Artificial Intelligence in Agriculture 6 (2022) 90–99

Contents lists available at ScienceDirect

Artificial Intelligence in Agriculture

journal homepage: http://www.keaipublishing.com/en/journals/artificial-
intelligence-in-agriculture/

Analysis of land surface temperature using Geospatial technologies in

Gida Kiremu, Limu, and Amuru District, Western Ethiopia
Mitiku Badasa Moisa a,⁎, Bacha Temesgen Gabissa b, Lachisa Busha Hinkosa a,
Indale Niguse Dejene c, Dessalegn Obsi Gemeda d
a
Department of Agricultural Engineering, Faculty of Technology, Wollega University Shambu campus, Shambu, Ethiopia
b
Department of Fisheries and Aquatic Sciences, Faculty of Agriculture, Wollega University Shambu campus, Shambu, Ethiopia
c
Department of Earth Sciences, College of Natural and Computational Sciences, Wollega University Nekemte Campus, Nekemte, Ethiopia
d
Department of Natural Resource Management, College of Agriculture and Veterinary Medicine, Jimma University, Jimma, Ethiopia

a r t i c l e i n f o a b s t r a c t

Article history: Degradation of vegetation cover and expansion of barren land are remained the leading environmental
Received 22 February 2022 problem at global level. Land surface temperature (LST), Normalized Difference Vegetation Index
Received in revised form 26 May 2022 (NDVI), Normalized Difference Barren Index (NDBaI), and Modified Normalized Difference Water Index
Accepted 20 June 2022
(MNDWI) were used to quantify the changing relationships using correlation analysis. This study
Available online 23 June 2022
attempted to analyze the relationship between LST and NDVI, NDBaI, and MNDWI using Geospatial tech-
Keywords:
nologies in Gida Kiremu, Limu, and Amuru districts in Western Ethiopia. All indices were estimated by
LST using thermal bands and multispectral bands from Landsat TM 1990, Landsat ETM+ 2003, and Landsat
NDVI OLI/TIRS 2020. The correlation of LST with NDVI, NDBaI and MNDWI were analyzed by using scatter
NDBaI plot. Accordingly, the NDBaI was positive correlation with LST (R 2 = 0.96). However, NDVI and
MNDWI MNDWI were substantially negative relationship with LST (R 2 = 0.99, 0.95), respectively. The result
Satellite data shows that, LST was increased by 5 °C due to decline of vegetation cover and increasing of bare land
over the study periods. Finally, our result recommended that, decision-makers and environmental ana-
lysts should give attention on the importance of vegetation cover, water bodies and wetland in climate
change mitigation, particularly, LST in the study area.
© 2022 The Authors. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co., Ltd. This is an open
access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction increasing trend of LST (Moisa et al., 2022b; Wolteji et al., 2022;
Merga et al., 2022; Dissanayake et al., 2019). Similar to LST, the atmo-
The majority of human activities are the primary cause of the contin- spheric temperature is significantly increasing as a results of land use
uous loss of vegetation cover on the earth's surface (Sahana et al., 2016). land cover change (Gemeda et al., 2020; Gemeda et al., 2021; Gemeda
The decrease in vegetation was a contributing factor to the increase in et al., 2022; Moisa et al., 2022c).
land surface temperature (LST). Song et al. (2021) highlights that vege- To quantitatively describe degraded land, the Normalized Difference
tation is the main influencing factors that driven LST. The LST is one of Barren Index (NDBaI) and the Modified Normalized Difference Water
the crucial indicators of ecological functioning of the environment. The Index (MNDWI) were used (Gao, 1996; Zha et al., 2003). The LST
LST is indicates the temperature of the Earth's skin, which is an impor- were compared to the Normalized Difference Vegetation Index
tant component in regional and global land surface processing studies (NDVI), NDBaI, and MNDWI to provide ambient environmental condi-
(Thanh et al., 2018). The LST is increasing at global level due to land tion for living things (Qinqin et al., 2012). Vegetation based indices
use land cover conversion (Moisa et al., 2022a). Studies conducted can indicates the existence and abundancies of vegetation cover
in different parts of Ethiopia confirmed that there is a substantial (Moisa et al., 2022a; Iang and Tian, 2010).
Analysis of spatial flexibility of the NDVI, LST, NDBaI, and MNDWI
are tremendously crucial for decision making and natural resources
Abbreviations: LST, Land surface temperature; MIR, Middle infrared; MNDWI, monitoring in natural and environmental investigations (Zareie et al.,
Modified normalized difference water index; NDBaI, Normalized difference barren
index; NDVI, Normalized difference vegetation index; NIR, Near infrared.
2016). In addition, NDVI has been used to confirm the importance for
⁎ Corresponding author. represent distribution of green space (Yuan and Bauer, 2007). Further-
E-mail address: [email protected] (M.B. Moisa). more, several scholars raised that, remote sensing indices like NDVI

https://doi.org/10.1016/j.aiia.2022.06.002
2589-7217/© 2022 The Authors. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co., Ltd. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).
M.B. Moisa, B.T. Gabissa, L.B. Hinkosa et al. Artificial Intelligence in Agriculture 6 (2022) 90–99

and MNDWI were gradually declined due to forest fires, deforestation, 2. Materials and methods
expansion of urbanization, agricultural land expansion, and expansion
of grazing land from which NDBaI, and LST were increased (Zhou and 2.1. Description of the study area
Wang, 2011; Mimbrero et al., 2014).
Although these indices have been used to model LST in previous This study is conducted in East Wollega and Horo Guduru Wollega
study (Chen et al., 2006), only a few of them have compared index re- Zones and located between 9027′00′′ and 10018′00′′ N, and 36019′30′′
sults from different years to monitor surface temperature. Even though, and 37010′30′′ E. Two districts namely: Gida Kiremu and Limu selected
agricultural expansion is the main causes that aggravated climate is located in East Wollega Zone while Amuru district is administratively
change particularly LST in western parts of Ethiopia. No research has located in the Horo Guduru Wollega Zone, Oromia National Regional
been conducted around Gida Kiremu, Limu, and Amuru district. It is ob- State Western Ethiopia (Fig. 1). The study area elevation ranges from
vious that the LST has increased over time due to a decline in vegetation 713.32 to 2496.61 m above mean sea level. It has a total area of
cover for agricultural expansion. In addition, wetland degradation and 5086.65km2.
the increment of barren land were other causes for the rising of LST.
Therefore, this study attempted to analyze LST based on remote sensing
indices such as NDVI, NDBaI, and MNDWI in western Ethiopia. This 2.2. Climate
paper is organized as follows: the first section introduces the general in-
troduction on LST and its correlation with different indices; Section two The average monthly rainfall of the current study area is ranged
describes materials and methods which includes the study area descrip- between 14.32 and 338 mm /year. The study area receive high rain-
tions, data types and sources, and method of data analysis; Section three fall during summer that starts in June and ends in September and
presents' results and discussions and the final section concludes the short rain season is in, spring, which encompass March, April, and
findings of the study. May.

Fig. 1. Location map of the study area.

91
M.B. Moisa, B.T. Gabissa, L.B. Hinkosa et al. Artificial Intelligence in Agriculture 6 (2022) 90–99

2.2.1. Soil types band 3 for Landsat 8) and middle infrared (band 5 for Landsat 5 and 7,
Soil types of the study area were classified into sixteen in number band 6 for landsat 8) reflectance values (Eq. (2)).
namely: - calcic cambisols,calcic xerosols,chromic cambisols, chromic
Green  MIR
luvisols, dystric gleysols, dystric nitisols, eutric cambisols, eutric nitisols, MNDWI ¼ ð2Þ
gypsic yermosols, haplic xerosols, leptosols, orthic acrisols, orthic Green þ MIR
luvisols, orthic solonchaks, phaeozems, vertic cambisols from which
dystric nitisols is the most dominants(2123.5 km2) and least dominant
2.5.3. Normalized difference barren index (NDBaI)
soil type is phaeozems with an area of (2 km2).
When estimating the thermal environment, NDBaI was chosen to
represent barren terrain areas, which often show large variances in
2.3. Socio economic activities
thermal characteristics (Zhifeng and Jianjun, 2012). It is estimated
using the reflectance of medium infrared (band 5 for Landsat 5 and 7,
Mixed agriculture namely crop production and livestock production are
band 6 for Landsat 8) and thermal infrared (band 6 for Landsat 5 and
the most popular income sources of smallholders farmers. From crop pro-
7, band 10 and 11 for Landsat 8) satellites (Eq. (3)).
duction, Maize, groundnut, niger seed, lentil, beans, and peas and some veg-
etables (potato, onions, garlic) and fruits like mango, papaya, oranges and MIR  TIR
bananas are mostly dominants agricultural products of the study areas. NDBaI ¼ ð3Þ
MIR þ TIR
The agricultural activity of the study area is mainly depending on rain-fed.

2.4. Data types and sources 2.5.4. Retrieval of land surface temperature (LST)
The mono-window algorithm (Qin et al., 2001), single channel, was
Landsat images from three years were used in this study. The USGS created from Landsat TM, ETM+, and OLI/TIRS data (Jiménez-Muñoz
(http://earthexplorer.usgs.gov/) provides free downloads of the ther- and Sobrino, 2003). LST was extracted from Landsat 8 data with one
mal and multispectral bands of Landsat TM 990, Landsat ETM+ 2003, bands using mono window method from band 10. LST is determined
and Landsat OLI/TIRS 2020. Detailed information regarding these data by employing the brightness temperature of the two bands of Thermal
is presented in (Table 1). Software used for this study were, ArcGIS Infrared (TIR), as well as the mean and difference in land surface emis-
10.3 version and ERDAS imagine 2015. sivity (Cheng et al., 2015).

2.5. Method of data analysis 2.5.4.1. Step I: Conversion of DN in to radiance.

A. Mono window algorithm
The LST, NDVI, MNDWI, and NDBaI were calculated in this study. The
methodology flow diagram of the study was presented in (Fig. 2).
Before calculating the brightness temperature, the mono window
technique converted the digital data into an at-sensor radiance sensor.
2.5.1. Normalized difference vegetation index (NDVI)
The TM and ETM+ DN value ranges between 0 and 255 (Eq. (4)).
This index was used to calculate the amount of vegetation covering
the earth's surface. The NDVI was estimated using multi-spectral LMAXλ  LMINλ
bands from Landsat images collected over the course of the investiga- Lλ ¼ x ðDN  QCALMIN Þ þ LMINλ ð4Þ
QCALMAX  QCALMIN
tion. Band 4 was used to measure near-infrared bands on Landsat 5
and 7, and band 5 was used for Landsat 8. The red band of the Landsat where:
data were measured using band 4 for Landsat 8 and band 3 for Landsat QCLA = quantized calibrated pixel value in Digital Number (DN).
5 and 7. The formula for this index is presented in (Eq. (1)) LMINλ = spectral radiance scaled to QCALMIN in watts/(meter
squared*ster* μm).
NIR  R LMAXλ = spectral radiance scaled to QCALMAX in watts/(meter
NDVI ¼ ð1Þ
NIR þ R squared*ster* μm).
QCALMIN = minimum quantized calibrated pixel value correspond-
The NDVI scale spans from −1.0 to 1.0. For healthy and dense vege- ing to LMINλ in DN.
tation, the NDVI values are always between 0.2 and 0.9 (Bustos and QCALMAX = maximum quantized calibrated pixel value corre-
Meza, 2015). Vegetation such as rock, water, and barren plains, on the sponding to LMAXλ in DN = 255.
other hand, were represented by values less than 0.1 (Fu and Burgher, For Landsat 8, the mono window algorithm (MWA) was also applied
2015). to estimate the LST as used by other experts (Aik et al., 2020; Sahana
et al., 2016; Atitar and Sobrino, 2009). The Landsat 8 TIRS Digital Num-
2.5.2. Modified normalized difference water index (MNDWI) bers (DNs) for band 10 were first converted into spectral radiance
When modelling the thermal environment, the MNDWI is assigned (Eq. (5)).
to represent water areas, which typically display notable variations in
Lλ ¼ ðML∗QCalÞ þ AL ð5Þ
thermal characteristics (Xu, 2008; Zhifeng and Jianjun, 2012). The for-
mula was developed using the green band (band 2 for landsat 5 and 7,
where;
Lλ is Top of Atmosphere (TOA) spectral radiance (Wm−2 sr−1
Table 1 μm−1).
Remote sensing data used for the study.
ML is Band-specific multiplicative rescaling factor from the metadata
Satellite Path/Row Sensor Resolution No of Date of (RADIANCE_MULT_BAND x, where x is number of bands).
Image (m) bands Acquisitions AL is a band-specific additive rescaling factor from the metadata
Landsat 170/53 TM 30 7 12-01-1990 (RADIANCE_ADD_BAND_x, where x is the band number).
5 QCal is quantized and calibrated standard product pixel values (DN).
Landsat 170/53 ETM+ 30 7 20-02-2003
7
Landsat 170/53 OLI/TIRS 30 11 11-02-2020 2.5.4.2. Step II. Conversion to temperature (ETM+). Based on land surface
8 emissivity, atmospheric trans-emissivity, brightness temperature, and

92
M.B. Moisa, B.T. Gabissa, L.B. Hinkosa et al. Artificial Intelligence in Agriculture 6 (2022) 90–99

Fig. 2. Methodological flowchart of the study.

average atmospheric temperature, the mono-window algorithm was W is content of water vapors in the atmosphere;
used to determine the LST (Zhang et al., 2006). The spectral radiance ℇ is LSE of band 10 to estimate LST (Eq. (7)):
(as mentioned above) of TM and ETM+ Band 6 imaging can also be
K2
transformed into a more physically usable quantity (Eq. (6)). The con- BT ¼
ð7Þ
Lλ þ 1
K1
ln
version formula is as follows:

K2 where;
T¼
ð6Þ BT: is effective at-sensor brightness temperature (K);
Lλ þ 1
K1
ln
K2: is calibration constant 2 (K);
where T = Effective at-satellite temperature in Kelvin. K1: is calibration constant 1 (W/ (m2 * sr * μm));
K2 = Calibration constant 2. Lλ: is spectral radiance at the sensor's aperture (W/ (m2 * sr * μm));
K1 = Calibration constant 1. and.
Lλ = Spectral radiance in watts/(meter squared * ster * μm). Ln: is natural logarithm.
For this investigation, mono window algorithm (MW) was also used
2.5.4.3. Step III: Land surface emissivity estimation. According to Sobrino
to calculate LST in landsat 8. It calculates mean land surface emissivity
et al. (2004), the emissivity is calculated using (Eq. (9))
and then to estimates the brightness temperatures from band 10 of
landsat 8. ε ¼ 0:004∗PV þ 0:986 ð8Þ
TB10 is brightness temperature of band 10 (Kelvin K);
ℇ is mean value of Land Surface Emissivity (LSE) of TIR bands; where PV is the vegetation proportion obtained according to.

93
M.B. Moisa, B.T. Gabissa, L.B. Hinkosa et al. Artificial Intelligence in Agriculture 6 (2022) 90–99

Carlson and Ripley formula (10); Finally, the findings of Landsat TM, ETM+ and OLI/TIRS LST mea-
  surements were converted to degrees Celsius by subtracting 273.15.
NDVI  NDVImin 2 (Eq. (11)) was used to convert temperature in degrees Kelvin (K) to de-
PV ¼ ð9Þ
NDVImax  NDVImin grees Celsius (°C).

The calculated radiant surface temperatures will be corrected for 0C ¼ K  273:15 ð11Þ
emissivity using the equation (Eq. (11)):
where: °C = LST result in degree Celsius;
TB K = LST result in degree Kelvin.
LST ¼
ð10Þ
1 þ λ TB
P Inε
3. Result and discussion
where;
LST: land surface temperature (in Kelvin); 3.1. Analysis of land surface temperature
TB: radiant surface temperature (in Kelvin).
λ: the wavelength of emitted radiance (10.8 μm), and Inε is The spatial pattern of LST in the study area was determined for
emissivity. the years 1990, 2003, and 2020, respectively. In all of the years stud-
ρ: h × c/σ (1.438 × 10−2 mK); h is Planck's constant (6.26 × 10–34 J s); ied, the northeastern and southwestern parts of the study area had
c is the velocity of light (2.998 × 108 m/s); σ is Stefan Boltzmann's high LST (Fig. 3). The increasing of LST in the study area is associated
constant (1.38 × 10 − 23J K−1); and ε is land surface emissivity. with the declining of vegetation cover and increasing of bare land.

Fig. 3. LST map of the study area.

94
M.B. Moisa, B.T. Gabissa, L.B. Hinkosa et al. Artificial Intelligence in Agriculture 6 (2022) 90–99

Fig. 4. Normalized Difference Vegetation Index map of the study area.

Due to the existance of substantial vegetation cover, the eastern, Sub-basin between the year 1991, and 2020. As compared to 1990,
central, and western parts of the research region experienced rela- and 2003, the year 2020 is the most extreme temperature. The increas-
tively low LST. ing trend of LST is associated with the declining of vegetation cover and
The mean LST gradually increased from 23.70 °C in 1990 to 24.30 °C
in 2003 to 28.70 °C in 2020. From 1990 to 2020, an average temperature
was increased by 5 °C. Comparable result has been reported by Moisa 50 y = -38.996x + 38.422
45 R² = 0.9989
et al., 2022a, which found an increment of LST by 5.6 °C in Anger River
40
35
LST (0C)

30
Table 2 25
Correlation between LST, NDVI, NDBaI and MNDWI. 20
15
Correlation LST NDBaI NDVI MNDWI 10
5
LST 1 0
NDBaI 0.95767⁎ 1 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6
NDVI −0.9995⁎ −0.9627⁎ 1 NDVI
MNDWI −0.9581⁎ −0.979⁎ 0.98452⁎ 1
⁎ shows that Correlation values between them. Fig. 5. Relationship of LST and NDVI of study area.

95
M.B. Moisa, B.T. Gabissa, L.B. Hinkosa et al. Artificial Intelligence in Agriculture 6 (2022) 90–99

wetland degradation driven by expansion of agricultural land. The LST is confirmed that, LST was increased as decline of vegetation cover and
expected to increase due to global warming. For instance, Rasul et al. both of them had strong negative relationship.
(2012) predict that by 21 century, the temperature will have increased
by 1.40 °C to 5.80 °C during the twentieth century which confirmed the
3.3. Correlation between LST and NDBaI
present study.
Barren land of the study area is increasing due to agricultural ex-
pansion and decline of vegetation cover. The results demonstrate a
3.2. Correlation between LST and NDVI
substantial positive correlation between LST and NDBaI (R 2 =
0.96). It demonstrates that high LST can be found on degraded or
LST values were discovered to range from maximum temperature
barren terrain with a high NDBaI value (Fig. 6). The relationship
(43.2 °C) to minimum temperature (16.9 °C), whereas NDVI values
between two parameters was shown in (Fig. 7). The result of the
are found to range from 0.50 to −0.53 maximum to minimum. The re-
study more consistency with the result of (Zhifeng and Jianjun,
sults show that LST and NDVI have a strong negative correlation (R2 =
2012) and they stated that barren land was substantial positive
0.99). Wolteji et al. (2022) found a moderate negative relationship be-
correlation with LST.
tween NDVI and LST over the Rift Valley Region of Ethiopia. It shows
that high LST are more closely related to low vegetation cover and
vice versa. The distribution of NDVI over the study period is presented 3.4. Correlation between LST and MNDWI
in (Fig. 4). The relationship between the two factors (LST and NDVI) in
2020 is shown in (Table 2 and Fig. 5). This result of study is in line Vegetation water content (wetland) was decreased due to de-
with (Alemu, 2019; Wedajo et al., 2019; Merga et al., 2022) they cline of vegetation and increasing of LST. Expansion of agricultural

Fig. 6. Normalized Difference Barren Index (NDBaI) map of the study area.

96
M.B. Moisa, B.T. Gabissa, L.B. Hinkosa et al. Artificial Intelligence in Agriculture 6 (2022) 90–99

50 y = 16.13x + 33.652
R² = 0.96 MNDWI, with coefficient determination of (R2 = 0.95). MNDWI is
45
40 high in the northern and southern parts of the study area (Fig. 8).
35 The findings reveal that high LST are more commonly recorded in
LST (0C)

30 low-water locations. The relationship between LST and MNDWI

25 was presented in (Fig. 9). The result is agreed with the result of pre-
20
vious studies (Zhou and Wang, 2011; Zhifeng and Jianjun, 2012). The
15
10 relationship between LST and other indices were reported in
5 (Table 2). The NDBaI has a direct association with LST, while NDVI
0 and MNDWI have an inverse relationship with LST.
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
NDBaI

3.5. Correlation between LST and Time of study

Fig. 7. Relationship of LST and NDBaI of study area.

The mean LST of the study area was correlated the study period. The
land and increasing of barren land which together increase LST was result shows that mean of LST was positive relationship with time of
the main factors for decreasing of MNDWI in the study area. The study with R2 = 0.89. As time increased mean of LST was increased
results shows that a negative strong association between LST and (Fig. 10).

Fig. 8. Modified Normalized Difference Water Index (MNDWI) map of the study area.

97
M.B. Moisa, B.T. Gabissa, L.B. Hinkosa et al. Artificial Intelligence in Agriculture 6 (2022) 90–99

50 y = -30.204x + 23.743
R² = 0.9581 Author contributions
45
40
35 MBM involved in research design, data collection, data analysis, and
LST (0C)

30 draft manuscript writing. LBH and END involved in data analysis. DOG
25 and BTG participated in methodology, data analysis and manuscript
20
edition. All authors read and approved the final manuscript.
15
10
5 CRediT authorship contribution statement
0
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4
Mitiku Badasa Moisa: Conceptualization, Data curation, Formal
MNDWI
analysis, Investigation, Methodology, Resources, Software, Validation,
Fig. 9. Relationship of LST and MNDWI of study area.
Visualization, Writing – original draft, Writing – review & editing.
Bacha Temesgen Gabissa: Formal analysis, Investigation, Methodology,
Resources, Writing – review & editing. Lachisa Busha Hinkosa: Formal
analysis, Investigation, Methodology, Resources, Writing – review &
editing. Indale Niguse Dejene: Formal analysis, Investigation, Method-
35
ology, Resources, Writing – review & editing. Dessalegn Obsi Gemeda:
30 Formal analysis, Investigation, Methodology, Resources, Writing –
25 review & editing.
LST mean (0C)

20 y = 0.1713x - 317.74
Declaration of Competing Interest
R² = 0.8909
15
10 The authors declared no competing interest.
5
Acknowledgments
0
1985 1990 1995 2000 2005 2010 2015 2020 2025
Year The authors would like to acknowledge Wollega University Shambu
Campus Faculty of Technology, Faculty of Agriculture, Wollega Univer-
Fig. 10. Relationship of LST with year of study. sity Nekemte campus College of Natural and computational sciences
and Jimma University College of Agriculture and Veterinary Medicine
for the existing facilities to conduct this study.
4. Conclusions
References
In this article, we use remote sensing data to assess the spatial- Aik, D.H.J., Ismail, M.H., Muharam, F.M., 2020. Land use/land cover changes and the rela-
temporal LST in relation to NDVI, NDBaI, and MNDWI indices tionship with land surface temperature using Landsat and MODIS imageries in Cam-
in three districts of western Ethiopia: Gida Kiremu, Limu, and eron Highlands, Malaysia. Land 9, 372.
Alemu, M.M., 2019. Analysis of Spatio-temporal land surface temperature and normalized
Amuru. Between 1990 and 2020, the temperature of the earth's sur- difference vegetation index changes in the Andassa watershed, Blue Nile Basin,
face increased by 5 °C. Our results concluded that the NDVI and Ethiopia. J. Resour. Ecol. 10 (1), 77–85.
MNDWI have a substantial negative relationship with LST, whereas Atitar, M., Sobrino, J., 2009. A split-window algorithm for estimating LST from meteosat 9
data: test and comparision with data and MODIS LSTs. IEEE Geosci. Remote Sens. Lett.
NDBaI has a positive correlation with LST. Expansion of agricultural 6, 122–126.
land is the main cause for decline of vegetation cover in the study Bustos, E., Meza, F.J., 2015. A method to estimate maximum and minimum air tempera-
area. Decline of vegetation cover is the cause for decrement of veg- ture using MODIS surface temperature and vegetation data: application to the
Maipo Basin, Chile. Theor. Appl. Climatol. 120 (1–2), 211–226. https://doi.org/10.
etation water content (wetland) and increasing of barren land. 1007/s00704-014- 1167-2.
Moreover, LST is increasing over time and expanded to all directions Chen, X., Zhao, H., Li, P., Yin, Z., 2006. Remote sensing image-based analysis of the rela-
in study area. Based on our findings, this study recommended the tionship between urban heat island and land use/cover changes. Remote Sens. Envi-
ron. 104, 133–146.
community awareness creation to promote the wise use of natural
Cheng, X., Wei, B., Chen, G., Song, C., 2015. Influence of park size and its surrounding
resources for its sustainability. In addition, further studies on the urban landscape patterns on the park cooling effect. J. Urban Plan. Dev. 141,
impacts of LST agricultural production can enhance our understand- A4014002.
ing of the impacts of environmental changes on community Dissanayake, D., Morimoto, T., Murayama, Y., Ranagalage, M., 2019. Impact of landscape
structure on the variation of land surface temperature in sub-Saharan region: a
livelihhoods. case study of Addis Ababa using Landsat data (1986–2016). Sustainability 11, 2257.
https://doi.org/10.3390/su11082257.
Fu, B., Burgher, I., 2015. Riparian vegetation NDVI dynamics and its relationship with cli-
Consent for publication mate, surface water and groundwater. J. Arid Environ. 113, 59–68. https://doi.org/10.
1016/j.jaridenv.2014. 09.010.
Gao, B.C., 1996. NDWI a normalized difference water index for remote sensing of vegeta-
The authors agreed to publish the manuscript on Artificial Intelli- tion liquid water from space. Remote Sens. Environ. 58, 257–266.
gence in Agriculture. Gemeda, D.O., Feyssa, D.H., Garedew, W., 2020. Meteorological data trend analysis and
local community perception towards climate change: a case study of Jimma City,
southwestern Ethiopia. Environ. Dev. Sustain. 23, 5885–5903.
Gemeda, D.O., Korecha, D., Garedew, W., 2021. Evidences of climate change presences in
Availability of data the wettest parts of southwest Ethiopia. Heliyon 7 (9), e08009. https://doi.org/10.
1016/j.heliyon.2021.e08009.
Gemeda, D.O., Korecha, D., Garedew, W., 2022. Monitoring climate extremes using stan-
Available in this manuscript. dardized evapotranspiration index and future projection of rainfall and temperature
in the wettest parts of Southwest Ethiopia. Environ. Challenges 7, 100517. https://doi.
org/10.1016/j.envc.2022.100517.
Funding Iang, J., Tian, G., 2010. Analysis of the impact of land use/ land cover change on land sur-
face temperature with remote sensing. Procedia Environ. Sci. 2, 571–575.
Jiménez-Muñoz, C., Sobrino, J., 2003. A generalized single-channel method for retrieving
No funding received for this research. land surface temperature from remote sensing data. J. Geophys. Res. 108.

98
M.B. Moisa, B.T. Gabissa, L.B. Hinkosa et al. Artificial Intelligence in Agriculture 6 (2022) 90–99

Merga, B.B., Moisa, M.B., Negash, D.A., Ahmed, Z., Gemeda, D.O., 2022. Land surface tem- 2019. Int. J. Appl. Earth Obs. Geoinf. 104 (2021), 102537. https://doi.org/10.1016/j.
perature variation in response to land-use and land-cover dynamics: a case of jag.2021.102537.
Didessa River sub-basin in Western Ethiopia. Earth Syst. Environ. https://doi.org/10. Thanh, Noi Phan, Martin, Kappas, Trong, Phuong, 2018. Land surface temperature varia-
1007/s41748-022-00303-3. tion due to changes in elevation in Northwest Vietnam. Climate 6, 28. https://doi.
Mimbrero, M.R., Vlassova, L., Pé, F., 2014. Analysis of the relationship between land sur- org/10.3390/cli6020028.
face temperature and wildfire severity in a series of landsat images. Remote Sens. Wedajo, G.B., Muleta, M.K., Gessesse, B., Koriche, S.A., 2019. Spatiotemporal cli-
6, 6136–6162. https://doi.org/10.3390/rs6076136. mate and vegetation greenness changes and their nexus for Dhidhessa River
Moisa, M.B., Dejene, I.N., Merga, B.B., Gemeda, D.O., 2022a. Impacts of land use/land cover basin, Ethiopia. Environ. Syst. Res. 8, 31. https://doi.org/10.1186/s40068-
dynamics on land surface temperature using geospatial techniques in Anger River 019-0159-8.
sub-basin, Western Ethiopia. Environ. Earth Sci. 81, 99. https://doi.org/10.1007/ Wolteji, B.N., Bedhadha, S.T., Gebre, S.L., Alemayehu, E., Gemeda, D.O., 2022. Multiple in-
s12665-022-10221-2. dices based agricultural drought assessment in the Rift Valley region of Ethiopia. En-
Moisa, M.B., Merga, B.B., Gemeda, D.O., 2022b. Multiple indices-based assessment of agri- viron. Challenges 7, 100488.
cultural drought: a case in Gilgel gibe sub-basin, Southern Ethiopia. Theor. Appl. Xu, H., 2008. A new index for delineating built-up land features in satellite imagery. Int.
Climatol. 148, 455–464. J. Remote Sens. 29, 4269–4276.
Moisa, M.B., Merga, B.B., Gemeda, D.O., 2022c. Urban heat island dynamics in response to Yuan, F., Bauer, M.E., 2007. Comparison of impervious surface area and normalized differ-
land use land cover change: a case of Jimma city, southwestern Ethiopia. Theor. Appl. ence vegetation index as indicators of surface urban heat island effects in Landsat im-
Climatol. https://doi.org/10.1007/s00704-022-04055-y. agery. Remote Sens. Environ. 106, 375–386.
Qin, Z., Karnieli, A., Berliner, P., 2001. A mono-window algorithm for retrieving land sur- Zareie, S., Khosravi, H., Nasiri, A., 2016. Using Landsat thematic mapper (TM) sensor to de-
face temperature from Landsat TM data and its application to the Israel-Egypt border tect change in land surface temperature in relation to land use change in Yazd, Iran.
region. Int. J. Remote Sens. 21, 3719–3746. Solid Earth 7, 1551–1564. https://doi.org/10.5194/se-7-1551-2016.
Qinqin, Sun, Zhifeng, Wu., Jianjun, Tan, 2012. The relationship between land surface tem- Zha, Y., Gao, J., Ni, S., 2003. Use of normalized difference built-up index in auto-
perature and land use/land cover in Guangzhou, China. Environ. Earth Sci. 65, matically mapping urban areas from TM imagery. Int. J. Remote Sens. 24,
1687–1694. https://doi.org/10.1007/s12665-011-1145-2. 583–594.
Rasul, G., Mahmood, A., Sadiq, A., Khan, S.I., 2012. Vulnerability of the Indus Delta to cli- Zhang, Y., Yu, T., Gu, X., Zhang, Y., Chen, L., 2006. Land surface temperature retrieval from
mate change in Pakistan. Pakistan J. Meteorol. 2012, 8(16). CBERS-02 IRMSS thermal infrared data and its applications in quantitative analysis of
Sahana, M., Ahmed, R., Sajjad, H., 2016. Analyzing land surface temperature distribution urban heat islands effects. J. Remote Sens. Beijing 10, 789.
in response to land use/land cover change using split window algorithm and spectral Zhifeng, W.U., Jianjun, Jim Tan, 2012. The relationship between land surface temperature
radiance model in Sundarban biosphere reserve, India. Model. Earth Syst. Environ. 2, and land use/land cover in Guangzhou, China. Environ. Earth Sci. 65, 1687–1694.
https://doi.org/10.1007/s12665-011-1145-2.
81. https://doi.org/10.1007/s40808-016-0135-5.
Zhou, Xiaolu, Wang, Yi-Chen, 2011. Dynamics of land surface temperature in response to
Sobrino, J.A., Jimenez-Munoz, J.C., Paolini, L., 2004. Land surface temperature retrieval
land-use/cover change. Geogr. Res. 49 (1), 23–36. https://doi.org/10.1111/j.1745-
from LANDSAT TM5. Remote Sens. Environ. 90, 434–440.
5871.2010.00686.x.
Song, Z., Yang, H., Huang, X., Yu, W., Huang, J., Ma, M., 2021. The spatiotemporal pattern
and influencing factors of land surface temperature change in China from 2003 to

99