Content uploaded by Aysar Jameel Abdalkadhum
Author content
All content in this area was uploaded by Aysar Jameel Abdalkadhum on Feb 12, 2019
Content may be subject to copyright.
International Journal of Engineering & Technology
608
Copyright © 2018 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unre-
stricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
International Journal of Engineering & Technology, 7 (4.20) (2018) 608-612
International Journal of Engineering & Technology
Website: www.sciencepubco.com/index.php/IJET
Research paper
Land Surface Temperature Retrieval from LANDSAT-8
Thermal Infrared Sensor Data and Validation with Infrared
Thermometer Camera
Muhammad Mejbel Salih1, Oday Zakariya Jasim1, Khalid I. Hassoon2, Aysar Jameel Abdalkadhum3*
1 Civil Engineering Department, University of Technology
2 Ministry of Science and Technology
3College of Water Resources Engineering, Al-Qasim Qreen University
*Corresponding author E-mail: Aysar@wrec.uoqasim.edu.iq:
Abstract
This paper illustrates a proposed method for the retrieval of land surface temperature (LST) from the two thermal bands of the LANDSAT-
8 data. LANDSAT-8, the latest satellite from Landsat series, launched on 11 February 2013, using LANDSAT-8 Operational Line Imager
and Thermal Infrared Sensor (OLI & TIRS) satellite data. LANDSAT-8 medium spatial resolution multispectral imagery presents particular
interest in extracting land cover, because of the fine spectral resolution, the radiometric quantization of 12 bits. In this search a trial has
been made to estimate LST over Al-Hashimiya district, south of Babylon province, middle of Iraq. Two dates images acquired on 2nd &18th
of March 2018 to retrieve LST and compare them with ground truth data from infrared thermometer camera (all the measurements contacted
with target by using type-k thermocouple) at the same time of images capture. The results showed that the rivers had a higher LST which
is different to the other land cover types, of less than 3.47 C ◦, and the LST different for vegetation and residential area were less than 0.4
C ◦ with correlation coefficient of the two bands 10 and 11 Rbnad10= 0.70, Rband11 = 0.89 respectively, for the imaged acquired on the 2nd of
march 2018 and Rband10= 0.70 and Rband11 = 0.72 on the 18th of march 2018. These results confirm that the proposed approach is effective
for the retrieval of LST from the LANDSAT-8 Thermal bands, and the IR thermometer camera data which is an effective way to validate
and improve the performance of LST retrieval. Generally the results show that the closer measurement taken from the scene center time, a
better quality to classify the land cover. The purpose of this study is to assess the use of LANDSAT-8 data to specify temperature differ-
ences in land cover and compare the relationship between land surface temperature and land cover types.
Keywords: land cover type; Land surface temperature (LST); LANDSAT-8 thermal bands; IR thermometer camera (thermocouple).
1. Introduction
Land surface temperature (LST) play an important role in research
on agricultural analyses, effects of urban heat island, and environ-
mental monitoring [1]. However, significant advancement has been
made in these applications with a series of remote sensing satellites
being launched [1-4]. LST divergences in space and time, measured
by satellite remote sensing data, were used for the assessment of a
lot of geophysical variables, such as soil moisture, vegetation water
stress, evapotranspiration, and thermal inertia [5-7]. With a view to
accurately retrieve LST, methods have been continuously devel-
oped, they can be roughly gather into three groups: the single-chan-
nel algorithm [8], multi-channel algorithm [9-15], and multi-angle
algorithm [16]. In general, using multi-channel algorithm is the
split-window algorithm that takes the interest of atmospheric water
vapor absorption difference between two channels centered at 11.0
µm and 12 µm to remove the effectiveness of atmosphere [9]. In
recent decades, a series of sensors which have been sent into space,
such as the Moderate-resolution Imaging Spectroradiometer
(MODIS) aboard Terra and Aqua, Advanced Very High Resolution
Radiometer (AVHRR) on series of National Oceanic and Atmos-
pheric Administration (NOAA) satellite, have provided public
range global thermal data twice daily, using two long wave infrared
(LWIR) bands [17]. Thematic Mapper (TM) and Enhanced The-
matic Mapper Plus (ETM+) aboard previous LANDSAT-5 and
LANDSAT-7 satellites provide thermal data using only one
(LWIR) band, with higher spatial resolution within 16-days tem-
poral resolution. LANDSAT-8 was successfully launched on the
11th of February 2013 into space with two instrument on board, the
Operational Land Imager (OLI) and the Thermal Infrared Sensor
(TIRS) [18]. The OLI instrument with nine spectral bands in the
visible (VIS), near infrared (NIR) (Table 1), and the shortwave in-
frared (SWIR) spectral regions, while the TIRS instrument with two
thermal infrared spectral bands in the LWIR respectively centered
at 10.9 µm and 12 µm (Table 2). The relative spectral response of
both OLI and TIRS channels are illustrated in Fig 1. Further, an-
other advantage is that the OLI, and TIRS instruments observed
Earth’s surface with resolution from 15 meters to 100 meters. Ac-
cording to the technical specification, the LANDSAT-8 are very
suitable for retriever LST. Therefore, the objective of this paper is
to retrieve LST and compare them with trained and validated
ground truth data from infrared thermometer camera (IR thermom-
eter) at the same time the images were acquired.
International Journal of Engineering & Technology
608
Copyright © 2018 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unre-
stricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Fig. 1: Spectral response of OLI and TIRS channels [19]
Table 1: Technical specification of operational image (OLI) on LAND-
SAT-8[18]
Band Name
Spectral
range (µm)
Spatial resolution
(m)
Band 1 Visible Ultra Blue
(Coastal)
0.435-0.451
Band 2 Visible (Blue)
0.452-0.512
Band 3 Visible (Green)
0.533-0.590
Band 4 (Red)
0.636-0.673
Band 5 Near-Infrared (NIR)
0.851-0.879
Band 6 Shortwave Infrared
(SWIR 1)
1.566-1.651
Band 7 Shortwave Infrared
(SWIR 2)
2.107-2.294
Band 8 Panchromatic (Pan)
0.503-0.676
Band 9 (Cirrus)
1.363-1.384
Table 2: Technical specification of thermal infrared sensor (TIRS) on
LANDSAT-8[18].
Band
Name
Central wave-
length (µm)
Spectral range
(µm)
Spatial resolu-
tion (m)
Band 10
TIRS 1
(30)*
Band 11
TIRS 2
(30)*
* TIRS bands are acquired at 100 meter resolution, but are
resampled to 30 meter in delivered data product.
2. Study Area and Dataset
2.1 Study area
Al-Hashimiya district is a part of Babil province, Iraq, which con-
sist of four cities. It is located along the Shatt Al-Hashimiya, while
the city of Al-Qassim is located to the south, and Al-Madhatiya and
Al-Shomali cities are located to the east. Al-Hashimiya district oc-
cupies an area of 1836.74 square kilometers (183674.166 hectares)
and it is located to the south of Babil province approximately be-
tween 44◦ 30‵ - 45◦ 15‵ E longitude and 32◦ 00‵ -32◦ 30‵ N lati-
tudes. The population of the district is about 600, 000 people. It is
characterized by its agricultural and rural areas and the moderate
weather ranges between the maximum and minimum temperature
(45◦- 4◦) C◦. The Date palm trees cover a wide are, up to more than
2 million trees. It is characterized by the cultivation of field crops,
especially wheat, barley and corn, where the agricultural area is
covered more than 123000 hectares, according to the statistics of
the province of Babylon in 2017 .The Al-Hashimiya district is con-
sidered the geographical center of Iraq, where the distance from east
to west of Iraq is equal to the distance from north to south.
Fig. 2: Study area Al-Hashimiya district, Iraq
2.2 Satellite data
The multispectral remote sensing images of Al-Hashimiya city of
two dates were collected from USGS website. LANDSAT-8 satel-
lite images revisit earth once in 16 days. Bands descriptions of
LANDSAT-8 are as given in table 1 and 2 (https://landsat-
look.usgs.gov). Satellite data over Al-Hashimiya city of the 2nd and
the 18th of March 2018 (day time, level-1TP product, path/raw
168/38) have been used in this study. Both were downloaded from
USGS with image quality 9. The study area consist vegetation
cover, bare soil, water and residential area. All the data were re-
projected to a Universal Transverse Mercator (UTM) coordinate
system, datum WGS84, zone 38.
2.3 Infrared Thermometer Camera (IR Thermometer)
Infrared thermometer measure the surface temperature of an opaque
object [20]. The thermometer’s optics sense emitted, reflected, and
transmitted energy, which is gathered and centered onto a detector.
The unit’s electronics translate the information into a temperature
reading which the displays on screen (see Figure 3). The Infrared
Thermometer Camera as shown in figure (4-a) is for non-contact
temperature measurement. This thermometer determines object’s
surface temperature by measuring the amount of infrared energy ra-
diated by the object’s surface [20]. The thermometer also support
contact temperature measurement via K-type thermocouple as
shown in figure (4-b). In this paper all the land surface temperature
measurements (ground-truth) taken by using type-K (thermocou-
ple). The main Specifications of the IR temperature camera is illus-
trated in Table 3. Field measurements were observed by using In-
frared Thermometer. The observations were measured with direct
connection with land surface at the same time of the acquisition of
images. The meteorological data recorded the minimum and maxi-
mum temperature degree ranges 18 to 25 C◦ and 9 to 24 C◦ for the
days 02/03/2018 and 18/03/2018 respectively.
International Journal of Engineering & Technology
609
Fig.3: How the Thermometer Works [20].
Fig. 4: a-Infrared thermometer camera, b- K-type thermocouple [20].
Table 3: IR Thermometer measurement non-contact (infrared) tempera-
ture measurements [20]
Temperature range
-50 to 1000 C◦
Accuracy
1 C◦
Resolution
640*480 pixels
Display resolution
0.1 C◦ < 1000 C◦
Response time
150 ms
Spectral response
8~14 µm
Display resolution
Digitally corrected from 0.1 to 1.0
Type-k thermocouple measurements with connection with land surface
Temperature range
-50 to 1370 C◦
Accuracy
0.5 C◦
Display resolution
0.1 C◦ < 1000 C◦ < 1000 C◦
Display resolution
Adjustable emissivity
3. Methodology
The algorithm of the proposed work to retrieve LST is shown in the
Fig. 5. This approach can only use to process LANDSAT-8 data. In
this paper, two bands 10 and 11 are used to estimate brightness tem-
perature and the red band (4), near infrared band (5) are used to
compute the Normalized Difference vegetation Index (NDVI).
Fig. 5: Flow diagram for LST retrieval
The steps included in the proposed work are detailed in the follow-
ing literature.
1- The LANDSAT-8 data products were geometrically corrected data
set. The metadata of the satellite images is shown in the table 4.The
first step of the proposed work is the satellite-based digital number
(DN) is converted to at-sensor spectral radiance (Lλ) using the fol-
lowing equation [21]:
(1)
Where ML= Band specific multiplicative rescaling factor from the
metadata, Qcal= Quantized and calibrated standard product DN
value of pixel, and AL= Band-specific additive rescaling factor from
the metadata.
Table 4: Metadata oh the LANDSAT-8 satellite image.
Band
Varia-
ble
Description
Value
10
Thermal band
K1
Thermal constant
774.8853
K2
1321.0789
ML
RADIANCE_MULT_BAND
0.000 33420
Qcal
DN value of pixel
AL
RADIANCE_ADD_BAND
0.1
11
Thermal band
K1
Thermal constant
480.8883
K2
1201.1442
ML
RADIANCE_MULT_BAND
0.000 33420
Qcal
DN value of pixel
AL
RADIANCE_ADD_BAND
0.1
2- The thermal infrared bands should be converted to brightness tem-
perature (BT) using metadata and the following equation [21].
(2)
Where K1 and K2 are the thermal infrared (TIRS) bands 10 and
11 which can be find in the metadata file linked with the satellite
image. To have the results in Celsius (C◦) it is needed to refine by
adding absolute zero which is approximately equal to – 273.15.
3- (NDVI) is necessary to identify land cover types of the area study.
The NDVI ranges between -1 to +1. NDVI is computed on per-pixel
basis as the normalized difference between the red band (0.636-
610
International Journal of Engineering & Technology
0.673µm) and near infrared band (0.851-0.879µm) of images using
following equation
(3)
Where NIR is the near infrared band value of pixel and RED is the
red band of same pixel. The NDVI is essential to calculate propor-
tional vegetation (Pv) and emissivity (ɛ).
4- From NDVI values obtained in the previous step calculate pro-
portional vegetation (Pv). This proportional vegetation gives the es-
timation of area under each land cover type. The vegetation and
bare soil proportions are acquired from the NDVI of pure pixel.
(4)
Computation of land surface emissivity (LSE) is required to esti-
mation LST. Land surface emissivity that is describes the radiative
absorption ability of a surface in loge wave radiation spectrum [22].
LSE is largely dependent on the target surface top layer such as type
of soil, surface roughness, and nature of vegetation cover [23]. To
obtained LSE, the following equation using [23]:
(5)
Where ɛ = Land Surface Emissivity
ɛSλ= emissivity of soil
ɛVλ = emissivity of vegetation
PV = proportion of vegetation
Cλ = surface roughness taken as a constant value of 0.009
The emissivity of water bodies is very settled in comparison with
land surfaces. Since the emissivity depends on the wavelength, the
NDVI can be used to estimate the emissivity of different land sur-
faces in the 10-12 μm range [22].
(6)
6- The final step is to calculate LST using brightness temperature (BT) of
two bands 10, 11 and LSE that is derived from Pv and NDVI [23].LST can
be retrieved using the following equation:
(7)
Where, BT is at- sensor BT in Celsius (C◦), λ is the mean wavelength of
band 10 or 11, ɛλ is the emissivity calculated from equation 5 and ρ is
which is = 1.438 x 10-2 mk in which, σ is the Boltzmann constant =
1.38 x 10-23 J/K, h is Plank constant = 6.626 x 10-34 and c is the velocity
of light = 3 x 108 m/s.
4. Results
Two LANDSAT-8 images were acquired. One image was obtained
on March the 2nd, 2013; the other image was obtained on March
the 18th, 2013 and both were used in this paper. The land cover was
classified into four types on cloud-free LANDSAT-8 image which
was acquired on the 02/03/2018 and ten types on cloud-free LAND-
SAT-8 image which was acquired on 18/03/2018. The two images
were taken at the same time the ground-truth field was measured
with a direct connection with land surface, as shown consecutively
on tables 5 and 6. Were taken at the same time the ground-truth
field was measured with a direct connection with land surface, as
shown consecutively on tables 5 and 6. As shown in Table 5 the
difference between the IR thermometer camera measurements
(type-k) and LST retrieved ranges between 0.4 C◦ and 1.77 C◦ with
the correlation coefficient of the two bands 10 and 11 Rbnad10= 0.70,
Rband11 = 0.89 respectively.
The results shown are satisfactory and this is due to the less meas-
urement undertaken and the scene center time recorded which was
in 10:33:42 local time according to Metadata. On the other hand
Table 6 shows an extraordinary results between the IR thermometer
camera measurements (type-k) and the LST retrieved measure-
ments Rband10 =0.70 and Rband11 = 0.72, the errors ranges between -
0.02 C◦ and 3.47 C◦. The results reveal that there is a high error in
water (3.47 C◦) and less errors in vegetation which ranges between
-0.02 C◦ – 1 C◦
International Journal of Engineering & Technology
611
Fig. 6: LST of Al-Hashimiya (a) and (b) LST maps of the two bands 10 and 11 respectively image acquired on 2-3-2018, (c) and (d) LST maps of the two
bands 10 and 11 respectively image acquired on 18-3-2018.
Table. 5: Retrieved LST and IR thermometer camera for TIRS bands 10 and 11 for 02/03/2018
Time of
measured
Am
Latitude
Longitude
Land cover
types
Land
cover
class
02/03/2018 band 10
02/03/2018 band 11
LST Re-
trieved
(C◦)
Type-k ther-
mometer cam-
era (C◦)
Error
(C◦)
LST Re-
trieved
(C◦)
Type-ther-
mometer cam-
era (C◦)
Error
(C◦)
9:45
467483
3582422
Water
River
18.77
17.00
1.77
18.64
17.0
1.64
10:00
467554
3582366
Soil
Clay soil
19.86
18.20
1.66
18.60
18.20
0.40
10:45
468906
3581914
Vegetation
Alfalfa
19.99
21.40
-1.41
19.75
21.40
-1.65
10:30
468886
3581957
Building
Asphalt
19.86
19.2
0.66
18.68
19.20
-0.52
The errors in vegetation recorded were due to a close scene center
time (10:33:34) acquired compared to the measurements of the
buildings, soil and water. However the least errors shown in Table
5 were recorded in soil and buildings that ranges between -0.52 C◦
– 0.4 C◦. Generally the results show that the closer measurement
taken from the scene center time, a better quality to classify the land
cover
Table 6: Retrieved LST and IR thermometer camera for TIRS bands 10 and 11 for18/03/2018
Time of
measured
Am
Latitude
Longitude
Land cover
types
Land
cover
class
18/03/2018 band 10
18/03/2018 band 11
LST Re-
trieved
(C◦)
Type-k ther-
mometer
camera (C◦)
Error
(C◦)
LST Re-
trieved
(C◦)
Type-k ther-
mometer
camera (C◦)
Error
(C◦)
10:15
467659
3582611
water
River
24.27
20.80
3.47
23.19
20.80
2.39
9:45
470449
3574270
soil
Clay
soil
25.34
23.10
2.24
24.13
23.10
1.03
10:20
468194
3582870
Soil
Sandy
soil
23.50
24.7
-1.2
22.58
24.70
-2.12
10:00
468300
3579318
Soil
Bare
soil
28.48
28.6
-0.12
26.94
28.6
-1.66
10:20
467251
3573651
Vegetation
Barley
24.86
23.00
1.86
24.02
23.00
1.02
10:33
469203
3582082
Vegetation
Date
Palms
24.97
24.3
0.67
24.03
24.3
-0.27
10:30
468897
3581936
Vegetation
Alfalfa
24.80
23.9
0.90
23.88
23.9
-0.02
10:05
468118
3579330
Building
Con-
crete
28.31
27.6
0.71
26.82-0.78
27.6
-0.78
11:00
470819
3573715
Building
Brick
25.55
25.10
0.45
24.34
25.10
-0.76
9:51
471153
3574666
Building
Asphalt
28.21
25.6
2.61
26.74
25.6
1.14
5. Conclusions
The algorithm created in the ArcGis, to estimate the LST for the
selected datasets with direct contact with the study area. the pro-
posed algorithm was created using the brightness temperature of
thermal infrared of two bands 10 and 11 and emissivity of different
land cover types, obtained from visible and near infrared bands of
LANDSAT-8.the estimated LSTs were verified using the infrared
thermometer camera with contact directly with the land surface. By
looking at the tables above, and by comparing the measurement re-
sults of the two images, the results obtained show that:
• The first image was gives high correlation coefficient of the two
bands 10 and 11 Rbnad10= 0.70, Rband11 = 0.89 respectively, for
the imaged acquired on the 2nd of march 2018 due to lack of meas-
urements taken, in comparison with the second image whose corre-
lation coefficient Rband10=0.70 and Rband11 = 0.72. The results
showed a clear indication of the difference in measured temperature
by IR thermometer camera and LST which showed the closer scene
center time, the more accurate land cover classification.
• The errors in band 11 is always less than the errors occurring in
band 10 in both images.
612
International Journal of Engineering & Technology
• The difference showed in the LST for the same class is due to the
canopies (buildings or trees) that make a shadow on these classes.
References
[1] Lu D, Weng Q.,” Spectral mixture analysis of ASTER images for
examining the relationship between urban thermal features and bio-
physical descriptors in Indianapolis”, Remote Sensing of Environ-
ment, Vol.104, No.2, (2006), pp.157-167.
[2] Weng Q, Lu D.,” A sub-pixel analysis of urbanization effect on land
surface temperature and its interplay with impervious surface and
vegetation coverage in Indianapolis”, International Journal of Ap-
plied Earth Observation and Geoinformation,Vol.10, No.1, (2008),
pp.68-83.
[3] Patel N, Parida B, Venus V, Saha S, Dadhwal ,” Analysis of agricul-
tural drought using vegetation temperature condition index (VTCI)
from Terra/MODIS satellite data”, Environmental Monitoring and
Assessment, Vol.184, NO.12, (2012), pp. 7153–7163.
[4] Krishna A P, Sharma A. s.l.,” Snow cover and land surface temper-
ature assessment of Gangotri basin in the Indian Himalayan Region
(IHR) using MODIS satellite data for climate change inferences”,
International Society for Optics and Photonics, SPIE Remote Sens-
ing, (2013), pp. 889314-889319.
[5] Karnieli, A., et al.,” Use of NDVI and Land Surface Temperature for
Drought Assessment”, American Meteorological Society, Vol. 23,
(2010), pp. 618-633.
[6] Kustas, W. and Anderson, M.,” Advances in Thermal Infrared Re-
mote Sensing for Land Surface Modeling. Kustas”, Agricultural and
Forest Meteorology, Vol.149, No.12, (2009), pp. 2071-2081.
[7] Agam, N., et al.,” Utility of Thermal Image Sharpening for Monitor-
ing Field-Scale Evapotranspiration Over Rainfed and Irrigated Agri-
cultural Regions”, Geophysical Research Letters, Vol.35, No.2,
(2008), pp. L02402.
[8] C. Ottlé and D. Vidal-Madjar,” Estimation of land surface tempera-
ture with NOAA9 data “, Remote Sensing of Environment, Vol.40,
No.1, (21992), pp.27-41.
[9] L. M. Mcmillin,” Estimation of sea surface temperatures from two
infrared window measurements with different absorption”, J. Ge-
ophys.Res., Vol.80, No.36, (1975), pp. 5113-5117.
[10] Li, F. Becker and Z.-L.,” Towards a local split window method over
land surfaces”, International Journal of Remote Sensing, VOL.11,
NO.3, (1990), pp.369-393.
[11] Z. Wan and J. Dozier,” A generalized split-window algorithm for re-
trieving land-surface temperature from space”, IEEE Transactions on
Geoscience and Remote Sensing, Vol.34, No.4, (1996), pp. 892 –
905.
[12] Z. Qin, G. Dall'Olmo, A. Karnieli, and P. Berliner,” Derivation of
split window algorithm and its sensitivity analysis for retrieving land
surface temperature from NOAA-advanced very high resolution ra-
diometer data”, J. Geophys.Res. Atmospheres, Vol.106, NO.D19,
(2001), pp. 22655-22670.
[13] Li, G.-M. Jiang and Z.-L,” Split-window algorithm for land surface
temperature estimation from MSG1-SEVIRI data”, International
Journal of Remote Sensing, Vol.29, No.20, (2008), pp. 6067-6074.
[14] G.-M. Jiang, W. Zhou, and R. Liu.,” Development of split-window
algorithm for land surface temperature estimation from the
VIRR/FY-3A measurements”, IEEE Geoscience and Remote Sens-
ing Letters, Vol.10, No.4, (2013), pp. 952-956.
[15] G.-M. Jiang and R. Liu,” Retrieval of sea and land surface tempera-
ture from SVISSR/FY-2C/D/E measurements”, IEEE Transactions
on Geoscience and Remote Sensing, Vol.52, No.10, (2014), pp. 6132
- 6140.
[16] Sobrino, G. Sòria and J. A.,” ENVISAT/AATSR derived land sur-
face temperature over a heterogeneous region”, Remote Sensing of
Environment, Vol.111, (2007), pp. 409-422.
[17] Rozenstein, Zhihao Qin, Yevgeny Derimian and Arnon Karnieli,”
Derivation of Land Surface Temperature for Landsat-8 TIRS Using
a Split Window Algorithm”, Sensors Open Access Journal, Vol.14,
No.4, (2014), pp. 5768-5780.
[18] Internet-Website, (https://landsat.usgs.gov).
[19] Shanshan LI and Geng-Ming Jiang,” Land Surface Temperature Re-
trieval from Landsat-8 Data with the Generalized Split-Window Al-
gorithm”, IEEE Access, Vol.6, (2018), pp. 18149 – 18162.
[20] IR Video Thermometer User Manual Model DT-9862.
[21] Walawender, J. p., Szymanowski, M., Hajto, M. J., Bokwa A.,” Land
syrface temperature patterns in the urban agglomeration of Krakow
(Poland) mderived from Landsat-7/ETM+ data”, Pure and Applied
Geophysics, Vol.171, NO.6, (2014), pp. 913–940.
[22] Jose A. Sobrino, Juan C. Jimenez-Munoz, Guillem Sòria, Mireia Ro-
maguera, Luis Guanter, Antonio Plaza, & Pablo Martínez,” Land
Surface Emissivity Retrieval from Different VNIR and TIR Sensors”,
IEEE Transactions on Geoscience and Remote Sensing, Vol.64,
(2008), pp. 316 – 327.
[23] M. Stathopoulou, & C. Cartalis,” Daytime urban heat islands from
Landsat ETM+ and Corine land cover data: An application to major
cities in Greece” , Solar Energy, Vol.81, No.3, (2007), pp. 358-368.
