See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/345976049

Land Surface Temperature Estimation of Mandya district using LANDSAT-8

data

Article · May 2020

CITATIONS READS

7 649

4 authors, including:

Aryalekshmi B N Rajashekhar Biradar

HKBK College of Engineering REVA University, Bangalore, Inda
14 PUBLICATIONS 65 CITATIONS 147 PUBLICATIONS 2,055 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Aryalekshmi B N on 17 November 2020.

The user has requested enhancement of the downloaded file.

 Journal of Applied Science and Engineering, Vol. 23, No 4, Page 583-591 583

Land Surface Temperature Estimation of Mandya district using

LANDSAT-8 data
Aryalekshmi B N1,* , Mohammed Ahamed J2 , Dr. R C Biradar1 , and Dr. Chandrasekar K3
1 School of ECE, REVA University, Bangalore, India
2 RRSC – South, NRSC, ISRO, Bengaluru, India
3 RRSC-Hyderabad, NRSC, ISRO, India
* Corresponding author. E-mail: aryalks88@gmail.com

Received: Dec. 10, 2019, Accepted: May. 29, 2020

This paper focuses on the estimation of LST over Mandya district (120o 31’N, 760o 53’E), which is a part of
the Cauvery river basin in Karnataka, India, using LANDSAT-8 satellite data for 12 meteorological stations.
Landsat 8 data is used for experimentation during 21st Jan 2018 and 6th Feb 2018 for LST analysis of identified
area. More than half of the land area in the Mandya district is under agricultural use. Fluctuations in LST
affects the quality of agricultural production. As a result, proper estimation of LST is very much required for
the management of crop growth, development, and yield component. Spectral radiance from band 10 and
emissivity of thermal infrared bands was used as input for LST estimation. Surface emissivity was obtained
with the help of NDVI (Normalized Difference Vegetation Index), Leaf Area Index (LAI), and Albedo in the
Mandya region. ERDAS IMAGINE and ArcGIS software were used for modeling. The obtained data compared
with weather station data collected by KSNDMC (Karnataka State Natural Disaster Monitoring Centre). For this
analysis, we considered two sets of data taken on different dates during the said period. The Root Mean Square
Error (RMSE) and correlation coefficient are used as statistical criteria for evaluating the model efficiency. From
the comparison, it is found that RMSE and R for the first date are 0.94o C, 0.62, and for the second date, it is
1.26o C, 0.88, respectively. It is observed that LST and weather data do follow the same pattern for the selected
12 meteorological stations.

Keywords: Emissivity; Land surface temperature (LST); NDVI; LAI; LANDSAT-8; Mandya
http://dx.doi.org/10.6180/jase.202012_23(4).0002

1. Introduction any region increases, it completely changes the Land Use

Land Cover (LULC) of those cities, making it warmer than
Land Surface Temperature (LST) is a crucial factor for an- their surrounding areas. Urban Heat Island (UHI) is one of
alyzing many aspects of biology and climate, affecting or- the phenomena that occurs due to an increase in tempera-
ganisms and ecosystems from local to global scales. It ture. UHI leads to the replacement of the natural vegetation
measures the radiative temperature of the surface of the of that region. With the help of satellite-derived data, we
earth. It has many applications in the field of monitoring can analyze the effect of urbanization as well as UHI. LST
vegetation, crop yield, evapotranspiration (ET) [1], urban estimation helps in analyzing the UHI effect properly [4].
climate change [2–4], forest fire detection [5, 6], etc. [7, 8].
It also helps in estimating soil moisture [9, 10] in the area Modeling of the temporal and spatial patterns of LST
of interest. For the study of climate variation over many is required for understanding the environmental impacts
years, surface temperature analysis plays an important role. of LST. When the dense network of the meteorological sta-
When the atmosphere is warmer, the chances of evapora- tion is unavailable in a particular area, the values obtained
tion of water from the river, ocean, and land will be more from satellite imagery can be used as a parameter for an-
[11]. When the built-up areas that are the concrete zones of alyzing water resource management, agro-environmental
584 Aryalekshmi B N et al.

studies, etc. Remote sensing-based estimation of surface Landsat-8 image for Mandya district is not known as of the
temperature produces accurate data when compared with time of writing this paper.
the weather station, which is point data [12, 13]. Surface The accurate result of LST is mainly based on the type
temperature changes rapidly with the change in the geogra- of data, condition of the environment, and the type of algo-
phy of the area, such as due to change in vegetation cover, rithm used. The retrieval of LST from remotely sensed TIR
soil type, surface roughness, etc. Due to all these factors, data had attracted much attention in researchers. Radiative
the estimation of LST is an important parameter for the Transfer Equation (RTE) is used for linking LST to the TIR
analysis of environmental changes in the whole part of the data [21]. Since the radiances calculated by the onboard
world. satellite radiometers not only depend on the temperature
Many algorithms are proposed for the estimation of and emissivity but also on atmospheric effects, it is diffi-
LST using different RS data [14–17]. As already discussed, cult to obtain accurate data from TIR based LST estimation.
many satellite sources are available for getting environ- Therefore, both emissivity, as well as atmospheric correc-
mental information. Among this, LANDSAT-8, which is tion, need to be done for TIR based LST estimation. In
a part of the Landsat program, is most commonly used LANDSAT-8 level-2 data, as already discussed, the images
for LST estimation due to the free availability of satellite will be already atmospherically corrected. Based on the
images that helps in, accurate estimation of LST [16, 18]. studies proposed by researchers for LST estimation, this can
Landsat-8 level 2 data obtained from the United States Ge- be mainly classified into two categories. The first one with
ological Survey (USGS) earth explorer website was used known surface emissivity (SE) such as Single Channel (SC)
in this study. Since these images were already atmospher- algorithm, mono-window (MW) algorithm, split-window
ically corrected using ground data by USGS, no separate techniques [15, 22], multi-channel techniques [14]. The sec-
atmospheric correction was done. LANDSAT-8 has got ond category is with unknown SE as simultaneous retrieval
two sensors, the Operational Land Imager (OLI) and Ther- of LST and emissivity with known atmospheric informa-
mal Infrared Sensor (TIRS) [19]. Visible and Near-infrared tion, simultaneous retrieval with unknown atmospheric
bands data were provided by band 2-7 of LANDSAT-8; it is information, and stepwise retrieval method [23].
of 30m*30m pixel size. Longwave (thermal) radiation data Users usually face some problems to estimate LST from
is provided by band-10 and band-11, which are acquired at the above techniques due to its complicated steps. Due
100m resolution but are resampled at 30m in delivered data to this, some researchers came up with some tools which
product. From LANDSAT-8, only band-10 is currently used use RS and GIS software such as ERDAS IMAGINE [2],
in this study due to uncertainty in calibration reported for ArcGIS [24], ENVI, Python, etc. for the estimation of LST
band-11 by the USGS [20]. [17]. ERDAS IMAGINE has got the MODEL MAKER tool,
which helps in effective geospatial modeling, whereas Ar-
The reason for choosing Mandya district is that agricul-
cGIS helps in the proper analysis of the obtained data. In
tural sector is the principal patron of the Indian economy,
this paper, the ERDAS IMAGINE model maker tool and
and Mandya is one of the significant agriculturally flourish-
ArcGIS tool were used to develop geospatial modeling for
ing district of Karnataka . The study of LST with satellite
automated LST estimation using the Single channel (SC)
data over this region is limited when compared with the
method and LANDSAT-8 data. In the SC method out of
other part of the world. The Mandya district lies in the Cau-
two thermal bands (band 10 and 11), only band-10 will be
very river basin, and it shares its border with the districts
used for LST estimation. In addition to thermal bands, this
of Hassan, Mysore, Tumkur, Ramanagara, and Chamara-
method requires surface emissivity for the computation of
janagar. The major portion of the total land area of Mandya
LST. The surface emissivity was calculated using soil and
is under agricultural use. Some of the major crops of this
vegetative indices. The obtained LST data were compared
region are ragi, rice, sugarcane, groundnut, sesame, and
with 12 weather station point temperature data (obtained
pulses. One of the major commercial crops of Mandya is
from KSNDMC). The detailed step by step procedure is ex-
Sugarcane. Weather parameters such as temperature, rain-
plained further. The rest of the paper organized as follows:
fall, evapotranspiration (ET), etc. are important factors for
Section 2 describes the study area and data sets. The de-
agricultural production. Variation in weather parameters
tailed methodology is in section 3. In section 4, results and
results in a change in the quality of agricultural products.
discussion is given, and in section 5 conclusion is given.
Mainly LST variation will lead to changes in climatic con-
ditions, UHI, less crop yield, etc. As a result, a proper
2. Study Area and Datasets
study of the LST of Mandya is required for the agricultural
production growth of Karnataka. Estimation of LST from Study area
 Journal of Applied Science and Engineering, Vol. 23, No 4, Page 583-591 585

The current study(as shown in Figure 1) was based

on the Cauvery river basin [25] through Karnataka state,
mainly the Mandya region. Cauvery river is one of the
major rivers in India which originates at Talakaveri (12o
25’N, 75o 34’E) in Kodagu district, which is in the Western
Ghats at an altitude 1341 m. The river basin is estimated to
have an 8115 km2 area, which occupies nearly 2.5% of the
total geographical area of the country. Cauvery river basin
catchment lies in Karnataka (41.2%), Tamil Nadu (55.5%),
Kerala (3.3%).

Fig. 2. Location map of the Study Area

up area, vegetation, water, forest, grassland, bare soil, etc.

*TIRS bands are acquired at 100m resolution, but in deliv-

Table 1. Technical specification of Landsat-8 [20]

Spectral
Band Resolution (m)
wavelength (m)
Band-1 (Coastal) 0.43-0.450 30
Fig. 1. Location map of the Study Area Band-2 (Blue) 0.45-0.51 30
Band-3 (Green) 0.533-0.59 30
Weather data Band-4 (Red) 0.63-0.67 30
The weather data used in the present study was ob- Band-5 0.85-0.87 30
tained from KSNDMC (Karnataka State Natural Disaster (Infrared)
Monitoring Centre). The dataset consists of daily values of Band-6 1.56-1.65 30
maximum and minimum temperature (Tmax , Tmin ), which (Short wave )
was used for comparing with the obtained surface temper- infrared SWIR 1)
ature. Band-7 2.10-2.29 30
Satellite data (Shortwave )
The LANDSAT-8 level 2 satellite images of the Mandya Infrared SWIR2)
region (cloud-free data) were obtained from the USGS web- Band-8 0.50-0.67 15
site as shown in Figure 2. LANDSAT-8 band designation (Panchromatic)
is as in Table 1. Satellite images over the Mandya region Band-9 (Cirrus) 1.36-1.38 30
of 21st January and 6th February 2018 have been used for Band-10 (TIRS1) 10.6-11.1 100 (30)*
this study (Table 2). The area of interest consists of a built- Band-11 (TIRS2) 11.5-12.5 100 (30)*
586 Aryalekshmi B N et al.

ered data product, it is resampled to 30m. For NDVI > 0

e NB = 0.97 + 0.0033LAI for LAI ≤ 3 (4)

3. Methodology
The methodology used for the study is as shown in the e NB = 0.98 for LAI > 3 (5)
flowchart (Figure 3). LST estimation is done in the ERDAS For NDVI ≤ 0
Imagine model maker tool. LANDSAT-8 metadata used for
this study is as shown in Table 4. Water α < 0.47 NB = 0.99 (6)
Step 1: Albedo (α)Calculation
Snow α ≥ 0.47 NB = 0.99 (7)
Surface albedo is the ratio of radiation reflected from
the surface to the shortwave radiation incident at the sur- The value of water and snow are set to the same in this
face. For LANDSAT-8 shortwave spectrum (0.2 to 3.2µm) study. But based on the locally calibrated emissivity value
is represented by band 2-7. Albedo expression is given as and user preference, this can be modified.
in equation 1. Step 4: Spectral radiance (L, ) Calculation
7  As the satellite image used which is geometrically cor-
α = ∑ ρs,b · ωb

(1)
b =2
rected data set; from that the DN(digital number) of band
10 is first converted to spectral radiance (Lλ ) [29] using the
here ρs,b is the input satellite image reflectance of the
equation 8
area, as shown in Figure 2.
ωb is the weighting coefficient of each band as given in Lλ = M L ∗ Qcal + A L (8)
Table 3. Step 5: Land Surface Temperature (LST) Estimation
Step 2: Vegetation Indices (NDVI, LAI) Calculation The last step is to obtain the land surface temperature
Normalized Vegetation Indices (NDVI) indicates the (LST) [30, 33, 34] by using modified Plank is given in equa-
vegetation cover of the region of interest and is calculated tion 9.
using band-4 and band-5 of Landsat-8 images using the K2
LST =   (9)
ENB K1
equations 2 [7, 10, 18, 26]. In the electromagnetic spectrum, ln Lλ + 1
the near-infrared part reflects more energy by healthy veg-
Where e NB is the narrow band emissivity, and Rc is
etation. Its value ranges from -1 to +1. The value within
the corrected thermal radiance from the respective surface.
0.2 to +1 indicates green vegetation, and that which is less
K1 and K2 values are constant, as in table 4. Finally, the
than 0, means water, snow, and clouds.
satellite-derived land surface temperature is verified using
( Band5 − Band4) weather station data, as shown in Table 5.
NDV I = (2)
( Band5 + Band4)
NDVI obtained for the two dates are shown in Figure 4. 4. Results and Discussion
Leaf Area Index (LAI) is the ratio of the total area of all Two LANDSAT-8 images (as given in Table 2) of the
leaves on a plant to the ground area represented by the Mandya region were taken from the USGS website for the
plant. It ranges from 0 to 6 and is expressed by using the study. The obtained image was layer stacked and applied
empirical formula is given in equation 3 [27]. Along with mathematical models using the ERDAS IMAGINE tool.
NDVI, LAI for two dates is also shown in Figure 4. After getting the spectral radiance from the DN values,
vegetation indices (NDVI, LAI), and surface albedo and
LAI = 10NDV Is3.5 (3)
emissivity were calculated. Weather station data was col-
Step 3: Surface emissivity ( e0 , eNB ) Calculation lected from KSNDMC and was used for comparing with
For LST estimation, it is necessary to obtain surface the obtained surface temperature results for 12 represen-
emissivity ( e0 , eNB ) [28]. Emissivity [29] helps in analyzing tative points. Due to the variation in the resolution of
the amount of radiance emitted from the surface of the LANDSAT-8, there can be a difference in the air tempera-
earth. It has more influence on LST. Based on the soil ture of a weather station and obtained surface temperature.
type, vegetation cover nature, and surface roughness, the It can be due to sensor characteristics as well as weather
surface emissivity is varied [3, 30, 31]. e0 is the broadband conditions.
emissivity, and eNB , which is narrowband emissivity, is The surface temperature of the Mandya region was cal-
used for LST calculation. Surface emissivity is calculated culated by processing LANDSAT-8 and by using equation
using an empirical equation 4, 5 by [32] based on soil and 9 in the ERDAS IMAGINE model maker. Figures 5 & 6 dis-
vegetative indices. play the surface temperature pattern of Mandya district for
 Journal of Applied Science and Engineering, Vol. 23, No 4, Page 583-591 587

Table 2. Satellite image details

DOY Lat/Long Acquisition dates Path/Row Local time

21 120 31’N, 76o 53’E 21/1/2018 144/51 10.40
37 120 31’N, 76o 53’E 06/2/2018 144/51 10.51

Fig. 3. Flow chart of LST estimation

Table 3. Weighing coefficient of LANDSAT-8 bands [20]

Term Band-2 Band-3 Band-4 Band-5 Band-6 Band-7 Band-10

ωb 0.254 0.149 0.147 0.311 0.103 0.036 0

21st Jan 2018 and 6th Feb 2018. In the figure, it is observed served from the values that the obtained LST and weather
that the portion covered by vegetation is having a surface data are following the same pattern for the selected 12 mete-
temperature value as low. In contrast, barren land and orological stations (Figure 7). The Root Mean Square Error
urban area are having the highest amount of surface tem- (RMSE) and correlation coefficient (R) are used as statisti-
perature. As degree Celsius is used as a basic temperature cal criteria for evaluating the model efficiency. From the
unit for all analyses, the obtained LST value is converted comparison, it is found that RMSE and R for the first date
to degree Celsius and shown here. are 0.94o C,0.62, and for the second date, it is 1.26o C,0.88,
Comparison of estimated LST with the weather data: respectively. It is observed that LST and weather data are
For the region of interest, 12 points were selected for following the same pattern for the selected 12 meteorologi-
comparing LST and weather data. The difference between
weather data and LST is shown in Table 5. It can be ob-
588 Aryalekshmi B N et al.

Fig. 4. Vegetation Indices values for 21st Jan 2018 6th Feb 2018

Table 4. LANDSAT-8 Metadata

Thermal
Variable Description Value
Band
10 K1 774.8853
Constant
K2 1321.0789
RADIANCE M U 0.00033420
ML
LTB and
RADIANCE A D 0.1
AL
DB and
Qcal Pixel DN value

cal stations.
s
2
∑ ( Tw − LST ) Fig. 5. Land Surface Temperature for 21st Feb 2018
RMSE = (10)
n
∑ ( Tw − Tw0 ) ( LST − LST 0 )
R= q q (11)
2 2
∑ ( Tw − Tw0 ) Σ ( LST − LST 0 ) station temperature and Land surface Temperature, respec-
Where Tw and LST are the weather station temperature, tively.
and Land Surface Temperature, respectively, and n repre- For the analysis of variation in the surface temperature
sents the number of points. Tw0 and LST’ are mean weather of Mandya district, Karnataka, India, the remote sensing
 Journal of Applied Science and Engineering, Vol. 23, No 4, Page 583-591 589

Table 5. Weighing coefficient of LANDSAT-8 bands[20]

Sl Location Weather data LST Difference Weather data LST Difference

No: (Temp o C) (o C) (Temp o C) (o C)
21st Jan 2018 6th Feb 2018
1 Belluru 29.7 29 0.7 32.5 32 0.5
2 Bindiganavile 29.1 28 1.1 29.8 28 1.8
3 Kikkeri 30.8 31 -0.2 33.1 32 1.1
4 Krishnarajpet 30.5 31 -0.5 32.5 31 1.5
5 Akkehebbalu 30.4 30 0.4 33.5 32 1.5
6 Bukanakere 29.6 29 0.6 31.8 31 0.8
7 Sheelanere 30.7 29 1.7 32.4 31 1.4
8 Devalapura 30.6 29 1.6 33.5 33 0.5
9 Koppa 30 30 0 33.5 32 1.5
10 Athaguru 31.1 30 1.1 32.8 31 1.8
11 Maddur 30.2 30 0.2 32.3 31 1.3
12 Boppagowdanapura 30.2 29 1.2 31 31 0

and estimated for the region of interest. It is observed that

LST and weather data are following the same pattern for
the selected 12 meteorological stations. The study results
verified that LANDSAT-8 data could be further used for
LST estimation of the Mandya region as well, extended to
neighboring districts that lie on the Cauvery river basin.
From the study, it is seen that as the vegetation index, i.e., as
NDVI increases LST value has come down and vice versa.
Thus, more green areas are recommended for controlling
the land cover changes in the area of interest due to LST.
We conclude that all the study results can be used for the
development of both urban as well as the agricultural area
of the Mandya district and its surroundings.

Fig. 6. Land Surface Temperature for 6th Feb 2018 Acknowledgement

The authors are incredibly grateful to RRSC-South Banga-
techniques were used in the present study. The analysis of lore for providing the resources to carry out the work. The
LST helps in understanding climate variations, which can LANDSAT-8 satellite data from USGS (US geological sur-
also be used for understanding the urban heat island effect. vey) and weather data from KSNDMC are acknowledged
It is seen that the vegetation and river basin areas that are in this research.
acting as heat sink have a relatively lower temperature.
The more vegetation leads to lower LST, as it enhances
the ET that maintains the heat flux. It is also seen that for
different LULC, LST and NDVI have different effects. The
conclusion of this paper is summarized as follows:

5. Conclusion
The availability of remote sensing thermal data having
high temporal and spatial resolution offers many oppor-
tunities. The TIR images correlated with weather station
data allows the spatial distribution of LST to be modeled
590 Aryalekshmi B N et al.

Fig. 7. Plot showing LST and weather data for 6th Feb 2018

References Zaccaria, Blake L Sanden, Richard L Snyder, Jinbao

Jiang, and Jan W Hopmans. Evapotranspiration esti-
[1] M Bisquert, J. M. Sánchez, R. López-Urrea, and
mate over an almond orchard using Landsat satellite
V. Caselles. Estimating high resolution evapotranspira-
observations. Remote Sensing, 9(5), 2017.
tion from disaggregated thermal images. Remote Sens-
[8] B. N. Aryalekshmi, Rajashekhar C. Biradar, and J. Mo-
ing of Environment, 187:423–433, 2016.
hammed Ahamed. Thermal imaging techniques in agri-
[2] Qinqin Sun, Jianjun Tan, and Yonghang Xu. An ER-
cultural applications. International Journal of Innovative
DAS image processing method for retrieving LST and
Technology and Exploring Engineering, 8(12):2162–2168,
describing urban heat evolution: A case study in the
oct 2019.
Pearl River Delta Region in South China. Environmental
[9] Ann Kathrin Holtgrave, Michael Förster, Felix
Earth Sciences, 59(5):1047–1055, jan 2010.
Greifeneder, Claudia Notarnicola, and Birgit Klein-
[3] Marina Stathopoulou and Constantinos Cartalis. Day-
schmit. Estimation of Soil Moisture in Vegetation-
time urban heat islands from Landsat ETM+ and Corine
Covered Floodplains with Sentinel-1 SAR Data Us-
land cover data: An application to major cities in
ing Support Vector Regression. PFG - Journal of Pho-
Greece. Solar Energy, 81(3):358–368, 2007.
togrammetry, Remote Sensing and Geoinformation Science,
[4] Stefania Bonafoni and Chaiyapon Keeratikasikorn.
86(2):85–101, apr 2018.
Land surface temperature and urban density: Multi-
[10] Syed Muhammad Zubair Younis and Javed Iqbal. Es-
year modeling and relationship analysis using modis
timation of soil moisture using multispectral and FTIR
and landsat data. Remote Sensing, 10(9), 2018.
techniques. Egyptian Journal of Remote Sensing and Space
[5] Lidia Vlassova, Fernando Pérez-Cabello, Marcos Ro-
Science, 18(2):151–161, 2015.
drigues Mimbrero, Raquel Montorio Llovería, and Al-
[11] S K Dash and J. C.R. Hunt. Variability of climate
berto García-Martín. Analysis of the relationship be-
change in India, 2007.
tween land surface temperature and wildfire severity
[12] Lingli Wang and John J. Qu. Satellite remote sensing
in a series of landsat images. Remote Sensing, 6(7):6136–
applications for surface soil moisture monitoring: A
6162, 2014.
review, 2009.
[6] Salman Özupekce. Use of Land Surface Temperature
[13] Ricardo Trezza, Richard G Allen, and Masahiro
(Lst) Data in the Determination of High Areas of Forest
Tasumi. Estimation of actual evapotranspiration along
Fire Risk in Erdemli District. Journal of International
the Middle Rio Grande of New Mexico using MODIS
Social Research, 11(57):227–232, 2018.
and landsat imagery with the METRIC model. Remote
[7] Ruyan He, Yufang Jin, Maziar M Kandelous, Daniele
 Journal of Applied Science and Engineering, Vol. 23, No 4, Page 583-591 591

Sensing, 5(10):5397–5423, 2013. Journal of Remote Sensing and Space Science, 19(2):285–
[14] Geng Ming Jiang and Ronggao Liu. Retrieval of 295, 2016.
sea and land surface temperature from SVISSR/FY- [27] W G M Bastiaanssen. Remote Sensing in Water Resources
2C/D/E measurements. IEEE Transactions on Geoscience Management : The State of the Art. 1998.
and Remote Sensing, 52(10):6132–6140, 2014. [28] José A. Sobrino, Juan C. Jiménez-Muñoz, Guillem
[15] C. Ottlé and D Vidal-Madjar. Estimation of land sur- Sòria, Mireia Romaguera, Luis Guanter, José Moreno,
face temperature with NOAA9 data. Remote Sensing of Antonio Plaza, and Pablo Martínez. Land surface emis-
Environment, 40(1):27–41, 1992. sivity retrieval from different VNIR and TIR sensors.
[16] A Rajeshwari. ESTIMATION OF LAND SURFACE In IEEE Transactions on Geoscience and Remote Sensing,
TEMPERATURE OF DINDIGUL DISTRICT USING volume 46, pages 316–327, 2008.
LANDSAT 8 DATA. International Journal of Research [29] Kalifa Goïta and Alain Royer. Surface temperature
in Engineering and Technology, 03(05):122–126, 2014. and emissivity separability over land surface from com-
[17] Ugur Avdan and Gordana Jovanovska. Algorithm for bined tir and swir avhrr data. IEEE Transactions on
automated mapping of land surface temperature using Geoscience and Remote Sensing, 35(3):718–733, 1997.
LANDSAT 8 satellite data. Journal of Sensors, 2016, 2016. [30] Andrew M. Coutts, Richard J. Harris, Thu Phan,
[18] Emily G. Kullberg, Kendall C. DeJonge, and José L. Stephen J. Livesley, Nicholas S.G. Williams, and Nigel J.
Chávez. Evaluation of thermal remote sensing indices Tapper. Thermal infrared remote sensing of urban heat:
to estimate crop evapotranspiration coefficients. Agri- Hotspots, vegetation, and an assessment of techniques
cultural Water Management, 179:64–73, 2017. for use in urban planning. Remote Sensing of Environ-
[19] William Kustas and Martha Anderson. Advances in ment, 186:637–651, 2016.
thermal infrared remote sensing for land surface mod- [31] De Cai Wang, Gan Lin Zhang, Ming Song Zhao,
eling. Agricultural and Forest Meteorology, 149(12):2071– Xian Zhang Pan, Yu Guo Zhao, De Cheng Li, and Bob
2081, 2009. Macmillan. Retrieval and mapping of soil texture based
[20] USGS. Landsat Missions, 2019. on land surface diurnal temperature range data from
[21] Larry M. McMillin. Estimation of sea surface temper- MODIS. PLoS ONE, 10(6), 2015.
atures from two infrared window measurements with [32] Masahiro Tasumi, Richard G Allen, Ricardo Trezza,
different absorption. Journal of Geophysical Research, and James L Wright. Satellite-based energy balance to
80(36):5113–5117, dec 1975. assess within-population variance of crop coefficient
[22] Offer Rozenstein, Zhihao Qin, Yevgeny Derimian, and curves. Journal of Irrigation and Drainage Engineering,
Arnon Karnieli. Derivation of land surface tempera- 131(1):94–109, jan 2005.
ture for landsat-8 TIRS using a split window algorithm. [33] Rizqi I. Sholihah, Bambang H. Trisasongko, Diar Shid-
Sensors (Switzerland), 14(4):5768–5780, 2014. diq, La Ode S. Iman, Selamet Kusdaryanto, Manijo, and
[23] Zhao Liang Li, Bo Hui Tang, Hua Wu, Huazhong Dyah R. Panuju. Identification of Agricultural Drought
Ren, Guangjian Yan, Zhengming Wan, Isabel F. Trigo, Extent Based on Vegetation Health Indices of Landsat
and José A. Sobrino. Satellite-derived land surface Data: Case of Subang and Karawang, Indonesia. Proce-
temperature: Current status and perspectives, 2013. dia Environmental Sciences, 33:14–20, 2016.
[24] ARCGIS. Chapter 5 Geospatial Modeling 5.1 Ratio- [34] Anju Bala, Prakashkiran S Pawar, Anil Kumar Misra,
nale 5.2 SEBAL modeling for evapotranspiration 5.3 and Kishan Singh Rawat. Estimation and validation of
Penman-Monteith method for evapotranspiration 5.4 actual evapotranspiration for wheat crop using SEBAL
Runoff modeling 5.5 Recharge modeling 5.6 Weighed model over Hisar district, Haryana, India. Current
Overlay Analysis using GIS 5.1 Rationale. Technical Science, 113(1):134–141, 2017.
report.
[25] H. R. Shwetha and D. Nagesh Kumar. Performance
evaluation of satellite-based approaches for the estima-
tion of daily air temperature and reference evapotran-
spiration. Hydrological Sciences Journal, 63(9):1347–1367,
jul 2018.
[26] Athos Agapiou, Nikos Papadopoulos, and Apostolos
Sarris. Monitoring olive mills waste disposal areas in
Crete using very high resolution satellite data. Egyptian

View publication stats