Geocarto International

ISSN: 1010-6049 (Print) 1752-0762 (Online) Journal homepage: http://www.tandfonline.com/loi/tgei20

Land surface temperature and emissivity

estimation for Urban Heat Island assessment
using medium- and low-resolution space-borne
sensors: A review

Abdelmoneim Abdelsalam Mohamed, John Odindi & Onisimo Mutanga

To cite this article: Abdelmoneim Abdelsalam Mohamed, John Odindi & Onisimo Mutanga (2017)
Land surface temperature and emissivity estimation for Urban Heat Island assessment using
medium- and low-resolution space-borne sensors: A review, Geocarto International, 32:4, 455-470,
DOI: 10.1080/10106049.2016.1155657

To link to this article: http://dx.doi.org/10.1080/10106049.2016.1155657

Accepted author version posted online: 19

Feb 2016.
Published online: 16 Mar 2016.

Submit your article to this journal

Article views: 135

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=tgei20

Download by: [UNIVERSITY OF KWAZULU-NATAL] Date: 07 February 2017, At: 05:03

Geocarto International, 2017
VOL. 32, NO. 2, 455–470
http://dx.doi.org/10.1080/10106049.2016.1155657

Land surface temperature and emissivity estimation for Urban

Heat Island assessment using medium- and low-resolution
space-borne sensors: A review
Abdelmoneim Abdelsalam Mohamed, John Odindi and Onisimo Mutanga
Discipline of Geography, School of Agricultural, Earth and Environmental Sciences, University of KwaZulu-Natal,
Pietermaritzburg, South Africa

ABSTRACT ARTICLE HISTORY

The Urban Heat Island (UHI) phenomenon, a typical characteristic on urban Received 29 July 2015
landscapes, has been recognised as a key driver to the transformation of local Accepted 1 February 2016
climate. Reliable retrieval of urban and intra-urban thermal characteristics KEYWORDS
using satellite thermal data depends on accurate removal of the effects Remote sensing; urban;
of atmospheric attenuations, angular and land surface emissivity. Several thermal; medium-resolution;
techniques have been proposed to retrieve land surface temperature (LST) land surface temperature
from coarse resolution sensors. Medium spatial resolution sensors like (LST)
the Advanced Space-borne Thermal Emission and Reflection Radiometer
and the Landsat series offer a viable option for assessing LST within urban
landscapes. This paper reviews the theoretical background of LST estimates
from the thermal infrared part of the electromagnetic spectrum, LST retrieval
algorithms applicable to each of the commonly used medium-resolution
sensors and required variables for each algorithm. The paper also highlights
LST validation techniques and concludes by stipulating the requirements
for LST temporal and spatial configuration.

1. Introduction
Over 50% of the current global population live in urban areas, this percentage is expected to reach
69.6% by 2050 (United Nations 2010). Increasing urbanisation is commonly associated with transfor-
mation from natural to built-up landscapes. Such transformation results in changes in local climate
which threaten social and ecological sustainability (Basara et al. 2010). The Urban Heat Island (UHI),
a phenomena characterised elevated urban temperature in comparison to the rural periphery, has been
identified as a major driver of local, regional and global climate change (Weng 2001, 2003; Arnfield
2003; Hage 2003; Streutker 2003; Voogt & Oke 2003; Xian & Crane 2006). The elevated urban thermal
properties may among others lead to deterioration of the local environment by facilitating reaction of
atmospheric chemical compounds, leading to an increase in ozone gases, increase energy demand in
urban areas by necessitating air conditioning, induce heat waves, alter local wind and rainfall patterns,
and increase smog and atmospheric aerosol loads (Curriero et al. 2002; Konopacki & Akbari 2002;
Crutzen 2004; Huang et al. 2008; Johnson et al. 2009; US Environmental Protection Agency 2012).
Traditionally, UHI is determined by establishing differences in air temperature between urban
and rural areas using pairs of ground-based meteorological observation measurements (Henry et

CONTACT  Abdelmoneim Abdelsalam Mohamed  Ahmeda1@ukzn.ac.za

© 2016 Informa UK Limited, trading as Taylor & Francis Group
456    A. A. Mohamed et al.

al. 1989; Gallo & Owen 1999; Morris et al. 2001; Gaffin et al. 2008; Chow & Svoma 2011; Shao et al.
2011). Whereas these methods are commonly characterised by high temporal resolution, effective for
illustrating UHI temporal variation, they have limited spatial coverage due to poor spatial resolution
(Xiaoma et al. 2013). The emergence of thermal sensors on satellite platforms has opened opportuni-
ties to determine UHIs at meso- and low-resolution spatial scales. Typically, LST is determined from
thermal infrared (TIR) data retrieved by the radiative transfer equation (RTE) (Li et al. 2013a). In
UHI studies, Voogt and Oke (2003) note that LST, which replaces specific air temperature, is used as
a surrogate for surface thermal characteristics.
To date, studies on UHI have been dominated by low spatial resolution (8 km) Advanced Very High
Resolution Radiometer (AVHRR) and the 1 km Moderate Resolution Imaging Spectroradiometer
(MODIS) – (Gallo & Owen 1998; Friedl et al. 2002; Streutker 2003; Cui & De Foy 2012). However, stud-
ies adopting satellite thermal infrared (TIR) data, with varying low spatial resolutions for intra-urban
thermal assessment have suggested that such data do not accurately account for surface temperature
and heat fluxes on heterogeneous and complex landscapes (Pu et al. 2006; Weng 2009; Chakraborty
et al. 2015). Validation of MODIS for instance has established a 6.5˚ C error on complex landscapes
(Chakraborty et al. 2015). Therefore, adoption of medium-resolution thermal sensors like Landsat’s
Thematic Mapper (TM), Enhanced Thematic Mapper Plus (ETM+) and Advanced Space-borne
Thermal Emission and Reflection Radiometer (ASTER) with 120 and 60 and 90 m spatial resolutions,
respectively, may generate more reliable urban LST values, hence better assessment of surface tem-
perature and heat fluxes on the heterogeneous urban landscapes (Weng 2001; Chen et al. 2002; Hartz
et al. 2006; Lu & Weng 2006; Yuan & Bauer 2007; Stathopoulou et al. 2009; Xu et al. 2009; Zhang et
al. 2009, 2013; Gao et al. 2011; Tomlinson et al. 2011; Li et al. 2012; Tan et al. 2012; Zhao et al. 2012;
Mallick et al. 2013). According to Sobrino et al. (2012), sensors with approximately 50 pixel size may
be most ideal for accurate intra-urban thermal mapping. Currently, thermal sensors with relatively
higher spatial resolution include ASTER (90 m) and the Landsat series (120 m for 4 and 5, 60 m for
7 and 100 m for 8 (Schott et al. 2012).
Commonly, the split-window (SW) land surface temperature algorithms, derived from a first-order
Taylor-series linearization of the radiative transfer equation (RTE) in long-wave infrared spectral bands
are used to retrieve LST from thermal satellite data (McMillin & Crosby 1984). Using estimates of the
surface spectral emissivity, the SW LST algorithms simultaneously converts brightness temperatures to
land surface temperatures. This technique is particularly popular due to its simplicity and robustness
(Yu et al. 2008). The SW technique is used in the 11 and 12 μm channels of the AVHRR and MODIS
products and could be adopted for TIR region two-channel medium-resolution sensors like ASTER
and Landsat-8. This technique uses adjacent channels with different properties to calculate atmospheric
attenuation and therefore cannot be adopted for single-channel sensors like Landsat TM/ETM+.
It is also critical to obtain information on urban surfaces emissivity which can be used to reduce
error when retrieving LST from thermal satellite data. Emissivity (ɛ) is defined as the ‘emitting ability’
of a natural material, compared to that of an ideal blackbody at the same temperature. The values of
emissivity range between 0 and 1. Generally, the retrieval of land surfaces emissivity (LSE) from space
and direct estimation of LSE from passive satellites are impractical due to combined effects of the LST,
LSE and atmospheric contamination (Li et al. 2000; Jiang et al. 2006). Over the past decades, LST
estimation from satellite TIR measurements has significantly improved. Many algorithms have been
proposed to deal with the characteristics of various sensors onboard different satellites. These algo-
rithms utilise different assumptions and approximations for the radiative transfer equation (RTE) and
land surface emissivity (LSE). If LSEs are known a priori, these algorithms can generally be grouped as
single-channel methods, multi-channel methods and multi-angle methods. If the LSEs are unknown,
the algorithms can be categorised into stepwise retrieval method, simultaneous retrieval of LSEs and
LST with known atmospheric information, and simultaneous retrieval with unknown atmospheric
information (see; Li et al. 2013a).
Due to the potential of medium-resolution sensors to map the thermal characteristics on heter-
ogeneous urban landscapes, this review details LST and LSE retrieval methods and their potential
 Geocarto International   457

for adoption on medium spatial resolution. The review further comments on spatial and temporal
prospects of effective intra-urban surface thermal mapping.

2.  Derivation of land surface temperature

A fundamental requirement for thermal remote sensing is the detection of electromagnetic radiation
(EMR) by sensors facilitated by unique emission of objects in the relevant sections of the EMR. Within
the EMR, the most valuable wavelength for LST retrieval is the 8 to 15 μm range of the thermal infrared
(TIR) section (Tomlinson et al. 2011).
Satellite TIR sensors receive EMR, which can be quantified as Top of Atmosphere (TOA) radiance.
This include upwelling radiance emitted from the ground, up-welling radiance from the atmosphere
and the down-welling radiance emitted by the atmosphere and reflected from the ground. During day,
there is both emission and reflection of EMR, at night, sensed EMR is restricted to emission. The inverse
of Planck’s law (the energy emitted by a surface is directly related to its temperature) is used to derive
blackbody/brightness temperatures from TOA radiances. The TOA radiances is converted to LST by
correcting for angular effects, atmospheric attenuation and spectral emissivity values at the surface.
As high zenith angles result in lengthened atmospheric path that can result in less accurate ther-
mal values, angular effects, a product of the variety in viewing angles resulting in wavelength shifting
must be compensated when estimating radiance (Dash et al. 2002; Streutker 2003). Thermal images
are commonly accompanied by additional metadata that can be used to correct for this anomaly.
Atmospheric attenuation (absorption, reflection or refraction and scattering) alter the EMR as the
target to sensor signal is modified as it passes through the atmosphere, resulting in differences between
TOA radiances and LST. Within TIR wavelengths, attenuation is mainly caused by the presence of
water vapour. When atmospheric conditions are known, emission and absorption of radiation in the
atmosphere can be quantified and corrected using one of the radiative transfer computer codes, e.g.
MODerate resolution atmospheric TRANsmission (MODTRAN) (Berk et al. 2011). Atmospheric
conditions are typically assessed using in situ atmospheric profile data. Whereas on-line atmospheric
databases or estimations based on empirical models (Precipitable Water at the VLA 1990; –1998) are
available online, such data are often not available during image acquisition (Barsi et al. 2003, 2005).
Two main algorithmic approaches are used for atmospheric correction for retrieval of LST; the
radiative transfer equation (RTE) and the split window (SW) technique. These techniques are detailed
in Dash et al. (2001) and Weng (2009). In addition to relevant radiative transfer model and knowledge
on atmospheric profile, data on emissivity is also required for LST estimation. For satellites with a single
thermal band, such as Landsat TM and ETM+, LST retrieval is often more challenging. Commonly,
the single TIR channel (SC) method, also known as the model emissivity method is used to retrieve
LST from mono-TIR channel like the Landsat TM and ETM+ (Hook et al. 1992). The SC uses the
radiance measured by the satellite sensor in a single-channel chosen within an atmospheric window.
The radiance is then corrected for residual atmospheric attenuation and emission using atmospheric
transmittance/radiance code from atmospheric profiles data. LST is then retrieved from the radiance
measured in this channel by inverting the RTE, provided the LSE is well known or estimated in
advance (Price 1983; Susskind et al. 1984; Chedin et al. 1985; Hook et al. 1992; Ottlé & Vidal-Madjar
1992; LI et al. 2004; Mushkin et al. 2005). The radiance measured at a sensor can be transformed into
LST by inverting the Radiative Transfer Equation (RTE) applied to a particular thermal IR band or
wavelength as,
Lsensor = 𝜏𝜀LTs + Lu + 𝜏(1 − 𝜀)Ld (1)
where Lsensor is the radiance registered by the sensor, also referred to as top of atmosphere radiance,
LTs is the blackbody radiance related to the surface temperature by Planck’s law and Ts is the LST, Lu
and Ld are the upwelling and down-welling atmospheric radiances, respectively (all the radiances in
W sr−1 m−2 μm−1), τ is the atmospheric transmissivity and ε is the LSE. All these parameters are inte-
grated into the spectral response function of the waveband. Recently, several single-channel methods
458    A. A. Mohamed et al.

based on approximations from the RTE for single-channel sensors have been noted (Li et al. 2013a;
Tang & Li 2014). However, a major challenge when using SC algorithms is the high sensitivity to
the input atmospheric parameters, which leads to inaccurate LST retrievals under moderate to high
atmospheric absorption.
In the SW technique, adjacent channels (e.g. in ASTER and Landsat-8) with different properties
are used to calculate atmospheric attenuation. Using two thermal bands in the 10.5–12.5 μm region,
the SW algorithms have been widely used for estimating LST (Yu et al. 2008). However, as aforemen-
tioned, due to the two-channel requirement, this technique cannot be used in mono-channel bands
like the Landsat TM and ETM+. Adoption of most SW algorithms is based on the assumptions that
LST (Ts) is linearly related to the brightness temperature of two thermal channels of known surface
emissivity. This technique eliminates atmospheric effect based on differential absorption between the
two bands for the LST estimation (Du et al. 2015). A typical SW is expressed as:
Ts = a0 + a1 Ti + a2 Tj (2)
where ai are the coefficients, and Ti and Tj are the brightness temperature at bands i and j. Coefficients
ai depend on surface spectral emissivity and the atmospheric conditions, particularly when the atmos-
pheric humidity is very high (Becker & Li 1995). It should be noted that in all the SW algorithms,
brightness temperature rather than band radiance is used. According to Du et al. (2015), the SW
technique allows for more stable LST retrievals in different regions.

3.  Determination of LSE

Measurements of parameters from thermal remote sensors such as the LST depend on surface emissiv-
ity (Voogt & Oke 2003). According to Van de Griend and owe (1993), inaccuracy of 1% in emissivity
estimation can cause up to 0.78 K LST error. Emissivity is an intrinsic property of the surface and is
independent of irradiance. It is a function of a number of factors that include water content, chem-
ical composition, surface structure and roughness (Snyder et al. 1998). Quantification of emissivity
is achieved by considering the ratio of energy emitted by a surface in relation to the energy emitted
by a black body at the same temperature. However, calculations are often complicated as natural sur-
faces do not emit energy like a black body and thus need correction using typical emissivity values
(Table 1). Since LSE can change substantially over a small area, it is important to estimate its value
for every pixel prior to applying the RTE or SW. Several methods have been suggested to estimate the
emissivity for different TIR sensors. Generally, there are three distinctive ways to estimate LSE from
space; semi-empirical methods (SEMs), multi-channel temperature/emissivity separation methods
and physical-based methods (PBMs). Whereas all these methods can be used to determine emissivity
from space, with varying strengths and limitations, their applicability differ significantly (Dash et al.
2002; Li et al. 2013b). Recently, three major techniques have been proposed for LSE estimation before
LST inversion; temperature emissivity separation (TES), day/night temperature-independent spec-
tral-indices (TISI) and NDVI-based emissivity method (NBEM). These techniques are often taken as

Table 1. Typical emissivity values of common materials (Lillesand et al. 2004).

Material Typical average emissivity (over 8–14 μm)

Wet snow 0.98–0.99
Healthy green vegetation 0.96–0.99
Wet soil 0.95–0.98
Brick 0.93–0.94
Wood 0.93–0.94
Dry vegetation 0.88–0.94
Dry snow 0.85–0.90
Glass 0.77–0.81
Aluminium foil 0.03–0.07
 Geocarto International   459

reference for comparison to others (Li et al. 2013b). Generally, TISI and TES are theoretically robust
and do not require a priori information on the emissivity or surface type (Dash et al. 2005; Sobrino et
al. 2008). However, the wide adoption of TISI for instance has been limited by the need for estimated
atmospheric corrections and the MIR and TIR concurrence requirement (Sobrino & Raissouni 2000).
Then, the surfaces must be observed under similar observation conditions, e.g. viewing angle, during
both day and night (Dash et al. 2002, 2005). Furthermore, it requires accurate image co-registration
(Dash et al. 2005). The NBEM on the other hand is operationally simple and provides satisfactory LSEs
estimates for heterogeneous soil/vegetation landscapes (Dash et al. 2005; Li et al. 2013b), therefore, the
NBEM method has already been applied to various sensors with access to visible and near-infrared
(VNIR) and TIR bands (Sobrino & Raissouni 2000; Sobrino et al. 2003, 2008; Sobrino & Romaguera
2004; Momeni & Saradjian 2007). The TES algorithm was developed by Gillespie et al. (1998) to pro-
duce Standard surface temperature and emissivity products from ASTER data.
In this review, two commonly used methodologies for retrieval of LSE from TIR region of the
medium spatial scale (ASTER, Landsat Series and landsat-8) are discussed.

3.1.  NDVI-based emissivity methods (NBEM)

As earlier mentioned, Landsat TM/ETM+ is characterised by a single window for thermal data, hence
direct retrieval of emissivity without using other spectral bands and/or ancillary data and assumptions
is impossible. The NDVI-based methods, due to their simplicity, have gained popularity. One such
technique is the fractional cover mixture model. Under this technique, it is assumed that the soil
background and the vegetation have known emissivities that ‘mix’ according to the fractional cover
estimated from the NDVI (Sobrino et al. 2001). Some studies e.g. Valor and Caselles (1996) and Van
de Griend and owe (1993) have established empirical models between NDVI and emissivity. Since
pixels representing the land surface are usually mixed, i.e. are a combination of surfaces types such as
vegetation and soil, effective emissivity of a pixel can be estimated by summing up the contributions
from its respective surface types. The Botswana water and surface energy balance research programme
for instance generated a large series of emissivity measurements and retrieved visible and near-infrared
spectral reflectances to determine NDVI. Using emissivity-box and a combined red and near-IR radi-
ometer of spectral bands corresponding to AVHRR, spatial variation of thermal emissivity (8–14 μm)
and Normalised Difference Vegetation Index (NDVI) was determined for a series of natural surfaces
within a savanna environment in Botswana. According to Van de Griend and owe (1993), after log-
arithmic transformation, results established a high correlation (R2 = 0.94) between measured ε and
NDVI. The correlation can be expressed as:
𝜀 = a + b ln(NDVI) (3)
where a = 1.0094 and b = 0.047 derived from regression analysis.
Whereas, the relationship may be most suited for areas with characteristics similar to the Botswana
region, it can be adopted for other areas as it was based on AVHRR (Global Area Coverage and Local
Area Coverage) and Landsat (TM) data to demonstrate spatial variability of various spatial scales.
Inspired by the work of Van de Griend and owe (1993), using NDVI to describe the ‘cavity effect’, Valor
and Caselles (1996) proposed a theoretical method derived from the model generated by Caselles and
Sobrino (1989) to estimate the effective LSE for a row-distributed rough system in which the observed
radiance is the weighted sum of the radiances from the ground, top and the sides. In their method,
the effective LSE can be numerically expressed as the sum of the area-weighted emissivities of each
sub-component and the cavity effect as functions of the NDVI. Valor and Caselles (1996) focused on
the 10.5–12.5 μm region, where most thermal sensors on board satellites work. From the theoretical
model, which defines the effective emissivity of the surface from the emissivity of their simple com-
ponents, soil and vegetation, Caselles and Sobrino (1989) developed an operational methodology to
obtain the effective emissivity combining satellite images and field measurements. The error of the
methodology range from 0.5% (due to experimental limitations of field methods) to 2% (considering
460    A. A. Mohamed et al.

the case in which no information is available about the studied areas). The model was validated under
different atmospheric conditions and degrees of roughness in the mid-latitudes and tropical regions
and flat and rough surfaces, concluding that it was applicable on varied soil and vegetation types
with an estimated error of 0.6% on emissivity. An operational model of this method can be found in
Sobrino et al. (2008). Generally, the methodology assigns different values of emissivity for each pixel
depending on its NDVI value. In areas with NDVI values lower than 0.2, the pixels are considered as
bare soil and the emissivity is obtained from reflectance values in the red section of the electromag-
netic spectrum expressed as;
𝜀 = a + b𝜌red (4)
where ε is the emissivity, ρred is the reflectance in the red region and a and b are coefficients obtained
from laboratory spectra of soils and statistical fits (Sobrino et al. 2008).
When NDVI values are greater than 0.2 but lower than 0.5, the pixels are assumed to be a mixture
of bare soil and vegetation, and the emissivity is determined using the expression;
𝜀 = m ⋅ Pv + n (5)
with
m = 𝜀v − 𝜀s − (1 − 𝜀v)F 𝜀v (6)

n = 𝜀s + (1 − 𝜀s )F 𝜀v (7)
where εv is the vegetation emissivity and εs is the soil emissivity, Pv is the proportion of vegetation
obtained according to Carlson and Ripley (1997) and F is a shape factor, whose mean value is 0.55.
When the NDVI obtained is higher than 0.5, the pixel is considered as fully vegetated, and a constant
value for the emissivity is assumed, typically 0.99.

3.2.  Temperature emissivity separation (TES) method

The TES algorithm was developed by Gillespie et al. (1998). Whereas it was designed for the ASTER
thermal imagery (with 14 spectral channels of which 5 channels operate between 8 and 12 μm), Sobrino
and Jiménez-Muñoz (2014) have shown that a three- band (8.9–12 μm) configuration is capable of
providing similar results than a four or five-band configuration. The input parameters are the up- and
down-welling atmospheric radiance at the relevant wavelengths. The methodology is based on three
modules, the Normalised Emissivity Method (NEM) obtains preliminary values of LST and LSE,
the Ratio module estimates the emissivity normalised spectrum (β spectrum), and the Maximum
Minimum Difference (MMD) transforms the β spectrum into the absolute LSE and re-calculates the
LST using an empirical relation between the minimum emissivity and the spectral contrast (Oltra-
Carrió et al. 2012). See below for a brief explanation of each module.
NEM module: The NEM method, proposed by Gillespie (1985) assumes a constant emissivity for
all N channels in order to calculate N ground temperatures (TG). For each pixel, the maximum of
these N temperatures (TGmax) is taken as the surface temperature (TS), which is then used to estimate
emissivities for the other channels. This method is used as a separate module in the TES algorithm.
The emissivity ε(λ) -ratio between the radiance emitted by an object at wavelength λ and that emitted
by a black body at the same temperature is initially set to 0.97, in order to be within ± 0.03 for typical
land surfaces like vegetation, snow, soil and rocks (Schmugge et al. 1998). The reflected sky irradiance
is removed iteratively. The temperature passed to the next ratio module should be within ± 3 K at
340 K and within ± 2 K at 273 K.
Ratio module: The temperature T (the temperature of a Lambertian reflector) from the NEM
module is used to calculate emissivity ratios β (Equation 8). The emissivity ratios βi are calculated by
dividing the atmospherically corrected radiances by the average of the channels:
 Geocarto International   461

Ri B(T)
𝛽i = (8)
̄ i (T)
RB

where
N N
∑ ∑
R̄ = 1∕N Ri and B(T) = 1∕N Bi (T) (9)
i=1 i=1

For emissivities between 0.7 and 1.0, the ratios βi are generally within 0.7–1.4.
Maximum–minimum difference (MMD) module: The maximum–minimum difference between the
emissivity ratios βi is given by MMD = max(βi) – min(βi). An empirical relation between minimum
emissivity and MMD is:
0.737
𝜀min = 0.994 − 0.687 ⋅ MMD (10)

A relationship is determined by regression analysis using emissivity measured in the laboratory and
in a field experiment (Salisbury & D’Aria 1992). The MMD often ranges from 0.0 to 0.4.
From εmin given by Equation (11), the emissivities in all channels are calculated on a pixel-by-pixel
basis:

(11)
[ ]
𝜀i = 𝛽i 𝜀min ∕min(𝛽i )

Numerical modelling showed that TES can recover LST to within ~1.5 K and emissivity to within
0.015 (Gillespie et al. 1998).

4.  Medium-resolution LST sensors

As earlier mentioned, whereas there are several satellite remote-sensing platforms with sensors in
the TIR spectrum relevant to urban LST applications. This review is restricted to the potential for the
commonly used medium spatial resolution satellite-based sensors.
As highlighted in Section (2), different methods and algorithms have been developed to retrieve
LST and LSE from TIR data.
(1) The Single-Channel (SC) requires only one TIR band and can therefore be used on Landsat
TM and ETM+. They are based on the direct inversion of the RTE.
(2) The two-channel (TC) or split-window (SW) can be used on ASTER and Lansat-8, which
require two TIR bands.
(3) The temperature and emissivity separation (TES) algorithm, requires multispectral TIR data,
with at least three TIR bands and can be used on ASTER.

4.1.  Landsat series

The Landsat series of satellites are probably the most popular sensors, with the longest record of earth
observations. The Thematic Mapper (TM) on Landsat 4 and 5 had a visible resolution of 30 m and
a TIR resolution of 120 m (band 6, 10.4–12.5 μm). Landsat 4 and 5 are no longer in consistent data
collection. The Landsat 7’s Enhanced Thematic Mapper (ETM+) is a near-polar sun synchronous
orbit sensor with a 16-day temporal resolution. With a 60 m thermal band at 10.4 to 12.5 μm, it is
one of the highest thermal resolution measurements from space. However, data from 2003 onwards
has been impaired by the failed scan line corrector and only ∼80% of each scene is captured. The free
462    A. A. Mohamed et al.

availability of the Landsat data archive since 2008 has seen an increase in the adoption of the data-
sets. However, adoption of the Landsat data is commonly impeded by inability to collect night data
and limited thermal calibration. For satellites with a single thermal band, such as Landsat TM and
ETM+, obtaining LST is often difficult. An accurate radiative transfer model and knowledge on the
atmospheric profile and emissivity information is also required. For this reason, despite the higher
spatial resolution valuable for urban thermal applications, Landsat TM/ETM+ thermal data have
not been widely used in urban environment, however, they remain popular for regional and global
LST studies (Masek & Collatz 2006; Li et al., 2012). In addition, the 16-day revisit time, may prohibit
its application (Bai et al. 2015). The revisit cycle for a particular area may be extended due to poor
atmospheric conditions, such as cloud and haze (Ju & Roy 2008). According to Stathopoulou and
Cartalis (2007), future studies may focus on a time series of images as the UHI strongly depends on
synoptic weather conditions.

4.2. Landsat-8
The Landsat-8 was successfully launched on 11 February 2013 and deployed into orbit with two
instruments onboard; the Operational Land Imager (OLI) with nine spectral bands in the visible (VIS),
near infrared (NIR), and the shortwave infrared (SWIR) spectral regions and the Thermal Infrared
Sensor (TIRS) with two spectral bands in the long-wave infrared (LWIR). The TIRS spatial resolution
is 100 m, with a 16-day revisit time. Landsat-8 images are currently freely distributed through the
United States Geological Survey (USGS). The two TIRS bands were selected to enable the atmospheric
correction of the thermal data using a split window algorithm (SW) (Irons et al. 2012; Cuenca et al.
2013). According to Caselles et al. (1998), the use of two separate, relatively narrow, thermal bands has
been shown to minimise error in the retrieval of LST. Although Landsat-8 has two thermal infrared
channels in the atmospheric window that provide a new LST retrieval opportunity using the widely
used split-window algorithm rather than the single-channel method, some adaptations are required
in order to implement them for the TIRS spectral bands. Du et al. (2015) and Rozenstein et al. (2014)
suggest an adjusted split window algorithm for estimating land surface temperature from Landsat-8
Data. To the best of our knowledge, no implementation for LST retrieval from TIRS for SUHI assess-
ment has been published using Landsat-8.

4.3.  Advanced Space-borne thermal emission and reflection radiometer

The ASTER sensor is on the NASA Terra satellite platform. It is the only satellite sensor with mul-
tispectral TIR channels necessary for the TES algorithm. It has three visible and near-infrared (VNIR)
bands with a 15 m spatial resolution, nine shortwave infrared (SWIR) bands (30 m spatial resolution)
and five TIR bands (90 m spatial resolution). Further technical details are available in Yamaguchi et
al. (1998). A number of studies have adopted ASTER imagery in determining SUHI. Using linear
spectral mixture analysis and multiple regression models, Lu and Weng (2006) compared LST to
urban biophysical descriptors (such as impervious surfaces, green vegetation and soil) in Indianapolis,
USA using LSTs computed from ASTER’s band 13 (10.25–10.95 μm). A relationship between LST and
the derived fraction variables across ASTER’s pixels spatial resolutions (15 m to 90 m) and multiple
regression models were developed to determine how LSTs were related to the biophysical descrip-
tors to the thermal feature fractions (i.e. hot- and cold objects). Results indicated that impervious
surfaces and vegetation were positively and negatively correlated to LST, respectively. Hot objects
were more influential on LST patterns than cold objects. Nichol et al. (2009) used ASTER in concert
with a 148 km vehicle traverse of Hong Kong to compare air and remotely sensed thermal values and
concluded that thermal satellite images combined with ‘in situ’ ground data can be used to examine
models of heat island genesis and thus identify the main causes of UHI. The models, although pro-
posed over 30 years ago, have not been thoroughly evaluated due to a combination of inadequate
ground data and lack of high resolution thermal satellite data. Furthermore, there has been limited
 Geocarto International   463

understanding of the relevance of satellite-derived surface temperatures to local- and regional-scale

air temperatures. A cloud-free ASTER thermal image of urban and rural areas of Hong Kong was
obtained in a winter night with a well-developed heat island, accompanied by a 148 km vehicle traverse
of air temperatures. Over the whole traverse, a high R2 value of 0.80 was observed between surface
and air temperatures, with the two data-sets showing a similar amplitude and general trend, but with
the surface exhibiting much higher local variability than air temperature. Gradients in both surface
and air temperature could be related to differences in land cover, with little evidence of large-scale
advection, thus supporting the population/physical structure model of UHI causation, rather than the
advection model. However, the much higher surface and air temperatures observed over the largest
urban area, Kowloon, than over any smaller urban centre with similar physical structure in the New
Territories, would seem more indicative of the advection model. The image and ground data suggest
that Kowloon’s urban canopy layer climate is mainly influenced by local city structure, but it is also
modified by a strongly developed, regional-scale urban boundary layer which has developed over the
largest urban centre of Kowloon, and reinforces heating from both above and below. Cao et al. (2010)
combined ASTER imagery with IKONOS data to explore the cooling effect of urban parks in Nagoya,
Japan. Urban parks can help mitigate UHI effects and decrease cooling energy consumption in summer.
However, it is unclear how park characteristics affect the formation of a park cool island (PCI). In the
study, PCI intensity values for 92 parks in Nagoya were obtained from ASTER LST products and then
correlated to detailed information derived from high spatial resolution IKONOS satellite data. The
results indicated that (1) the cooling effect depends on the park size and seasonal radiation condition,
and park size is non-linearly correlated to PCI intensity; (2) PCI intensity is mainly determined by
the area of tree and shrub inside the park as well as the park shape, and grass has negative impact on
PCI formation. The park vegetation and shape index (PVSI) proposed well-predicted PCI intensity
of selected parks. These findings can help urban planners to understand PCI formation and design
cool parks to counteract UHI effects.
Unlike other sensors (e.g. Landsat series) discussed in this review, ASTER is request-only product
with payable fees. Consequently, the sensor’s historical data is limited and costly. ASTER’s Terra plat-
form has a twice daily overpass and the 90 m spatial resolution, valuable for urban thermal mapping.
However, limited archival data and difficulties in ensuring suitable atmospheric and weather conditions
for a specific future request impedes its wide adoption in SUHI studies.

5.  LST validation

Studies on among others evapotranspiration, climate change, surface energy budgets, the hydrological
cycle, vegetation monitoring and urban climate require knowledge on LST (Ramanathan et al. 2001;
Kalnay & Cai 2003; Wan et al. 2004; Chapin et al. 2005). Consequently, reliable spatial and temporal
LST estimates are critical (Li et al. 2013c). Many algorithms have been proposed and developed for
estimating the LST. However, major challenges when estimating LST from satellite data include cor-
rections for the effects of the atmosphere and the LSE (Li et al. 2013b). To date, addressing the above
remains a challenge. This is due to coupled LST, LSEs and the downward atmospheric radiance and
the large variability in the vertical profiles of atmospheric water vapour and temperature, which results
in highly variable atmospheric absorption, emission and emission-reflection. Research on validation
of satellite-derived LST have been focused on satellite sensors using thermal-infrared (TIR) channels,
including Landsat TM/ETM+, ASTER, AVHRR, Along-Track Scanning Radiometer (ATSR), Advanced
Along-Track Scanning Radiometer (AATSR), Moderate Resolution Imaging Spectroradiometer
(MODIS), Spinning Enhanced Visible and Infrared Imager (SEVIRI) and Visible Infrared Imaging
Radiometer Suite (VIIRS) data (Wan et al. 2002; Sobrino et al. 2004; Coll et al. 2005, 2012; Hook et al.
2007; Trigo et al. 2008; Wan 2008; Wan & Li 2008; Schneider et al. 2009; Freitas et al. 2010; Göttsche et
al. 2013; Guillevic et al. 2014; Qian et al. 2013; Li et al. 2014). The main focus of the present work is to
provide recommendations on the suitable methods for medium spatial resolutions sensors (60–120 m
spatial resolution and TIR: 8–15 μm) LST validation. The methods proposed for LST validation can
464    A. A. Mohamed et al.

generally be grouped into three: temperature-based method (T-based), radiance-based (R-based) and
inter-comparison (Tang et al. 2015).
The T-based method directly compares the satellite-derived LST with in situ LST measurements
at the satellite overpass (Wan et al. 2002; Sobrino et al. 2004; Coll et al. 2005; Hook et al. 2007; Wan
2008; Schneider et al. 2009; Freitas et al. 2010; Göttsche et al. 2013; Guillevic et al. 2014). This method
relies on the accuracy of the ground-measured LSTs and their representation at the satellite pixel scale.
Because the LST varies significantly over space and time, the T-based method is often restricted to
homogeneous surfaces, such as lakes and densely vegetated areas (Tang et al. 2015).
The R-based method is based on the satellite-derived LST and the in situ atmospheric profiles
measured at the satellite overpass, as well as the LSE as initial inputs to simulate the radiance at TOA.
Comparing the simulated satellite-measured TOA radiances and adjusting the initial LST to minimise
the difference between the simulated and measured radiances, the accuracy of the retrieved LST is
obtained from the difference between the adjusted LST and the initial satellite-derived LST (Coll et al.
2012; Wan & Li 2008). Whereas this method does not require ground LST measurements, it requires
LSE and measured atmospheric profiles at the satellite overpass. It depends on the accuracies of the
atmospheric profiles and the LSE at the pixel size. As suggested in previous studies (Wan & Li 2008;
Coll et al. 2012; Li et al. 2013c; Vlassova et al. 2014), LSTs simulated using radiative transfer code such
as MODTRAN 4,5 can be an alternative for validation when field measurements at a required spatial
scale are not available. The method has been previously applied to Landsat TM and MODIS sensors.
With regard to Landsat TM data, only values of emissivity measured in situ and radiosonde data were
available. From at-sensor data extracted from the Landsat image, emissivity values and atmospheric
parameters obtained from the radiosonde and the MODTRAN 3.5 code, LST values have been repro-
duced using a generalised single-channel method developed by Jiménez-Muñoz and Sobrino (2003).
The validation for different test sites shows root mean square deviations lower than 1.5 K for Landsat
Thematic Mapper (TM) band 6 (11.5 μm) (Jiménez-Muñoz & Sobrino 2003).
The V5 level 2 LST product of MODIS was validated over homogeneous rice fields in Valencia,
Spain, and the Hainich forest in Germany. For the Valencia site, ground LST measurements were com-
pared with the MOD11_L2 product in the conventional temperature-based (T-based) method. The
validation also applied the alternative radiance-based (R-based) method, with in situ LSTs calculated
from brightness temperatures in band 31 through radiative transfer simulations using temperature
and water vapour profiles and surface emissivity data. Profiles at the site were obtained from local
radiosonde measurements and from National Centres for Environmental Prediction (NCEP) data.
The R-based method was applied at the Hainich site using radiosonde profiles from a nearby sounding
station and NCEP profiles. The T-based validation showed average bias (MODIS minus ground) of
−0.3 K, standard deviation of 0.6 K and root mean square error (RMSE) of ± 0.7 K. For the R-based
method, the quality of the atmospheric profiles was assessed through the difference δ (T31–T32)
between the actual MODIS and the profile-based calculated brightness temperature difference in
bands 31 and 32. For the cases where −0.3 K < δ (T31–T32) < 0.5 K, the R-based method yielded LST
errors with small biases and RMSE = ±0.6 K for the two sites. These results show the high accuracy
and precision of the MODIS LST product for the two sites studied. The good performance of the
R-based method opens the possibility for a more complete validation including heterogeneous surfaces
where the T-based method is not feasible (Coll et al. 2009). Wan and Li (2008) reported radiance‐
based validation approach for MODIS - LST product. Surface emissivity spectra were retrieved by a
sun‐shadow method from surface‐leaving radiance spectra measured with a thermal infrared (TIR)
spectroradiometer in the 3.5–14 μm spectral region under sunshine and sun‐shadow conditions. By
using the measured surface emissivity spectrum and atmospheric profiles obtained by radiosonde
balloons and LST values at validation sites in the V5 MODIS level‐2 LST products, radiative transfer
simulations were made with the MODTRAN4 code to calculate the TOA radiance values in MODIS
band 31 (L31). By adjusting the LST input values in the simulations to match the calculated L31 values
to the MODIS measured radiance (MOD L31) values, MOD L31 inverted LSTs can be obtained. The
 Geocarto International   465

MODIS LST product was validated by comparison to the values of the MOD L31 inverted LSTs. This
approach compares well with the conventional temperature‐based approach. The results of the radi-
ance‐based validation indicate that the accuracy of the MODIS LST product is better than 1 K in most
cases, including lake, vegetation and soil sites in clear‐sky conditions. The errors in the split‐window
retrieved LSTs may be larger in bare soil and highly heterogeneous sites due to large uncertainties
in surface emissivities. The results of the radiance‐based validation also reveal the weakness of the
split‐window algorithm used for the generation of the MODIS LST product in two situations: one in
cases where LSTs are larger than the air temperature at the surface level (Ts‐air) by more than 16 K
and the columnar water vapour is larger than 1.5 cm and under the influence of thin cirrus clouds or
heavy aerosol loadings. These two situations were not considered in the development of the current
MODIS LST algorithm.
The most important improvements in MODTRAN 5, compared to MODTRAN 4 is the incor-
poration of band model parameters based on High-Resolution Transmission (HITRAN2008), with
2009 updates (Berk et al. 2011). MODTRAN 5 performs calculations based on the information about
observation geometry and atmospheric profiles at the moment of observation. The best results are
achieved when data from in situ radio soundings are synchronised with image acquisition.
The inter-comparison method only compares the satellite-derived LST with a well-validated LST
product from another satellite (Trigo et al. 2008; Qian et al. 2013; Li et al. 2014). This method does not
require any ground measurements and can be used anywhere, if well-validated LST product is available.
However, because of the large spatial and temporal variations in the LST, this method is very sensitive
to the scale differences over space, time and viewing angle between the two sensors (Tang et al. 2015).

6.  Future development

Whereas several studies have reported on the SUHI effect using different medium spatial resolution
imagery, e.g. ASTER and Landsat series, the existing satellite remote-sensing capabilities (temporal,
spatial and spectral resolutions) are not ideal for monitoring SUHI. Future increases in sensor spatial
resolutions in concert with high temporal resolution is likely to offer useful data for determining SUHI
effect (Sobrino et al. 2012).
The operational medium spatial resolution sensors (i.e. Landsat series), have only one TIR bands. In
this framework, only LST retrieval schemes such as SC are possible. For Landsat-8, which is currently
available for SUHI research, SW algorithms are applicable. However, input emissivities (LSEs) are esti-
mated by independent measurements, e.g. NDVI-based approach (i.e. not using TIR measurements)
is required for both different LST retrieval schemes.
For accurate LST estimation to address local climate applications over urban areas, a multispectral
configuration is required. A minimum configuration with only three TIR bands would be useful for
LST/LSE retrieval. This configuration should include one TIR band in the 8–9 μm spectral region,
and two TIR bands in the 10–12 μm spectral window, enabling use of both TES and SW algorithms
for LST retrieval (Sobrino & Jiménez-Muñoz 2014). Using the TES, Atmospheric correction is critical
for successful LST retrieval algorithm. It is therefore important to complement the TIR sensor with
appropriate bands for atmospheric characterisation (e.g. water vapour retrieval).
Higher spatial resolutions, which may offer higher classification accuracy, involve lower temporal
frequency and lower swath at nadir. Currently, no space-borne thermal sensors satisfy necessary spa-
tial/temporal requirements concurrently. The proposed 50 m resolution can be achieved with a revisit
time of 2 days (one day if two satellites are operative) and a swath of approximately 25 km, which
is sufficient to monitor most larger towns (Sobrino et al. 2012). It is therefore necessary that future
efforts focus on reduced pixel dimensions and increased temporal frequency. In concert with the large
‘traditional’ missions, there is an expected increase in ‘small satellites’ with enhanced spatial, spectral
and temporal resolutions valuable for intra-urban SUHI studies (Sandau et al. 2010).
466    A. A. Mohamed et al.

7. Conclusions
This review has provided an overview of remote sensing techniques, sensors and studies on LST
retrieval of intra-urban SUHI characteristics from medium-resolution sensors. Whereas conventional
ground-based temperature monitoring techniques are used for determining air temperature within
urban landscapes, they are often unsuitable in determining spatial thermal characteristics. Therefore,
despite challenges associated with the use of LST products for SUHI research, existing medium spatial
resolution sensors may offer more reliable UHI assessment.
The requirement for validation and accuracy assessment necessitate adoption of remotely sensed
data in concert with other ancillary datasets. In light of existing challenges using low spatial resolution
remotely sensed thermal data, this review has highlighted the potential of existing medium-resolution
images in determining thermal variability within urban landscapes.

Acknowledgements
We would like to thank the University of KwaZulu-Natal for funding the research. The authors would also like to thank
the anonymous reviewers for their valuable comments.

References
Arnfield AJ. 2003. Two decades of urban climate research: a review of turbulence, exchanges of energy and water, and
the urban heat island. Int J Climatol. 23:1–26.
Bai Y, Wong MS, Shi W-Z, Wu L-X, Qin K. 2015. Advancing of land surface temperature retrieval using extreme learning
machine and spatio-temporal adaptive data fusion algorithm. Remote Sens. 7:4424–4441.
Barsi JA, Barker JL, Schott JR. 2003. An atmospheric correction parameter calculator for a single thermal band earth-
sensing instrument. In: Processdings of the 2003 IEEE International Geoscience and Remote Sensing Symposium
IGARSS ‘03. Melbourne, Australia, 21–25 July 2003; Volume 5, p. 3014–3016.
Barsi JA, Schott JR, Palluconi FD, Hook SJ. 2005. Validation of a web-based atmospheric correction tool for single
thermal band instruments. In: JJ Butler, editor. Proceedings of SPIE, Earth Observing Systems X. Volume 5882, Paper
58820E, Bellingham, WA: SPIE Aug 22, 2005.
Basara JB, Basara HG, Illston BG, Crawford KC. 2010. The impact of the Urban Heat Island during an intense heat wave
in Oklahoma city. Adv Meteorol. Article ID. 230365:1–10.
Becker F, Li Z-L. 1995. Surface temperature and emissivity at various scales: definition, measurement and related
problems. Remote Sens Rev. 12:225–253.
Berk A, Anderson GP, Acharya PK, Shettle EP. 2011. ModTran 5.2.1 user’s manual. Burlington, MA: Spectral Sciences,
Inc. & Air Force Research Laboratory.
Cao X, Onishi A, Chen J, Imura H. 2010. Quantifying the cool island intensity of urban parks using ASTER and IKONOS
data. Landsc Urban Plan. 96:224–231.
Carlson TN, Ripley DA. 1997. On the relation between NDVI, fractional vegetation cover, and leaf area index. Remote
Sens Environ. 62:241–252.
Caselles V, Sobrino JA. 1989. Determination of frosts in orange groves from NOAA-9 AVHRR data. Remote Sens
Environ. 29:135–146.
Caselles V, Rubio E, Coll C, Valor E. 1998. Thermal band selection for the PRISM instrument: 3. Optimal band
configurations. J Geophys Res. 103:17057–17067.
Chakraborty SD, Kant Y, Mitra D. 2015. Assessment of land surface temperature and heat fluxes over Delhi using remote
sensing data. J Environ Manage. 148:143–152.
Chapin F, Sturm M, Serreze M, McFadden J, Key J, Lloyd A, McGuire A, Rupp T, Lynch A, Schimel J. 2005. Role of
land-surface changes in Arctic summer warming. Science. 310:657–660.
Chedin A, Scott NA, Wahiche C, Moulinier P. 1985. The improved initialization inversion method: a high resolution
physical method for temperature retrievals from satellites of the TIROS-N series. J Clim Appl Meteorol. 24:128–143.
Chen XL, Zhao H, Li P, Yin Z. 2002. Remote sensing image-based analysis of the relationship between urban heat island
and land use/cover changes. Remote Sens Environ. 104:133–146.
Chow WTL, Svoma BM. 2011. Analyses of nocturnal temperature cooling-rate response to historical local-scale urban
land-use/land cover change. J Appl Meteor Climatol. 50:1872–1883.
Coll C, Caselles V, Galve JM, Valor E, Niclos R, Sanchez JM, Rivas R. 2005. Ground measurements for the validation of
land surface temperatures derived from AATSR and MODIS data. Remote Sens Environ. 97:288–300.
Coll C, Wan Z, Galve JM. 2009. Temperature-based and radiance-based validations of the V5 MODIS land surface
temperature product. J Geophys Res. 114:D20102.
 Geocarto International   467

Coll C, Valor E, Galve JM, Mira M, Bisquert M, García-Santos V, Caselles E, Caselles V. 2012. Long-term accuracy
assessment of land surface temperatures derived from the Advanced Along-Track Scanning Radiometer. Remote
Sens Environ. 116:211–225.
Crutzen PJ. 2004. New directions: The growing urban heat and pollution “island” effect –impact on chemistry and
climate. Atmos Environ. 38:3539–3540.
Cuenca RH, Ciotti SP, Hagimoto Y. 2013. Application of Landsat to evaluate effects of irrigation forbearance. Remote
Sens. 5:3776–3802.
Cui YY, De Foy B. 2012. Seasonal variations of the Urban Heat Island at the surface and the near-surface and reductions
due to urban vegetation in Mexico city. J Appl Meteorol Climatol. 51:855–868.
Curriero FC, Heiner KS, Samet JM, Zeger SL, Strug L, Patz JA. 2002. Temperature and mortality in 11 cities of the
Eastern United States. Am J Epidemiol. 155:80–87.
Dash P, Göttsche F-M, Olesen F-S, Fischer H. 2001. Retrieval of land surface temperature and emissivity from satellite
data: physics, theoretical limitations and current methods. J Indian Soc Remote. 29:23–30.
Dash P, Göttsche F-M, Olesen F-S, Fischer H. 2002. Land surface temperature and emissivity estimation from passive
sensor data: theory and practice-current trends. Int J Remote Sens. 23:2563–2594.
Dash P, Göttsche F-M, Olesen F-S, Fischer H. 2005. Separating surface emissivity and temperature using two-channel
spectral indices and emissivity composites and comparison with a vegetation fraction method. Remote Sens Environ.
96:1–17.
Du C, Ren H, Qin Q, Meng J, Zhao S. 2015. A practical split-window algorithm for estimating land surface temperature
from Landsat 8 data. Remote Sens. 7:647–665.
Freitas SC, Trigo IF, Bioucas-Dias JM, Gottsche F-M. 2010. Quantifying the uncertainty of land surface temperature
retrievals from SEVIRI/Meteosat. IEEE Trans Geosci Remote Sens. 48:523–534.
Friedl MA, McIver DK, Hodges JCF, Zhang XY, Muchoney D, Strahler AH, Woodcock CE, Gopal S, Schneider A,
Cooper A, Baccini A, Gao F, Schaaf C. 2002. Global land cover mapping from MODIS: algorithms and early results.
Remote Sens Environ. 83:287–302.
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.
Gallo KP, Owen TW. 1998. Assessment of Urban Heat Islands: a multi‐sensor perspective for the Dallas‐Ft. worth, USA
region. Geocarto Int. 13:35–41.
Gallo KP, Owen TW. 1999. Satellite-based adjustments for the Urban Heat Island temperature bias. J Appl Meteor.
38:806–813.
Gao ZQ, Gao W, Chang NB. 2011. Integrating temperature vegetation dryness index (TVDI) and regional water stress
index (RWSI) for drought assessment with the aid of Landsat TM/ETM plus images. Int J Appl Earth Obs Geoinf.
13:495–503.
Gillespie AR. 1985. Lithologic mapping of silicate rocks using TIMS. In: Proceedings TIMS Data User’s Workshop, JPL
Pub. 86-38, Pasadena, CA: Jet Propulsion Laboratory, p. 29–44.
Gillespie AR, Rokugawa S, Matsunaga T, Cothern JS, Hook SJ, Kahle AB. 1998. A temperature and emissivity separation
algorithm for Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) images. IEEE Trans
Geosci Remote Sens. 36:1113–1126.
Göttsche F-M, Olesen F-S, Bork-Unkelbach A. 2013. Validation of land surface temperature derived from MSG/SEVIRI
with in situ measurements at Gobabeb, Namibia. Int J Remote Sens. 34:3069–3083.
Guillevic PC, Biard J, Hulley GC, Privette JL, Hook SJ, Olioso A, Göttsche F-M, Radocinski R, Román MO, Yu Y, Csiszar
I. 2014. Validation of land surface temperature products derived from the Visible Infrared Imager Radiometer Suite
(VIIRS) using ground-based and heritage satellite measurements. Remote Sens Environ. 154:19–37.
Hage K. 2003. On destructive Canadian prairie windstorms and severe winters – a climatological assessment in the
context of global warming. Nat Hazards. 29:207–228.
Hartz DA, Prashad L, Hedquist BC, Golden J, Brazel AJ. 2006. Linking satellite images and hand-held infrared
thermography to observed neighborhood climate conditions. Remote Sens Environ. 104:190–200.
Henry JA, Dicks SE, Wetterqvist OF, Roguski SJ. 1989. Comparison of satellite, ground- based, and modeling techniques
for analyzing the urban heat-island. Photogramm Eng Remote Sens. 55:69–76.
Hook SJ, Gabell AR, Green AA, Kealy PS. 1992. A comparison of techniques for extracting emissivity information from
thermal infrared data for geologic studies. Remote Sens Environ. 42:123–135.
Hook SJ, Vaughan RG, Tonooka H, Schladow SG. 2007. Absolute radiometric in-flight validation of mid infrared and
thermal infrared data from ASTER and MODIS on the terra spacecraft using the lake tahoe, CA/NV, USA, automated
validation site. IEEE Trans Geosci Remote Sens. 45:1798–1807.
Huang L, Li J, Zhao D, Zhu J. 2008. A fieldwork study on the diurnal changes of urban microclimate in four types of
ground cover and urban heat island of Nanjing, China. Build Environ. 43:7–17.
Irons JR, Dwyer JL, Barsi JA. 2012. The next Landsat satellite: the Landsat data continuity mission. Remote Sens Environ.
122:11–21.
Jiang GM, Li Z-L, Nerry F. 2006. Land surface emissivity retrieval from combined mid-infrared and thermal infrared
data of MSG-SEVIRI. Remote Sens Environ. 105:326–340.
468    A. A. Mohamed et al.

Jiménez-Muñoz JC, Sobrino JA. 2003. A generalized single-channel method for retrieving land surface temperature
from remote sensing data. J Geophys Res. 108:4688–4695.
Johnson DP, Wilson JS, Luber GC. 2009. Socioeconomic indicators of heat-related health risk supplemented with
remotely sensed data. Int J Health Geogr. 8:57.
Ju J, Roy DP. 2008. The availability of cloud-free Landsat ETM+ data over the conterminous United States and globally.
Remote Sens Environ. 112:1196–1211.
Kalnay E, Cai M. 2003. Impact of urbanization and land-use change on climate. Nature. 423:528–531.
Konopacki S, Akbari H. 2002. Energy savings for heat island reduction strategies in Chicago and Houston (including
updates for Baton Rouge, Sacramento and Salt Lake City). Draft Final Report, LBNL – 49638. California: University
of California, Berkeley.
Li Z-L, Petitcolin F, Zhang RH. 2000. A physically based algorithm for land surface emissivity retrieval from combined
mid-infrared and thermal infrared data. Science in China (Series E). 43:23–33.
LI F, Jackson TJ, Kustas WP, Schmugge TJ, French AN, Cosh M, Bindlish R. 2004. Deriving land surface temperature
from Landsat 5 and 7 during SMEX02/SMACEX. Remote Sens Environ. 92:521–534.
Li Y-Y, Zhang H, Kainz W. 2012. Monitoring patterns of urban heat islands of the fast-growing Shanghai metropolis,
China: using time-series of Landsat TM/ETM+ data. Int J Appl Earth Obs Geoinf. 19:127–138.
Li Z-L, Tang B-H, Wu H, Ren H, Yan G, Wan Z, Trigo IF, Sobrino JA. 2013a. Satellite-derived land surface temperature:
current status and perspectives. Remote Sens Environ. 131:14–37.
Li Z-L, Wu H, Wang N, Qiu S, Sobrino JA, Wan Z, Tang B-H, Yan G. 2013b. Land surface emissivity retrieval from
satellite data. Int J Remote Sens. 34:3084–3127.
Li X, Zhou W, Ouyang Z. 2013c. Relationship between land surface temperature and spatial pattern of greenspace: what
are the effects of spatial resolution? Landsc Urban Plan. 114:1–8.
Li H, Sun D, Yu Y, Wang H, Liu Y, Liu Q, Du Y, Wang H, Cao B. 2014. Evaluation of the VIIRS and MODIS LST products
in an arid area of Northwest China. Remote Sens Environ. 142:111–121.
Lillesand T, Kiefer R, Chipman J. 2004. Remote sensing and image interpretation. Chichester: John Wiley.
Lu D, Weng Q. 2006. Spectral mixture analysis of ASTER images for examining the relationship between urban thermal
features and biophysical descriptors in Indianapolis, Indiana, USA. Remote Sens Environ. 104:157–167.
Mallick J, Rahman A, Singh CK. 2013. Modeling urban heat islands in heterogeneous land surface and its correlation
with impervious surface area by using night-time ASTER satellite data in highly urbanizing city, Delhi-India. Adv
Space Res. 52:639–655.
Masek JG, Collatz GJ. 2006. Estimating forest carbon fluxes in a disturbed southeastern landscape: Integration of remote
sensing, forest inventory, and biogeochemical modeling. J Geophys Res. 111. Doi:10.1029/2005JG000062
McMillin LM, Crosby DS. 1984. Theory and validation of the multiple window sea surface temperature technique. J
Geophys Res. 89:3655–3661.
Momeni M, Saradjian M. 2007. Evaluating NDVI-based emissivities of MODIS bands 31 and 32 using emissivities
derived by Day/Night LST algorithm. Remote Sens Environ. 106:190–198.
Morris CJG, Simmonds I, Plummer N. 2001. Quantification of the influences of wind and cloud on the nocturnal Urban
Heat Island of a large city. J Appl Meteor. 40:169–182.
Mushkin A, Balick LK, Gillespie AR. 2005. Extending surface temperature and emissivity retrieval to the mid-infrared
(3–5 μm) using the Multispectral Thermal Imager (MTI). Remote Sens Environ. 98:141–151.
Nichol JE, Fung WY, Lam K, Wong MS. 2009. Urban Heat Island diagnosis using ASTER satellite images and ‘in situ’
air temperature. Atmos Res. 94:276–284.
Oltra-Carrió R, Sobrino JA, Franch B, Nerry F. 2012. Land surface emissivity retrieval from airborne sensor over urban
areas. Remote Sens Environ. 123:298–305.
Ottlé C, Vidal-Madjar D. 1992. Estimation of land surface temperature with NOAA9 data. Remote Sens Environ.
40:27–41.
Precipitable Water at the VLA. 1990–1998. Available from: http://legacy.nrao.edu/alma/memos/html-memos/alma237/
memo237.html
Price JC. 1983. Estimating surface temperatures from satellite thermal infrared data – A simple formulation for the
atmospheric effect. Remote Sens Environ. 13:353–361.
Pu R, Gong P, Michishita R, Sasagawa T. 2006. Assessment of multi-resolution and multi-sensor data for urban surface
temperature retrieval. Remote Sens Environ. 104:211–225.
Qian YG, Li Z-L, Nerry F. 2013. Evaluation of land surface temperature and emissivities retrieved from MSG/SEVIRI
data with MODIS land surface temperature and emissivity products. Int J Remote Sens. 34:3140–3152.
Ramanathan V, Crutzen P, Kiehl J, Rosenfeld D. 2001. Aerosols, climate, and the hydrological cycle. Science. 294:2119–
2124.
Rozenstein O, Qin Z, Derimian Y, Karnieli A. 2014. Derivation of land surface temperature for Landsat-8 TIRS using
a split window algorithm. Sensors. 14:5768–5780.
Salisbury W, D’Aria DM. 1992. Emissivity of terrestrial materials in the 8–14 μm atmospheric window. Remote Sens
Environ. 42:83–106.
 Geocarto International   469

Sandau R, Brieß K, D’Errico M. 2010. Small satellites for global coverage: Potential and limits. ISPRS J Photogrammetry
Remote Sens. 65:492–504.
Schmugge TJ, Hook SJ, Coll C. 1998. Recovering Surface temperature and emissivity from thermal infrared multispectral
data. Remote Sens Environ. 65:121–131.
Schneider P, Hook SJ, Radocinski RG, Corlett GK, Hulley GC, Schladow SG, Steissberg TE. 2009. Satellite observations
indicate rapid warming trend for lakes in California and Nevada. Geophys Res Lett. 36:1–6.
Schott J, Gerace A, Brown S, Gartley M, Montanaro M, Reuter DC. 2012. Simulation of image performance characteristics
of the Landsat Data Continuity Mission (LDCM) Thermal Infrared Sensor (TIRS). Remote Sens. 4:2477–2491.
Shao Q, Sun C, Liu J, He J, Kuang W, Tao F. 2011. Impact of urban expansion on meteorological observation data and
overestimation to regional air temperature in China. J Geogr Sci. 21:994–1006.
Snyder WC, Wan Z, Zhang Y, Feng Y-Z. 1998. Classification-based emissivity for land surface temperature measurement
from space. Int J Remote Sens. 19:2753–2774.
Sobrino JA, Jiménez-Muñoz JC. 2014. Minimum configuration of thermal infrared bands for land surface temperature
and emissivity estimation in the context of potential future missions. Remote Sens Environ. 148:158–167.
Sobrino JA, Raissouni N. 2000. Toward remote sensing methods for land cover dynamic monitoring: application to
morocco. Int J Remote Sens. 21:353–366.
Sobrino JA, Romaguera M. 2004. Land surface temperature retrieval from MSG1-SEVIRI data. Remote Sens Environ.
92:247–254.
Sobrino JA, Raissouni N, Li Z-L. 2001. A comparative study of land surface emissivity retrieval from NOAA data.
Remote Sens Environ. 75:256–266.
Sobrino JA, Kharraz JE, Li ZL. 2003. Surface temperature and water vapour retrieval from MODIS data. Int J Remote
Sens. 24:5161–5182.
Sobrino JA, Jiménez-Muñoz JC, Paolini L. 2004. Land surface temperature retrieval from LANDSAT TM 5. Remote
Sens Environ. 90:434–440.
Sobrino JA, Jiménez-Muñoz JC, Sòria G, Romaguera M, Guanter L, Moreno J, Plaza A, Martínez P. 2008. Land surface
emissivity retrieval from different VNIR and TIR sensors. IEEE Trans Geosci Remote Sens. 46:316–327.
Sobrino JA, Oltra-Carrió R, Sòria G, Bianchi R, Paganini M. 2012. Impact of spatial resolution and satellite overpass
time on evaluation of the surface urban heat island effects. Remote Sens Environ. 117:50–56.
Stathopoulou M, Cartalis C. 2007. Daytime Urban Heat Islands from Landsat ETM+ and corine land cover data: an
application to major cities in Greece. Sola Energy. 81:358–368.
Stathopoulou M, Synnefa A, Caralis C, Sanamouris M, Karless T, Akbari H. 2009. A surface heat island study of Athens
using high-resolution satellite imagery and measurements of the optical and thermal properties of commonly used
building and paving materials. Int J Sustain Energy. 28:59–76.
Streutker D. 2003. Satellite-measured growth of the urban heat island of Houston, Texas. Remote Sens Environ. 85:282–
289.
Susskind J, Rosenfield J, Reuter D, Chahine MT. 1984. Remote sensing of weather and climate parameters from HIRS2/
MSU on TIROS-N. J Geophys Res. 89:4677–4697.
Tan KC, Lim HS, MatJafri MZ, Abdullah K. 2012. A comparison of radiometric correction techniques in the evaluation
of the relationship between LST and NDVI in Landsat imagery. Environ Monit Assess. 184:3813–3829.
Tang H, Li Z-L. 2014. Introduction. In: H Tang, Z-L Li, editors. Quantitative Remote Sensing in Thermal Infrared.
Berlin/Heidelberg: Springer; p. 1–4.
Tang B-H, Shao K, Li Z-L, Wu H, Nerry F, Zhou G. 2015. Estimation and Validation of Land Surface Temperatures from
Chinese Second-Generation Polar-Orbit FY-3A VIRR Data. Remote Sens. 7:3250–3273.
Tomlinson CJ, Chapman L, Thornes JE, Baker C. 2011. Remote sensing land surface temperature for meteorology and
climatology: a review. Meteorol Appl. 18:296–306.
Trigo IF, Monteiro IT, Olesen F, Kabsch E. 2008. An assessment of remotely sensed land surface temperature. J Geophys
Res. 113. Doi:10.1029/2008JD010035
United Nations. 2010. World urbanization prospects: The 2009 revision population database. http://esa.un.org/unpd/
wup/index.htm.
US Environmental Protection Agency. 2012. Heat Island Impacts. Available from: http://www.epa.gov/hiri/impacts/
index.htm.
Valor E, Caselles V. 1996. Mapping land surface emissivity from NDVI: application to European, African, and South
American areas. Remote Sens Environ. 57:167–184.
Van de Griend AA, owe M. 1993. On the relationship between thermal emissivity and the normalized difference
vegetation index for natural surfaces. Int J Remote Sens. 14:1119–1131.
Vlassova L, Perez-Cabello F, Nieto H, Martín P, Riaño D, de la Riva Juan. 2014. Assessment of Methods for Land Surface
Temperature Retrieval from Landsat-5 TM Images Applicable to Multiscale Tree-Grass Ecosystem Modeling. Remote
Sens. 6:4345–4368.
Voogt JA, Oke TR. 2003. Thermal remote sensing of urban climates. Remote Sens Environ. 86:370–384.
Wan Z. 2008. New refinements and validation of the MODIS Land-surface temperature/emissivity products. Remote
Sens Environ. 112:59–74.
470    A. A. Mohamed et al.

Wan Z, Li Z-L. 2008. Radiance-based validation of the V5 MODIS land-surface temperature product. Int J Remote
Sens. 29:5373–5395.
Wan Z, Zhang Y, Zhang Q, Li Z-L. 2002. Validation of the land-surface temperature products retrieved from Terra
Moderate Resolution Imaging Spectroradiometer data. Remote Sens Environ. 83:163–180.
Wan Z, Wang P, Li X. 2004. Using MODIS land surface temperature and normalized difference vegetation index products
for monitoring drought in the Southern Great Plains, USA. Int J Remote Sens. 25:61–72.
Weng Q. 2001. A remote sensing-GIS evaluation of urban expansion and its impact on surface temperature in the
Zhujiang Delta. China Int J Remote Sens. 22:1999–2014.
Weng Q. 2003. Fractal analysis of satellite-detected Urban Heat Island effect. Photogramm Eng Remote Sens. 69:555–566.
Weng Q. 2009. Thermal infrared remote sensing for urban climate and environmental studies: Methods, applications,
and trends. ISPRS J Photogramm Remote Sens. 64:335–344.
Xian G, Crane M. 2006. An analysis of urban thermal characteristics and associated land cover in Tampa Bay and Las
Vegas using Landsat satellite data. Remote Sens Environ. 104:147–156.
Xiaoma L, Zhou W, Ouyang Z. 2013. Relationship between land surface temperature and spatial pattern of greenspace:
What are the effects of spatial resolution? Landscape Urban Plan. 114:1–8.
Xu HQ, Ding F, Wen XL. 2009. Urban expansion and Heat Island dynamics in the Quanzhou Region, China. IEEE J Sel
Top Appl Earth Observ Remote Sens. 2:74–79.
Yamaguchi Y, Kahle AB, Tsu H, Kawakami T, Pniel M. 1998. Overview of Advanced Spaceborne Thermal Emission and
Reflection Radiometer (ASTER). IEEE Trans Geosci Remote Sens. 36:1062–1071.
Yu Y, Privette JL, Pinheiro AC. 2008. Evaluation of split-window land surface temperature algorithms for generating
climate data records. IEEE Trans Geosci Remote Sens. 46:179–192.
Yuan F, Bauer ME. 2007. Comparison of impervious surface area and normalized difference vegetation index as indicators
of surface urban heat island effects in Landsat imagery. Remote Sens Environ. 106:375–386.
Zhang YS, Odeh IOA, Han CF. 2009. Bi-temporal characterization of land surface temperature in relation to impervious
surface area NDVI and NDBI, using a sub-pixel image analysis. Int J Appl Earth Obs Geoinf. 11:256–264.
Zhang H, Qi Z-F, Ye X-Y, Cai Y-B, Ma W-C, Chen M-N. 2013. Analysis of land use/land cover change, population shift,
and their effects on spatiotemporal patterns of urban heat islands in metropolitan Shanghai, China. Appl Geogr.
44:121–133.
Zhao H, Chen X, Zhang J, Yin Z. 2012. Land-use/-cover change spatial patterns and their impacts on sediment charge
in the Longchuan River catchment, south-western China. Int J Remote Sens. 33:4527–4552.