
constant values that depend on the type of sensor. K1 and K2 values for Landsat 5 TM are
607.76 and 1,260.56, respectively, while for Landsat 8, the values are 774.89 and 1,321.08. The
LULC data were split into two subsets urban buffer and other LULC classes except water class
because it contains high capacity of heat; consequently, it can over and underestimate the SUHI
at day and nighttime LST calculated for both urban and rural buffer to eventually calculate the
SUHI (Figure 2).

Land Surface Temperature Variations from 1990 to 2020:

The land surface temperature (LST) pattern of Lahore was calculated for the month of March
from 1990 to 2020. The results of LST are presented in Figure 5 below. The results revealed that
maximum temperature constantly increased, although there was only 3.2 C change from 1990 to
2020. Figure 5 also shows indicating that in 1990 the highest surface temperature was in the
central part of Lahore, but in 2010 and 2020, the maximum LST was spatially spread to the south
part of the city as well. The LST and built-up area increased parallel to each other from 1990 to
2020 (Figure 3 and Figure 5). Due to the urbanization process, the impervious surface absorbs
the heat rapidly but the emission process is very slow which is the reason of high surface
temperature apparent in this study. Built-up land contribute majorly in rise of LST (Bokaie et al.,
2016) since it is observed that in 1990 when built-up land has 22.9% of total area, the mean LST
of Lahore was 21.42 ± 1.36°C which is increased to 22.1 ± 1.9°C (2000), 22.7 ± 2.1°C (2010),
and 23.05 ± 1.6° C (2020) when urban area increased to 33.2%, 34.2%, and 46.4% in 2000,
2010, and 2020, respectively.
These results confirm the direct relationship of urban land and LST which is in line with a study
conducted by The environmental protection agency has also reported that surroundings of the
city center will have 1–3° C less temperature to urban land due to high absorption of solar
radiation by impervious surfaces (Mathew et al., 2018). Consistent rise in LST with urban area
leads to higher surface temperature than other LULC classes which exhibit the impact of urban
heat island (UHI). A consistent increase has been observed for surface UHI intensity. Changes in
LST with constant changes in thermal environment imply the impact of UHI (Srivanit et al.,
2012) which has been observed in this study. SUHI has been calculated for Lahore using the
LST difference of urban and rural pixels as presented in Table 4. An increasing trend of SUHI
has been observed in Lahore. The urban temperature in 1990 was 1.72 C high as compared to the
rural area which increased to 1.91°C, 2.22°C, and 2.41°C in 2000, 2010, and 2020, respectively
(Table 4). Urbanization is considered a main source of increase in LST which ultimately causes
SUHI (Hassan et al., 2021). Therefore, it is necessary to increase the plantation and green space
in urban environment because it can significantly decrease the temperature and impact of SUHI.
Relationship Between NDVI and LST NDVI and LST have an inverse relationship, that is, as the
vegetation increases the temperature decreases and vice versa. The correlation trend line is also
indicating the strong inverse relationship between two variables (Figure 6). The correlation co-
efficient (R2 ) values are shown in Figure 6 below for the years 1990, 2000, 2010, and 2020.

CONCLUSION:

Land use land cover change (LULC) analysis has been performed for Lahore using the
supervised image classification method with the maximum likelihood classifier algorithm.
Landsat 5 and 8 data have been acquired for the years 1990, 2000, 2010 (Landsat 5), and 2020
(Landsat 8) to examine the changes formed by rapid urbanization. The Landsat thermal band was
used to calculate the land surface temperature (LST) to examine the impact of urbanization, and
consequently surface urban heat island (SUHI) has been computed using urban LST and rural
LST. It was observed that built-up land increased from 22.9% in 1990 to 46.4% in 2020. In
addition, vegetation cover had a decrease in its area from 60.5% (1990) to 47.7% (2020) which
has caused a consistent increase in LST from 30.06° C in 1990 to 33.24 C in 2020. The
difference between urban and rural buffer was 1.72° C in 1990 which increased to 1.91° C in
2000, 2.22° C in 2010, and 2.41 C in 2020. Overall, 0.69°C SUHI change has been observed
from 1990 to 2020. NDVI and LST have been evaluated with each other, and it was examined
that LST and NDVI have an inverse relationship. Therefore, green rooftops, balanced urban
green spaces, and replacing low albedo with high-albedo surface materials can have a significant
positive influence on urban thermal environment.
REFERENCES:

Alam, A., Bhat, M. S., and Maheen, M. (2020). Using Landsat Satellite Data for Assessing the Land Use
and Land Cover Change in Kashmir Valley. GeoJournal 85, 1529–1543. doi:10.1007/s10708-019-10037-x
Avdan, U., and Jovanovska, G. (20162016). Algorithm for Automated Mapping of Land Surface
Temperature Using LANDSAT 8 Satellite Data. J. Sensors, 1480307. doi:10.1155/2016/1480307 Bokaie,
M., Zarkesh, M. K., Arasteh, P. D., and Hosseini, A. (2016). Assessment of Urban Heat Island Based on the
Relationship between Land Surface Temperature and Land Use/Land Cover in Tehran. Sustain. Cities
Soc. 23, 94–104. doi:10.1016/j.scs.2016.03.009 Buhaug, H., and Urdal, H. (2013). An Urbanization
Bomb? Population Growth and Social Disorder in Cities. Glob. Environ. Change 23, 1–10. doi:10.1016/j.
gloenvcha.2012.10.016 Chakraborty, T., and Lee, X. (2019). A Simplified Urban-Extent Algorithm to
Characterize Surface Urban Heat Islands on a Global Scale and Examine Vegetation Control on Their
Spatiotemporal Variability. Int. J. Appl. Earth Observation Geoinformation 74, 269–280.
doi:10.1016/j.jag.2018.09.015 Cheng, L. Y., Meng, X. Y., and Da, X. M. (2018). Simulation of Efficient
Classification for Image Characteristics of Fruit in Fruit Pest. Comput. Simul. 35 (02), 425–428.
Choudhury, D., Das, K., and Das, A. (2019). Assessment of Land Use Land Cover Changes and its Impact
on Variations of Land Surface Temperature in AsansolDurgapur Development Region. Egypt. J. Remote
Sens. Space Sci. 22, 203–218. doi:10.1016/j.ejrs.2018.05.004 Cleve, C., Kelly, M., Kearns, F. R., and
Moritz, M. (2008). Classification of the Wildland-Urban Interface: A Comparison of Pixel- and Object-
Based Classifications Using High-Resolution Aerial Photography. Comput. Environ. Urban Syst. 32, 317–
326. doi:10.1016/j.compenvurbsys.2007.10.001 Congalton, R. G. (1991). A Review of Assessing the
Accuracy of Classifications of Remotely Sensed Data. Remote Sens. Environ. 37, 35–46.
doi:10.1016/0034- 4257(91)90048-b Cui, F., Hamdi, R., Yuan, X., He, H., Yang, T., Kuang, W., et al. (2021).
Quantifying the Response of Surface Urban Heat Island to Urban Greening in Global North Megacities.
Sci. Total Environ. 801, 149553. doi:10.1016/j.scitotenv.2021.149553 Dilawar, A., Chen, B., Trisurat, Y.,
Tuankrua, V., Arshad, A., Hussain, Y., et al. (2021). Spatiotemporal Shifts in Thermal Climate in
Responses to Urban Cover Changes: A-Case Analysis of Major Cities in Punjab, Pakistan. Geomatics, Nat.
Hazards Risk 12, 763–793. doi:10.1080/19475705.2021.1890235 Estoque, R. C., Murayama, Y., and
Myint, S. W. (2017). Effects of Landscape Composition and Pattern on Land Surface Temperature: An
Urban Heat Island Study in the Megacities of Southeast Asia. Sci. Total Environ. 577, 349–359.
doi:10.1016/j.scitotenv.2016.10.195 Fahad, S., Li, W., Lashari, A. H., Islam, A., Khattak, L. H., and Rasool,
U. (2021). Evaluation of Land Use and Land Cover Spatio-Temporal Change during Rapid Urban Sprawl
from Lahore, Pakistan. Urban Clim. 39, 100931. doi:10.1016/j. uclim.2021.100931 Foody, G. M. (2002).
Status of Land Cover Classification Accuracy Assessment. Remote Sens. Environ. 80, 185–201.
doi:10.1016/s0034-4257(01)00295-4 Fu, P., and Weng, Q. (2018). Variability in Annual Temperature
Cycle in the Urban Areas of the United States as Revealed by MODIS Imagery. ISPRS J. Photogrammetry
Remote Sens. 146, 65–73. doi:10.1016/j.isprsjprs.2018.09.003 Goksel, Ç., Musaoglu, N., Gurel, M.,
Ulugtekin, N., Tanik, A., and Seker, D. (2006). Determination of Land-Use Change in an Urbanized District
of Istanbul via Remote Sensing Analysis. Fresenius Environ. Bull. 15, 798–805. Gull, R., and Gul, R. (2017).
In the Surge of Healing-The Case of Indus Basin. Malmo, Sweden: Malmö högskola/Kultur och samhälle.
Hassan, T., Zhang, J., Prodhan, F. A., Pangali Sharma, T. P., and Bashir, B. (2021). Surface Urban Heat
Islands Dynamics in Response to LULC and Vegetation across South Asia (2000-2019). Remote Sens. 13,
3177. doi:10.3390/rs13163177 Hosseini, A., Pourahmad, A., and Pajoohan, M. (2016). Assessment of
Institutions in Sustainable Urban-Management Effects on Sustainable Development of Tehran: Learning
from a Developing Country. J. Urban Plann. Dev. 142, 05015009. doi:10.1061/(asce)up.1943-
5444.0000301 Howard, L. (1833). in The Climate of London: Deduced from Meteorological Observations
Made in the Metropolis and at Various Places Around it. Editors H. J. Darton, A. Arch, Longman, S.
Hatchard, R. H. Highley, et al. Ihlen, V. (2019). Landsat 8 Data Users Handbook. Sioux Falls, SD, USA: US
Geological Survey. Imran, M., and Mehmood, A. (2020). Analysis and Mapping of Present and Future
Drivers of Local Urban Climate Using Remote Sensing: a Case of Lahore, Pakistan. Arabian J. Geosciences
13, 1–14. doi:10.1007/s12517-020- 5214-2 Javid, K., Nawaz Akram, M. A., Pervaiz, S., Siddiqui, R., and
Mazhar, N. (2021). Index-based Approach in Relation to Built-Up and LST Dynamics; A Study of Lahore,
Pakistan. Int. J. Econ. Environ. Geol. 12, 32–40. doi:10. 46660/ijeeg.vol12.iss1.2021.559 Kotharkar, R.,
Ramesh, A., and Bagade, A. (2018). Urban Heat Island Studies in South Asia: A Critical Review. Urban
Clim. 24, 1011–1026. doi:10.1016/j. uclim.2017.12.006 Malik, S., Pal, S. C., Sattar, A., Singh, S. K., Das, B.,
Chakrabortty, R., et al. (2020). Trend of Extreme Rainfall Events Using Suitable Global Circulation Model
to Combat the Water Logging Condition in Kolkata Metropolitan Area. Urban Clim. 32, 100599.
doi:10.1016/j.uclim.2020.100599 Manoli, G., Fatichi, S., Schläpfer, M., Yu, K., Crowther, T. W., Meili, N.,
et al. (2019). Magnitude of Urban Heat Islands Largely Explained by Climate and Population. Nature 573,
55–60. doi:10.1038/s41586-019-1512-9 Mathew, A., Khandelwal, S., and Kaul, N. (2018). Analysis of
Diurnal Surface Temperature Variations for the Assessment of Surface Urban Heat Island Effect over
Indian Cities. Energy Build. 159, 271–295. doi:10.1016/j. enbuild.2017.10.062 Matthews, T., Lo, A. Y.,
and Byrne, J. A. (2015). Reconceptualizing Green Infrastructure for Climate Change Adaptation: Barriers
to Adoption and Drivers for Uptake by Spatial Planners. Landsc. urban Plan. 138, 155–163.
doi:10.1016/j.landurbplan.2015.02.010 Mitchell, L., and Moss, N. (2012).