ASSESSMENT OF URBAN HEAT ISLAND EFFECT THROUGH DIFFERENT

SATTELITE DATA

1. INTRODUCTION

Temperatures in urban areas have long been studied as a function of urban morphology,
land use and anthropogenic activities than just core meteorological parameters. The
phenomenon of urban heat island is now an existent issue of concern in many cities of both
developed and developing nations of the world. Cities like New York (Cox 2011), London
(GLA 2006) and Tokyo (Ooka 2007) have long been affected by heat island phenomenon.
However, with growing development, heat island zones are also being observed in major as
well as upcoming cities of emerging economies like Nigeria (Akinbode et al. 2008), Ethiopia
(Kifle 2003), Oman (Charabi and Bakhit 2011), Mexico (Garcia-Cueto et al. 2007),
Argentina (Camilloni and Barrucand 2011), Malaysia (Takeuchi 2010) and others. In India,
urban heat island effect in major cities is being studied rigorously especially in the last
decade. Badarinath et al.(2005) observed heat island formation in urban regions of
Hyderabad and environs in the night- time satellite data with core urban regions showing
high temperatures. Ramachandra and Kumar (2010) studied local climate patterns in
Greater Bangalore and observed an increase of about 2 to 2.5 °C during the last decade.
Khandelwal et al. (2010) observed a maximum urban -suburban difference of 7 K in Jaipur
city during year 2007. Sundersingh (2003) studied heat island phenomenon in Chennai.
Most of the studies on urban heat island effect in Delhi were conducted in the 1980s (Bahl
and Padmanabhamurty 1979; Nand and Maske 1981; Padmanabhamurty and Bahl 1984).
Micrometeorological measurements for the assessment of urban heat island intensities
(UHI) were conducted in May 2008 (Mohan et al. 2012) over megacity Delhi. The study
revealed temper- ature hotspots in densely populated and commercial areas. Mohan et al.
(2011a, b) have shown a consistent increasing trend in the annual mean minimum
temperatures over four permanent weather stations (Safdarjung, Palam, Gurgaon and
Rohtak) across the National Capital Region of Delhi especially after the 1990s indicating an
overall warming trend. The present study aims to utilize both micrometeorological
measurements and satellite-derived land surface temperatures for urban heat island
assessment for varying land use land cover (LULC) over megacity Delhi. Satellite
observations can provide better spatial coverage, while in situ observations at fixed stations
provide better temporal resolution of data. Thus, a comparison of datasets of both satellite
observations and in situ observations facilitates a better comprehensive understanding of
the urban heat island phenomenon using both satellite thermal data as well as fixed station
temperature data can complement the strengths and weaknesses of each other. Satellite-
derived surface temperatures have been found to have good correlation and lower bias with
in situ temperatures and thus have been used to characterize urban heat island in places
such as Hong Kong (Fung et al. 2009) and several other Asian megacities in both
temperate and tropical climate region (Hung et al. 2006).
2. Data collection and experiment description

2.1. Micrometeorological measurements

Micrometeorological stations were set up at 28 locations at a height of about 1.5 m. At
these stations, temperature and relative humidity were measured and recorded by the
wireless data logger units (T&D Corporation (Japan) Wireless Thermo Recorder Models:
RTR-52 and RTR-72). The instruments comprised of thermistor sensor for air temperature,
polymer humidity sensor for relative humidity and were protected by naturally ventilated
radiation shields. The precision of the sensor was ±0.3 °C within temperature range of
−20 to 80 °C. A typical setup of the instrument at one of the riverside stations. The
instruments were calibrated prior to the experimental phase by standardizing their readings
with that of a standard psychrometer (Assmann Psychrometer). The calibration process
was carried out for a continuous stretch of 24 h. Temperature correction coefficients were
obtained for all data loggers by regression analysis of the measurements taken by the
standard psychrometer with all the temperature data loggers used during the campaign.
These correction coefficients were subsequently used for the measurements by these data
loggers. The average correlation coefficient and RMSE of the temperatures from the
Assmann Psychrometer and the temperature sensors are 0.997 and 0.502 °C, respectively.
In addition, three weather stations measured wind speed and direction, dry bulb
temperature, atmospheric pressure and global solar radiation at the rooftop level (15 m
approximately). Measurements were taken at interval of 1 min from 6 March 2010
(morning)–10 March 2010 (night extending into 11 March), and hourly averages were
obtained for further analysis. The instruments were installed at 1.5 m from the ground level
facing the South direction such that ‘immediate’ surroundings of the instruments were open
and not obstructed by any tree or building. UHI was computed with respect to the station
measuring lowest temperature within the urban area in the absence of a reference station
in the rural area close to the study area. Moreover, an additional station (Torni village,
Haryana) 61 km away from the downtown Delhi has been fixed in an open farmland with
cultivated vegetation in rural surroundings and utilized as the reference point to measure
urban heat island intensity as urban–rural temperature difference. Thus, UHIs comput ed
with respect to the station measuring lowest temperature within the urban area study are
compared with the reference station in the rural area at Torni village. Surface skin
temperatures at selected sites were measured for comparison with the satellite-derived
information on land surface tempera- tures and to provide spatial variability of these under
differ- ent LULC. Sites for measurement of skin temperature were chosen so as to have at
least one site representative for each LULC considered in this study. Skin temperature was
measured using HOBO instruments (HOBO-U12-014 data logger with T type
thermocouple) with precision of ±1.5 °C. A setup of the measurement of skin temperature.

2.2. Satellite-derived land surface temperatures

The land surface temperature data have been retrieved from the Monsoon Asia Integrated
Regional Study program which utilizes Terra and Aqua Moderate-Resolution Imag- ing
Spectroradiometer (MODIS) with 1-km spatial resolu- tion. This dataset consists of 8-day
period averages of land surface temperature and is available from the online data- base of
Goddard Earth Sciences Data and Information Serv- ices Center Interactive Online
Visualization and Analysis Infrastructure (GIOVANNI 2011) of NASA. The MODIS Terra
spacecraft passes over the region of Delhi during 1000 and 1130 hours approximately
 during daytime and similarly from 2200 to 2330 hours during nighttime (by local time).
Thus, the average data of this time period in the present field campaign have been used for
comparison with the satellite- derived temperatures. The present experiment was
conducted from 6 to 10 March 2010, and satellite data are available as 8-day period
average of 6–13 March 2010 which coincides with the experimental days.

Figure 1 Layout of field measurement observation points and their LULC category in the study area. Station
number and name are depicted with the figure, and WS with station name indicates observation point with
weather station.

3.Study area network and site classification

The national capital territory of Delhi covering an area of 32×32 km was chosen as the
study area as shown in Fig. 1. The entire area is divided into 16 major grid cells of 8×8 km
where at least one ground level station was allocated. Some of the grids with different land
use type had more than one station so as to get a representation of the terrains therein.
The locations and LULC categories of these stations are depicted in Fig. 1. Stewart (2007)
and Stewart and Oke (2009) noted that there is a lack of standard and homoge- nous
system of classification for different urban settings because of the remarkable diversity of
landscapes. Essen- tially, any LULC classification is based on built-up area coverage,
building density, green cover and open area cov- erage. In the present study, classification
is based on these LULC features and correlated with LULC classification of the satellite
imagery for each grid in the study area of Delhi (Mohan et al. 2011a). Thus, LULC
categories have been be classified into urban built-up areas, green areas, open areas and
riverside categories.
 I. Urban built-up areas: These include areas with significant infrastructure such as
buildings for residential and commercial purposes. These areas have been sub-
classified into three categories based on density of the built-up structures in the
study domain that corroborates well with the LULC classification for Delhi based on
satellite imagery (Mohan et al. 2009).

a) Dense canopy: These are highly built up areas (built up area>80 %) in the city with
very few open spaces and narrow lanes. One to two-storey residential or commercial
structures, which are generally old, may be intermixed with newer higher storey (up
to five storeys) buildings. These structures are attached at base or close set.
Buildings, houses and other built-up structures in these areas are usually old, and
thus, when the new structures come up, they are not built in any planned manner
thereby adding to the built-up clusters in these areas.
b) Medium dense urban canopy–1: These are well- planned colonies with three- to
four-storey buildings and some higher multistorey buildings, commercial areas,
shopping malls, etc. Built-up fraction accounts for 65–80 % of the total area.
c) Medium dense urban canopy–2: These are well- planned colonies with open spaces
and parks. Built-up area which is about 50–65 % is mostly covered by two- to three-
storey build- ings, shopping centres, small markets, commercial areas, etc.
d) Less dense urban canopy: These are the canopies with residential areas
intersparsed with plenty of green areas and open spaces. Built-up area is between
30 and 50 %.
II. Green areas: These areas include green coverage ranging from forests to cultivated
lands, parks and gardens. These areas have also been further divided into two sub-
categories based on density of vegetation in Delhi.
a) Medium dense forests: Typical dense forests are non-existent in Delhi. However, the
Ridge area stretch is marked with dense lush vegetation, and two such areas
(Sanjay van and Budha-Jayanti Park) have been chosen to represent this type of
land cover in the present study.

b) Parks and gardens: These are cultivated green areas characterized majorly by short
grass and shrub vegetation represented by District Park and Hauz-Khas.
III. Open areas: These areas are open lands with neither any significant built-up
structures nor any type of extensive natural or cultivated green coverage.

IV. Riverside areas: The River Yamuna, which flows through the city, is expected to
influence temperature in neighbouring areas. Thus, two riverside areas (Sail- ing
Club and Majnu-ka-Tila) were chosen for measurements of meteorological
parameters to analyse impact of proximity to river.
Table 1 details out all the stations classified under the above given groups.
4. Weather conditions during the field campaign

The field campaign was carried out in the month of March 2010 in Delhi. The month of
March in Delhi is a transition period between spring–summer characterised with warm days
and cool nights. As per the Climate Normals of Delhi (IMD 2011), the average maximum
temperature during the month of March is about 30 °C, while the average minimum
temperature is about 15.4 °C. During the present study, the minimum temperatures ranged
from about 13 to 18 °C, while the maximum temperatures ranged from 25 to 31 °C, and
daily maximum solar radiation ranged from 782 to 844 Wm−2. Major wind directions were
observed to be north-west, west and south-west. The wind speeds during the experimental
days ranged from calm to 3.0 m/s. There was an incidence of light drizzle on the night of 6
March 2010 between 11:00 and 11:30 pm. Other than this there was no incidence of
precipitation during the duration of the experiments. Cloud cover from 6 March 9:30 pm to
7 March 3:30 am varied from 4 to 6 octas.

5. Results and discussions

5.1 Urban heat island intensities from in situ measurements for varying
LULC
Analysis of in situ measurements revealed that urban heat island intensity (UHI) at various
sites can be grouped as per LULC categories. Accordingly, as per the average UHI
a. Very high UHI: Stations with maximum UHI ranging from 9.1 to 10.7 °C and average
UHI ranging from 4.0 to 4.3 °C
b. High UHI: Stations with maximum UHI ranging from 8.6 to 9.6 °C and average UHI
ranging from 3 to 3.8 °C
c. Medium UHI: Stations with maximum UHI ranging from 7.3 to 7.7 °C and average
UHI ranging from 3.1 to 3.7 °C.
d. Low UHI: Stations with maximum UHI ranging from 5.8 to 6.9 °C and average UHI
ranging from 2.2 to 2.9 °C.
e. Very low UHI: Stations with maximum UHI ranging from 3.1 to 6.4 °C and average
UHI ranging from 0.8 to 1.9 °C.
Figure 4 shows diurnal UHI profile of these groups. Very high UHIs were observed
mainly for dense and commercial urban canopies like Sitaram Bazar, Connaught Place
(CP), Bhikaji Cama and Noida which is a mix of residential and commercial land use.
The observation station at Badarpur lies in vicinity of a major thermal power plant
namely Badarpur Thermal Power Station and vigorous construction activities. This
indicates that anthropogenic heat emissions from nearby areas might have affected the
ambient temperature at Badarpur site. Medium and low UHI group include majorly the
medium dense and less dense urban canopies as identified in Table 1. As expected,
green areas such as a forest and a park showed very low UHI represented by the sites
namely, Sanjay Van and Buddha Jayanti Park. Out of the two riverside stations chosen,
Majnu-ka-Tila truly represented the riverfront, while Sailing Club, located not far from
residential areas, observed higher UHI comparatively. Therefore, there is a close
 corroboration between the sites selected to represent classified LULC categories
(Section 3) of this study with that from the appropriate UHIs barring two exceptions as
explained above. The maximum UHI was observed to be increasing through the
experimental dura- tion, and maximum hourly UHI observed was 10.7 °C at Sitaram
Bazar. Urban surfaces constitute dry impervious surfaces which lead to limited
evaporative cooling in com- parison to moist green surfaces. Urban geometry also leads
to lower sky view factor thereby preventing escape of radiation at nighttime (Oke 1982).

5.2 Reference point for estimation of UHI

As per conventional definition of UHI (Oke et al. 1972), it is the difference in ambient air
surface temperature between the city centre and a nearby rural area. However, over the
years the urban geometry has become more complex, and we have multifaceted land use
and cover in the cities with lots of commercial, industrial, residential and green areas
distributed. Therefore, one can also expect large spatial temperature gradients within the
city that can be relevant in the context of cooling and/or heating energy demand. With
growth of satellite towns, it is difficult to find a rural site free from influence of urban
canopies near the boundaries of urban areas. In the present study, UHI of a station has
been computed as the difference of temperature at that station and the lowest temperature
in the study area at that time. Lowest temperature stations were always located in the
green areas (large parks, forests, etc.) within the urban boundaries as expected as these
areas are least affected by built structures and anthropogenic heat emissions. The UHI has
been assessed from the two methods: (1) using difference between urban station
temperature under consideration and the rural site temperature and (2) using difference
between urban station temperature and the lowest temperature among all urban stations
(Fig. 5). In Fig. 5, ΔTstation-rural refers to UHI computed from method (1), and ΔTstation-min refers
to UHI estimated by method (2). Maximum UHI and average UHI computed from both the
methods are fairly close with maximum difference being only 1.2 °C. This shows that in the
absence of a reference rural site for urban heat island assessment, method (2) based on
the lowest temperature within the city could be a good representation of urban heat island
intensity. UHI intensity computed with reference to a rural site is in agreement with UHI
computed with reference to lowest temperature station. In the absence of a nearby rural
site due to growth of satellite towns, such proxy arrangement of computing UHI is expected
to work well. However, this arrangement would only work if the city has a large enough
undeveloped, vegetated region within its boundaries which could mimic a rural location.
Delhi with its variable LULC offers many zones which can substitute for a rural area.
However, the same may not be true for other cities. Especially in India, there are many old
metropolitan cities, where some areas/districts are very developed while others are
farmland/green areas. In addition, these major cities are then surrounded with satellite
towns, which are constructed in a planned manner. Thus, to find a rural area close to the
city of UHI studies becomes difficult. However, for cities which do not satisfy these
conditions, this may not be applicable.
Diurnal variation of various categories of UHI intensities across the city

5.3 Ambient temperature and skin temperature

The relationship of land surface temperature and ambient air temperature depends on land cover
types and can be derived by considering energy exchange between the atmospheres and the land
surface (Cheng et al. 2008). Prigent et al. (2003) studied the surface and air temperatures over
different land cover types and stated that energy exchanges at the land surface boundary are
largely controlled by the different reacting time and space scales of air and surface. The land
surface temperature responds more rapidly to changes of the local balance of energy than the air
temperature. Prigent et al. (2003) deduced that with decreasing soil and vegetation moisture, the
evaporation rate decreases, increasing the sensible heat flux thereby leading to larger surface air
temperature differences. In the present study as well, the difference between ambient and surface
skin (ΔTskin-ambient) temperature is greatest for dense built-up areas due to low moisture content both
during daytime as well as nighttime with a maximum difference of 14.2 and −5.3 °C, respectively.
For riverside areas, the maximum difference was 6.2 and −3.3 °C, and for open areas, it was 6.3
and −1.5 °C for daytime and night-time, respectively. It can be seen in that the LULC
characteristics are represented in the difference between ambient and skin temperature. Land
surface in urban areas acquires higher temperatures than ambient air during daytime and lower
temperature during nighttime. The urban canopy in built-up areas pro- vides multiple surfaces for
absorption and reflection of heat waves which both generates and traps heat. This eventually
leads to higher ambient temperature during nighttime. In contrast, green areas report higher skin
temperature than ambient temperature. Presence of moisture on green cover surfaces enables
lower temperatures. Open areas have reported least differences between skin temperature and
ambient temperature due to absence of heat-trapping canopy structures. Thus, difference between
ambient and skin temperature is closely related to LULC with maximum gradient for the built-up
areas.
5.4 Comparison of UHI and skin temperatures from satellite-derived land
surface temperatures and in situ measurements.

Urban heat island intensity with respect to lowest temperature and rural area temperature

Figure shows that between measured skin temperatures and MODIS-derived land surface
temperatures, the scatter for nighttime temperatures is less than for the daytime
temperatures, which indicates that nighttime in situ STs co-relate better with satellite-derived
STs. The correlation coefficient between in situ STs and MODIS Terra STs dur- ing nighttime
is 0.47, while during daytime it is 0.30. Roth et al. (1989) observed that due to under
sampling of active urban surface, daytime satellite temperatures pick up roof temperatures
which are the hottest surfaces during daytime and coldest during nighttime and thus diurnal
range of temperature is over-accentuated. In the present study also, satellite LSTs were
found to be greater than observed skin temperatures during daytime and vice versa during
night- time possibly for similar reasons. Considering the spatial limitation of in situ
measurements of depicting single point observations against 1 km 2 area in the MODIS STs,
the comparison between the two is quite reasonable. In UHI studies, temperature differences
are more impor- tant spatially in order to identify the hotspots. The top two UHI hotspots are
the same in case of MODIS and in situ observations as clearly seen in Fig. 8a and b,
respectively, which are identified at sites 30 (Sitaram Bazar) and 14 (Cannaught Place).
However, some different additional hot-spots are also prominent in which 12 (Janakpuri) and
9 (Lajpat Nagar) against 5 (Badarpur) and 7 (Bhikaji Cama) for in situ observations.
Maximum night-time (2200– 2330 hours) UHI intensity for MODIS LSTs is 5.4 °C, while that
based on ambient temperature is 7.6 °C. Thus, UHI obtained during night-time is in
concurrence with the remote sensing data of LST. These findings validate the UHI intensity
and major hotspots from both the in situ and satellite observations during night-time.
However for daytime, the correlation between satellite- derived and in situ observation is not
good for the following reasons. Temperature hotspots are not similar in both contours.
Further, there is large difference between the maximum UHI based on LST (8.1 °C) and
ambient temperatures (2.4 °C). Land surface and air differ in their heat capacities, and this
difference is more prominent during daytime in the presence of solar insolation at the ground.
Due to these differences, a poor correlation in UHI was established between satellite-derived
and in situ obser- vations during daytime. Thus, though satellite data can be used as a proxy
to in situ observation for UHI analysis during nighttime, the same is not recommended for
daytime. As per Roth et al. (1989), satellite-derived data are surface radiant temperatures of
only those surfaces seen by the radiometer as averaged across the area of a pixel. Thus, it
should not be equated with from other systems. Hence, satellite-derived surface data cannot
be expected to coincide with near surface-based measurements. The view is reiterated by
Voogt and Oke (2003) who explain that urban air temperatures are influenced by several
other processes in the urban canopy layer. However, in the present study, satellite data have
been used to qualitatively validate the results of the present study. The present study
involves air temperature at 1.5 m height above ground level at fixed stations over the entire
study area. In such an arrangement, many potential UHI hotspots can be missed. Satellite
data present a continuous spatial variation of data. Thus, if major hotspots as observed in the
near surface measurements are also observed in satellite-derived land surface temperatures,
it strengthens the findings of the study for utilizing the satellite data appropriately. Satellite-
derived temperature data have been used to complement other techniques of UHI
characterization such as fixed station temperature data, mobile surveys, etc. in some of the
recent studies. Fung et al. (2009) explored UHI intensity in Hong Kong using mobile
transverse and remote sensing techniques and observed that the in situ measured surface
temperature was consistent with satellite-derived surface temperature with low bias and good
correlation. Hung et al. (2006) used MODIS land surface temperature data to study spatial
patterns of UHIs for 18 Asian megacities. The maximum night-time UHI based on MODIS
satellite data LST for chronologically similar 8-day period as the experiment duration has
increased progressively from January 2010 (4.6 °C) onto February (5.2 ° C), March (5.4 °C)
and April (7.1 °C). Considering the good correlation be- tween nocturnal UHI of ambient
temperatures and LST in the present study and progressively increasing UHI, it could be
inferred that UHI could be higher with in situ observations in the subsequent summer
months.
5.5

Average UHI distribution for nighttime (2200–2330 hours) and daytime (1000–1130 hours) based on a nighttime MODIS
LST, b nighttime ambient in situ temperature, c daytime MODIS LST and d daytime ambient in situ temperature
Comparison with other cities

Table shows compilation of some reported UHIs in major cities of the world based on various
field observations. The average maximum daily UHI observed in present measurements (8.3
°C) is comparable to other cities like London, Beijing and Tokyo indicating that urban heat
island phenomenon in Delhi is of significant importance as in other megacities of the world.
With increasing urbanization, stronger heat island intensities are expected to be observed
here. Population size and associated anthropogenic activities are also expected to influence
UHI. Oke (1999) and Roth (2002) observed that under clear and calm conditions, UHI in a city
is proportional to its size in terms of population. The highest magnitude of UHI in this study
was observed on a clear and dry day (March 10, 2010). Oke (1973) analysed relationship
between UHImax and populations of various North American and European cities and deduced
that under clear and calm conditions, urban heat island magnitude is proportional to the city
population. Accordingly, the follow- ing relationships were proposed (1) for North American
cities and (2) European cities: For North America ΔTu—rðmaxÞ ¼ 2:96 log P — 6:41 ð1Þ
For Europe ΔTu—rðmaxÞ ¼ 2:01 log P — 4:06 ð2Þ Delhi, with a population of 13.8 million
(Census of India 2011), yields ΔTu−r(max) as 14.7 and 10.2 °C following Eqs. (1) and (2),
respectively. Thus, the observed daily ΔTu−r(max) for Delhi ( 8.3 °C) is significantly lower than the
results based on North American and European cities. Roth (2007) states that in these
relationships, population is used as a convenient and widely available surrogate for urban
geometry which has a more important control on heat island intensity. The relationship also
ignores the additional role of energy consumption per capita. Nonetheless, these relationships
signify the fact that Delhi being a developing megacity with mixed LULC and interspersed
green areas is likely to experience greater UHI effect with increasing anthropogenic heat
fluxes with growth of population. With growing population and consequently increase in built-
up structures, the urban heat island intensity is expected to strengthen in the city calling for
strategic and remedial actions.
Conclusion

After observing the details of the report we got to know some of the UHI main Characteristics
in the megacity of Delhi. Firstly the UHI Intensity Categories i.e. High intensity that was 8.6-
10.7 °C then the Medium intensity which was 7.3-7.7 °C and the Low intensity which was
observed at 3.1-6.9 °C these were the UHI intensities that we observed. Secondly, the LULC
Classification which was UHI intensity that correlated well with LULC classification. The
recorded intensities were, the maximum UHI intensity was 10.7 °C in the area Sitaram Bazar,
a dense residential and commercial area. Moreover the Air-Skin Temperature Gradient like
Reflective of LULC, urban canopy density, and the thermal characteristics. Apart from this the
Nocturnal Temperature Differences at Built-up areas was -5.3 °C, on riverside areas was -3.3
°C and in the open areas it recorded at -1.5 °C. After the Comparison with Global Cities
Maximum Daily UHI, Delhi was recorded at 8.3 °C UHI which was comparable to London,
Tokyo, and Beijing. The City-Specific Features were Empirical relations from American and
European cities may not be directly applicable to Delhi. The implications for the Urban
Planning were, at first understanding the UHI characteristics which also informs the urban
planning strategies. .initially the Environmental Sustainability for the City-specific UHI studies
which contributes to the sustainable urban environmental management. Some of the future
Future Research Directions can be, City-Specific UHI Modeling which Develops the empirical
relations and models tailored to Delhi's urban environment. After incorporating Local Factors
we can Consider some local factors like urban morphology, land use patterns, and climate
CITATION
1. Akinbode OM, Eludoyin AO, Fashae OA (2008) Temperature and relative humidity
distributions in a medium-size administrative town in southwest Nigeria. J Environ Manag
87:95–105
2. Badarinath K, Chand T, Madhavilatha K, Raghavaswamy V (2005) Studies on urban heat
islands using ENVISAT AATSR data. J Indian Soc Remote Sens 33:495–501
3. Bahl HD, Padmanabhamurty B (1979) Heat island studies at Delhi. Mausam 30:119–122
4. Camilloni I, Barrucand M (2011) Temporal variability of the Buenos Aires, Argentina, urban
heat island. Theor Appl Climatol. doi:10.1007/s00704-011-0459-z
5. Census of India (2011) http://censusindia.gov.in/2011-common/census dataonline.html.
Accessed 29 Jul 2011
6. Charabi Y, Bakhit A (2011) Assessment of the canopy urban heat island of a coastal arid
tropical city: the case of Muscat, Oman. Atmos Res 101:215–227
7. Cheng KS, Su YF, Kuo FT, Hung WC, Chiang JL (2008) Assessing the effect of landcover
changes on air temperature using remote sensing images—a pilot study in northern
Taiwan. Landsc Urban Plan 85:85–96
8. Cox JR (2011) Cool Neighborhoods within the New York Metro Heat Island. City Weathers:
Meteorology and Urban Design 1950–2010, Manchester, UK.
http://www.sed.manchester.ac.uk/architecture/
research/csud/workshop/programme/Cox_NewYork.pdf. Accessed 14 Aug 2011
9. McCormick, R. A., and D. M. Baulch, 1962: The variation with height of the dust loading
over a city, as determined from atmospheric turbidity. J. Air Pollut. Control, 12, 492-496,
10. Myrup, L. D., 1969: A numerical model of the urban heat island. J. Appl. Meteor., 8, 908-
918.
11. Fung WY, Lam KS, Nichol J, Wong MS (2009) Derivation of nighttime urban air
temperatures using a satellite thermal image. J Appl Meteorol Climatol 48:863–872
12. Gaffin SR, Rosenzweig C, Khanbilvardi R, Parshall L, Mahani S, Glickman H, Goldberg R,
Blake R, Slosberg RB, Hillel D (2008) Variations in New York city’s urban heat island
strength over time and space. Theor Appl Climatol 94:1–11
13. Garcia-Cueto OR, Jauregui-Ostos E, Toudert DY, Tejeda-Martinez A (2007) Detection of
the urban heat island in Mexicali, B. C., México and its relationship with land use.
Atmosfera 20:111–131
14. Duckworth, F. S., and J. S. Sandberg, 1954: The effect of cities upon horizontal and vertical
temperature gradients. Bull. Amer. Meteor. Soc., 35, 198-207,
15. Huff, F. A., and J. L. Vogel, 1978: Urban, topographic and diurnal effects on rainfall in the
St. Louis region. J. Appl. Meteor., 17, 565-577.
16. Rao, P. K. 1972: Remote sensing of urban heat islands from an environmental satellite.
Bull. Amer. Meteor. Soc., 53, 647-648.
17. Torrance, K. E., and J. S. W. Shum, 1976: Time varying energy consumption as a factor in
urban climate. Atmos. Environ., 10, 329-337.
18. Matson, M., E. P. McClain, D. F. McGinnis, Jr. and J. A. Pritchard, 1978: Satellite detection
of urban heat islands. Mon. Wea. Rev., 106, 1725-1734.
19. Heat Capacity Mapping Mission User's Guide, 1978; Goddard Space Flight Center,
Greenbelt, MD, 120 pp.
20. GIOVANNI (2011) Monsoon Asia Integrated Regional Study (MAIRS) 8-days Products.
http://gdata1.sci.gsfc.nasa.gov/daac-bin/G3/gui.cgi?instance_id=mairs_8day. Accessed 2
Aug 2011
21. Bohse, J. R., M. Bewtra and W. L. Barnes, 1979: Heat capacity mapping radiometer data
processing algorithm, calibration and flight performance evaluation. Goddard Space Flight
Center, Greenbelt, M.D., 162 pp.
22. Bornstein, R. D., 1968, Observations of the urban heat island effect in New York City. J.
Appl. Meteor., 7, 575-582.
23. Carlson, T. N., J. N. Augustine and F. E. Boland, 1977: Po- tential application of satellite
temperature measurements in the analysis of land use over urban areas. Bull. Amer.
Meteor. Soc., 58, 1301-1303.
24. Huff, F. A., and J. L. Vogel, 1978: Urban, topographic and diurnal effects on rainfall in the
St. Louis region. J. Appl. Meteor., 17, 565-577.
25. Matson, M., E. P. McClain, D. F. McGinnis, Jr. and J. A. Pritchard, 1978: Satellite detection
of urban heat islands. Mon. Wea. Rev., 106, 1725-1734.
26. Nunez, M., and T. R. Oke, 1977: The energy balance of an urban canyon. J. Appl. Meteor.,
16, 11-19.
27. Greater London Authority (GLA) (2006) London’s urban heat island: a summary for decision
makers. http://static.london.gov.uk/mayor/
environment/climate-change/docs/UHI_summary_report.pdf. Accessed 3 Mar 2011
28. Hung T, Uchihama D, Ochi S, Yasuoka Y (2006) Assessment with satellite data of the
urban heat island effects in Asian mega cities. Int J Appl Earth Obs Geoinf 8:34–48
29. India Meteorological Department (IMD) (2011) Climatological table.
http://www.imd.gov.in/section/climate/newdelhiweb.htm. Accessed 20 April 2011
30. Khandelwal S, Goyal R, Kaul N (2010) Study of seasonal and spatial pattern of urban heat
island of Jaipur City and its relationship with enhanced vegetation index. 13th Annual
International Conference and Exhibition on Geospatial Information Technology and
Applications.http://mapindia.org/2010/proceeding/pdf/186.pdf. Accessed 27 Jul 2011
31. Kifle B (2003) Urban heat island and its feature in Addis Ababa (a case study). Fifth
International Conference on Urban Climate, Lodz, Poland.
http://nargeo.geo.uni.lodz.pl/∼icuc5/text/P_6_11.pdf. Accessed 14 Aug 2011
32. Mohan M, Kikegawa Y, Gurjar BR, Bhati S, Kandya A, Ogawa K (2009) Assessment of
urban heat islands intensities over Delhi. The Seventh International Conference on Urban
Climate (ICUC7). http://www.ide.titech.ac.jp/∼icuc7/extended_abstracts/pdf/ 375621-2-
090515211144-004.pdf. Accessed 3 Mar 2011
33. Mohan M, Pathan SK, Narendrareddy K, Kandya A, Pandey S (2011a) Dynamics of
urbanization and its impact on land-use/land-cover: a case study. J Environ Prot 2:1274–
1283
34. Mohan M, Kandya A, Battiprolu A (2011b) Urban heat island effect over national capital
region of India: a study using the temperature trends. J Environ Prot 2:465–472
35. Mohan M, Kikegawa Y, Gurjar BR, Bhati S, Kandya A, Ogawa K (2012) Urban heat island
assessment for a tropical urban airshed in India. Atmos Clim Sci 2:127–138
36. Nand K, Maske SJ (1981) Mean heat island intensities at Delhi— assessed from urban
climatological data. Mausam 32:269–272
37. Oke TR (1973) City size and the urban heat island. Atmos Environ 7:769–779
38. Oke TR (1982) Energetic basis of urban heat island. Q J R Met Soc 108:1–24
39. Oke T (1999) Proceedings of International Conference on Urban Climate, Sydney
40. Oke TR, Yap D, Maxwell GB (1972) Comparison of urban rural cooling rates at night. Proc
Conf on urban environment and second conference on biometeorology, Phil- 31 Oct-2 Nov
1972
41. Ooka R (2007) Recent development of assessment tools for urban climate and heat-island
investigation especially based on experiences in Japan. Int J Climatol 27:1919–1930
42. Padmanabhamurty B, Bahl HD (1984) Isothermal and isohyetal patterns of Delhi as a
sequel of urbanization. Mausam 33:4
43. Prigent C, Aires F, Rossow WB (2003) Land surface skin temperatures from a combined
analysis of microwave and infrared satellite observations for an all-weather evaluation of
the differences between air and skin temperatures. J Geophys Res 108:4310–4313
44. Ramachandra TV, Kumar U (2010) Greater Bangalore: emerging urban heat island. GIS
Development. http://www.ces.iisc.ernet.in/energy/ paper/Bangalore_heatisland/index.htm.
Accessed 27 Jul 2011
45. Roth M (2002) Effects of Cities on Local Climates. Workshop of GES/ APN Mega-City
Project. http://enviroscope.iges.or.jp/contents/
13/data/PDF/24%20Jan%20PDF/ROTH_IEGS.pdf. Accessed 1 Aug 2011
46. Roth M (2007) Review of urban climate research in (sub)tropical regions. Int J Climatol
27:1859–1873
47. Roth M, Oke TR, Emery WJ (1989) Satellite-derived urban heat island from three coastal
cities and the utilization of such data in urban climatology. Int J Remote Sens 10:1699–
1720
48. Saitoh TS, Shimada T, Hoshi H (1996) Modeling and simulation of the Tokyo urban heat
island. Atmos Environ 30:3431– 3442
49. Sarkar A, Ridder K (2011) The urban heat island intensity of Paris: a case study based on a
simple urban surface parametrization. Boundary-Layer Meteorol 138:511–520
50. Stewart ID (2007) Landscape representation and the urban–rural dichotomy in empirical
urban heat island literature, 1950–2006. Acta Climatolog et Chorol 40–41:111–121
51. Stewart ID, Oke T (2009) Classifying urban climate field sites by “local climate zones the
case of Nagano, Japan. 7th International Conference on Urban Climate, Yokohama, Japan,
June 29–July 3, 2009. http://www.ide.titech.ac.jp/∼icuc7/extended_abstracts/pdf/ 385055-1-
090515165722-002.pdf Accessed 21 Jan 2011
52. Sundersingh SD (2003) Effect of heat islands over urban madras and measures for its
mitigation. Energy Build. doi:10.1016/0378-7788 (90)90136-7
53. Takeuchi W (2010) Application of remote sensing and GIS for monitoring urban heat island
in Kuala Lumpur Metropolitan area. Map Asia 2010 and the International Symposium and
Exhibition on Geoinformation, Kuala Lumpur, Malaysia.
http://www.mapasia.org/2010/proceeding/pdf/306.pdf. Accessed 14 Aug 2011
54. Voogt JA, Oke TR (2003) Thermal remote sensing of urban climates. Remote Sens Environ
86:370–384
55. Wang Y, Hu F (2006) Variations of the urban heat island in summer of the recent 10 years
over Beijing and its environment effects. Chin J Geophys 49:59–67