Theoretical and Applied Climatology (2022) 148:455–464

[Link]

ORIGINAL PAPER

Multiple indices‑based assessment of agricultural drought: A case

study in Gilgel Gibe Sub‑basin, Southern Ethiopia
Mitiku Badasa Moisa1 · Biratu Bobo Merga2 · Dessalegn Obsi Gemeda3

Received: 2 July 2021 / Accepted: 27 January 2022 / Published online: 1 February 2022
© The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature 2022, corrected publication 2022

Abstract
Agricultural drought seriously affects the economies of developing countries due to heavily dependent on rain-fed agriculture.
The present study aimed to assess agricultural drought based on vegetation condition index (VCI) and vegetation health index
(VHI) using geospatial techniques, the case of Gilgel Gibe sub-basin southern Ethiopia. Landsat images of TM 1990, ETM + of
2000 and OLI/TIRS of 2020 were used to calculate VCI and VHI for assessing agricultural drought. A normalized difference
vegetation index (NDVI) and land surface temperature (LST) were used to estimate agricultural drought through VCI and VHI.
The result revealed that NDVI declined from 1990 to 2020 due to expansion of agricultural lands and other socioeconomic
activities. The output of VCI values indicates that extreme drought was increased from 0.3 k­ m2 in 1990 to.4 k­ m2 and 0.7
­km2 in 2000 and 2020, respectively. In addition, severe drought occurred in 2020 on an area of 6 ­km2 (0.1%). Consequently,
moderate drought was increased by area of 3.3 ­km2, 22.2 ­km2 and 1051.1 ­km2 in 1990, 2000 and 2020, respectively, based on
VHI in the study area. Thus, to minimize the negative effects of agricultural drought on socioeconomic and well-being of the
society in the Gilgel Gibe sub-basin forest and other natural resources conservations are essentials not optional. To alleviate
the detrimental consequences of agricultural drought on socioeconomic and well-being, forest and other natural resources
conservation is necessary. Moreover, detailed research should be conducted in the study area by using high-resolution satellite
data with ground control points.

Keywords Agricultural drought · Geospatial techniques · Drought · Gilgel Gibe sub-basin · VCI · VHI · TCI

1 Introduction

Drought is a vital environmental threat that causes signifi-

cant damage to crop production, water resources and animals
(Sholihah et al. 2016; Sultana et al. 2021). The negative
effects of drought are more pronounced in tropical regions
* Mitiku Badasa Moisa (Gonzalez-Orozco et al. 2020). Agricultural drought sig-
mitikubadasa10@[Link] nificantly affects the economy of agrarian countries (Dutta
Biratu Bobo Merga et al. 2015). It is obvious that rain-fed agriculture is the
bira2bobo@[Link] most vulnerable to climate change as compared to irriga-
Dessalegn Obsi Gemeda tion-based agricultural systems. In addition, food shortages
dasoobsi@[Link] and household food insecurity were caused by the recurrent
1
Department of Agricultural Engineering, Faculty cycle of drought and extreme reliance on rain-fed agriculture
of Technology, Wollega University Shambu Campus, (Gebrehiwot et al. 2016; Gidey et al. 2018). Ethiopia has
Shambu, Ethiopia been experienced periodic droughts and dry spells in 1984,
2
Department of Geographic Information System, Institute 1991, 2002, 2009 and 2010 (Randell and Gray, 2016). Even
of Land Administration, Oda Bultum University, though some research has been conducted in the past, recent
Oda Bultum, Ethiopia environmental conditions that can accelerate the occurrence
3
Department of Natural Resource Management, College of agricultural drought are less studied in Ethiopia in gen-
of Agriculture and Veterinary Medicine, Jimma University, eral and in the Gilgel Gibe sub-basin in particular. Most of
Jimma, Ethiopia

13
Vol.:(0123456789)
456 M. B. Moisa et al.

drought characterization in the country is focused on pre- are growing. The flowchart methodology of this research is
cipitation and temperature data without considering other demonstrated in Fig. 2.
environmental factors. Thus, agricultural drought moni-
toring during the crop growing season is a crucial option
(Gidey et al. 2018). 4 Methodology
The vegetation health index (VHI), which is made up of
the vegetation condition index (VCI) and temperature con- 4.1 Vegetation‑Based Agricultural Drought
dition index (TCI), is used to monitor agricultural drought Monitoring
(Sholihah et al. 2016; Gidey et al. 2018). Detecting agricul-
tural drought at sub-basin scale is a prerequisite for better In this research, agricultural droughts were analyzed from
understanding of climate change of the study area, which 1990, 2000 and 2020 dry season Landsat data. Multispectral
is very crucial for policy makers to design climate change and thermal data from Landsat were used to construct nor-
adaptation and mitigation measures. Geospatial data-related malized difference vegetation index (NDVI) and land surface
climate data (VCI, TCI and VHI) helps to produce baseline temperature (LST) data, respectively.
information for mitigation options with relatively large area
coverage (Senamaw et al. 2021). In large coverage areas of
the study, geographic information systems and remote sens-
5 Normalized Difference Vegetation Index
ing play an important role in the analysis of agricultural
(NDVI)
drought (Senamaw et al. 2021).
Several studies were conducted to assess drought, par-
This NDVI was calculated from multi-spectral bands of
ticularly agricultural drought in Ethiopia (Gebrehiwot et al.
Landsat images over all the study period. Band 4 for Land-
2016; Gidey et al. 2018; Gebrechorkos et al. 2018; Temam
sat 5 and 7 and band 5 for Landsat 8 were used to measure
et al. 2019; Senamaw et al. 2021). From these previous
near-infrared bands. As well as band 3 for Landsat 5 and 7
research works, with the exception of Gebrehiwot, the other
and band 4 for Landsat 8 were used to measure red bands
study is focused on the northern parts of Ethiopia. How-
of the Landsat data. The formula for this index is presented
ever, agricultural drought frequently occurs in Gilgel Gibe
in (Eq. 1).
sub-basin, but they do not get the attention of researchers
and policy makers. Therefore, this study attempts to analyze NIR − R
NDVI = (1)
agricultural drought based on VHI, which composed of VCI NIR + R
and TCI from satellite data using geospatial techniques.
The NDVI values range between -1.0 and 1.0. The NDVI
values for health and dense vegetation always range between
0.2 and 0.9 (Bustos and Meza, 2015), whereas rock, water
2 Materials and Methods and barren lands have values less than 0.1 (Fu and Burger,
2015).
2.1 Description of study area

This study was conducted in the Gilgel Gibe sub-basin, 6 Vegetation Condition Index (VCI)
which is located between Oromia National Regional State
and Southern Nations, Nationalities, and People’s Region The VCI is the NDVI-based index that is used to monitor
(SNNPR). Geographically, the study area lies between vegetation status and to calculate drought (Gebrehiwot et al.
­7 0 43′00’’N to 8­ 0 54′30’’N latitude and 3­ 7 0 20′30’’ to 2011; Rojas et al. 2011). Gebrehiwot et al. (2016) stated that
­38026′30’’E longitude. The elevation of the Gilgel Gibe sub- VCI is more suitable for monitoring agricultural drought
basin varies from 971.351 m to 3605.43 m above the sea and provides reliable information for short-term agricultural
level (Fig. 1). The study area covers an area of 8899.7 ­km2. droughts during the crop growing season. As Kogan (1995)
confirmed, VCI was acquired from the NDVI. The VCI data
are derived by Eq. 2.
3 Data source and data types NDVIa − NDVImin
VCI = ∗ 100 (2)
NDVImax − NDVImin
For this study, remote sensing data were acquired from
USGS websites. These were Landsat TM 1990, ETM + 2000 where NDVIa represents NDVI value of current month,
and Landsat OLI/TIRS 2020, all of which were downloaded while NDVImin and NDVImax represent the minimum and
with path of 169 and row 54 during dry season when crops maximum NDVI values, respectively, throughout the period

13
Multiple indices‑based assessment of agricultural drought: A case study in Gilgel Gibe… 457

Fig. 1  Location map of the study area

of observation. Kogan (2002) and Gebrehiwot et al. (2016) with thermal bands of band 6 and Landsat 8 (OLI/TIRS)
reported that agricultural drought was classified into four with thermal bands of band 10 was used (Guo et al. 2020).
based on VCI (Table 1). Digital number (DN) was converted to spectral radiance of
The VCI has been suggested as drought tool (Dutta et al. thermal bands of Landsat TIRS (Dener and Alves, 2016).
2015); however, using VCI was not enough to define drought Consequently, LST was estimated by using the brightness
analysis accurately. To improve, the TCI was developed to temperature of the two bands of thermal infrared, mean and
capture different responses of vegetation to in situ tempera- difference in land surface emissivity (Cheng et al. 2015).
ture as additional information (Risqi et al. 2016). This can
be achieved by calculating LST of Landsat TM, ETM + and
TIRS of 1990, 2000 and 2020, respectively, for drought
monitoring. 8 Step 1: Conversion of digital numbers
to top of atmosphere (TOA) spectral
radiance
7 Retrieval of Land Surface Temperature
Thermal data in the Landsat sensor and deliver a manner of
The LST was calculated from thermal bands of Landsat TM signifying pixels are stored in the digital number that have
of 1990, Landsat ETM + of 2020 and Landsat OLI/TIRS not yet been calibrated and converted into radiance units
of 2020. Thus, LST was retrieved from Landsat 5 and 7 (Aik et al. 2020) by using (Eq. 3).

13
458 M. B. Moisa et al.

Fig. 2  Methodological flowchart of the study

Table 1   VCI-based agricultural drought classification scheme AL = Band-specific additive rescaling factor from the
adapted from Kogan, 2002. metadata (Radiance_Add_Band_x, where x is the band
Category Name Threshold number).
value (%) Q Cal = Quantized and calibrated standard product pixel
values (DN).
Extreme drought 0–10
Severe drought 11–35
Fair vegetation condition 35–50
No drought (optimal vegetation condition) 51–100
9 Step 2: Conversion of Radiance
to Brightness Temperature

After the DNs are converted to radiance, the thermal band’s

spectral radiance should be converted to brightness tempera-
L𝝀 = (ML ∗ QCal) + AL (3)
ture (BT) as indicated in (Eq. 4).
where.
K2
Lλ = Top of atmosphere (TOA) spectral radiance ­(Wm−2 𝐁𝐓 =
(4)
( )
­sr μm−1).
−1 K1
ln L𝝀 + 1
ML = Band-specific multiplicative rescaling factor from
the metadata (RADIANCE_MULT_BAND x, where x is BT = is effective at-sensor brightness temperature (K);
the band number). K2 = is Calibration constant 2.

13
Multiple indices‑based assessment of agricultural drought: A case study in Gilgel Gibe… 459

K1 = is Calibration constant 1. 𝐓𝐁
𝐋𝐒𝐓 = − 273.15
Lλ: is spectral radiance at the sensor's aperture (W/ (8)
( )
1 + 𝛌 𝐓𝐁 𝐈𝐧𝛆
(m2*sr* μm)); and. P

Ln: is natural logarithm.

where.
LST = is land surface temperature (in Kelvin);
TB = is radiant surface temperature (in Kelvin).
10 Step 3: Land surface Emissivity λ = is the wavelength of emitted radiance (11.5 μm).
Estimation Using the NDVI Technique ρ = is h × c/σ (1.438 × ­1 0 −2 mK); h is Planck’s con-
stant (6.26 × 10 − 34 J s); c is the velocity of light
The NDVI was used to categorize the distributions of veg- (2.998 × ­1 0 8 m/s); σ is Stefan Boltzmann’s constant
etation covers and their greenness as well. Consequently, it (1.38 × 10 − 23 J ­K−1); and ε is land surface emissivity.
also explores the transformation of NDVI into values associ- The TCI was calculated using the following equation
ated with cover portion using empirical relationships with (Eq. 9)
vegetation indices, as a possible basis function (Tomar et al.
(LSTmax − LSTa)
2013). The calculation of the NDVI is important to identify TCI = ∗ 100 (9)
LSTmax + LSTmin
the proportion of the vegetation (PV). In addition, emissiv-
ity (ε), which is related to the PV, should also be calculated where.
(Eq. 5). LSTa is the LST value of current month, and LSTmin
and LSTmax denote the minimum and maximum LST val-
NIR − R
NDVI = (5) ues, respectively, calculated from multiyear time series
NIR + R
data.
where NDVI is the normalized difference vegetation index, Consequently, Gidey et al. (2018) repor ted that
NIR is the near-infrared band and R is the red band. due to a lack of more accurate information on the
Landsat 5 and 7 use band 4 (Infrared) and band 3 (Red), inf luence of VCI and TCI on the VHI, the coeffi-
and Landsat 8 data use Bands 5 (Infrared) and 4 (red) to cient of the VHI equation was fixed at 0.5. Finally,
calculate the NDVI values. From the NDVI values acquired, the VHI was calculated to assess both vegetation
the next step in obtaining land surface emissivity (LSE), stress and temperature to evaluate drought severity.
namely the calculation of the proportion of vegetation (Pv), The VHI was calculated by the following equation
was achieved. (Eq. 10)
VHI = aVCI + (1 − a)TCI (10)

11 Step 4: Calculating the Proportion where.

of Vegetation (Pv) VHI = Vegetation health index, a = 0.5 (contribu-
tion of VCI and TCI), VCI = vegetation condition index,
The proportion of vegetation (Pv) is calculated according TCI = temperature condition index.
to equation (Eq. 6) Agricultural drought based on VHI was classified
into five, according to the previous work of (Kogan,
[
𝐍𝐃𝐕𝐈 − 𝐍𝐃𝐕𝐈𝐦𝐢𝐧
]2 2002). For this study, the following VHI classification
𝐏𝐕 = (6)
𝐍𝐃𝐕𝐈𝐦𝐚𝐱 − 𝐍𝐃𝐕𝐈𝐦𝐢𝐧 scheme for agricultural drought monitoring was pro-
posed (Table 2).

12 Step 5: Land Surface Emissivity (LSE)

Assessment

Land surface emissivity (ε) is significant to estimate LST. Table 2  Classification of VHI Drought class VHI in (%)
scheme for drought mapping
According to Sobrino et al. (2004), land emissivity is cal-
Extreme drought < 10
culated using (Eq. 7)
Severe drought 10–20
𝜺 = 0.004 ∗ PV + 0.986 (7) Moderate drought 20–30
Mild drought 30–40
The calculated radiant surface temperatures were cor-
No drought > 40
rected for emissivity using (Eq. 8)

13
460 M. B. Moisa et al.

13 Results and Discussion 13.2 Analysis of Vegetation Condition Index (1990–

2020)
13.1 Normalized Difference Vegetation Index
(1990–2020) Our results show that extreme drought was increased from
0.3 ­k m 2, 0.4 ­k m 2 and 0.7 ­k m 2 in 1990, 2000 and 2020,
The result of the study shows that vegetation cover respectively. Area affected by extreme drought in 1990
declined from 1990 to 2020 due to human action for the was almost null but later increased by 33% and 133% by
expansion of agricultural land in the study area. Accord- the year 2000 and 2020, respectively. Severe drought was
ingly, the NDVI value of the study area in 1990 is about expanded from 70.7 k­ m 2 to 829 ­k m 2 in 1990 and 2020
0.76 and declined to 0.57 and 0.56 by the year 2000 (Table 3). Severe drought was more expanded in 2020 due
and 2020, respectively. Thus, over the last 30 years, the to decline of vegetation cover (Figs. 4 and 5).
NDVI values declined by 0.2. This result is comparable to
Alemu (2019), who reported that the NDVI values over 13.3 Analysis of Land Surface Temperature (1990–
Andassa watershed, Blue Nile Basin. were 0.66 in 1986 2020)
and declined to 0.62 and 0.54 by the year 2000 and 2016,
respectively. Removal of vegetation covers increased In 1990, 2000, and 2020, the LST of the Gilgel Gibe sub basin
from the central part to the peripheral in all study period increased by a mean of 27.1 °C, 30.6 °C, and 33.6 °C, respec-
(Fig. 3). This is the main cause of barren land expansion. tively. This clearly show that the LST of the study area increased

Fig. 3  NDVI map of the study area

13
Multiple indices‑based assessment of agricultural drought: A case study in Gilgel Gibe… 461

Table 3  Table 3 VCI VCI-based Agri- Decadal expansion of agricultural drought- Decadal changes of agricultural
based agricultural drought cultural drought prone area drought-prone area [ compared to in
classification class 1990]
1990 2000 2020 2000 2020
(km2) (%) (km2) (%) (km2) (%) (km2) (%) (km2) (%)

Extreme drought 0.3 0 0.4 0 0.7 0 0.10 33.33 0.4 133.33

Severe drought 70.7 0.8 159.2 1.8 829 9.3 88.50 125.18 758.3 1072.56
Fair vegetation 3832.8 43.1 1557.7 17.5 5867.8 65.9 -2275.10 -59.36 2035 53.09
No drought 4996 56.1 7182.4 80.7 2202.3 24.7 2186.40 43.76 -2793.7 -55.92

Fig. 4  VCI map of the study area from 1990 to 2020

13
462 M. B. Moisa et al.

Fig. 5  VHI map of the study area

by 9.5 °C over the past 30 years. Other studies in Adama zuria and 11.8% by the year 2000 and 2020, respectively. Spatial dis-
in Ethiopia by Tafesse and Suryabhagavan (2019) found that the tributions of severe and moderate drought were increased the
LST increased from 33 °C in 1989 to 39 °C and 42 °C by the year of 2020 (Fig. 6). The findings of the study are more con-
year 1999, and 2016, respectively with an increased by 9 °C over sistent with earlier research, such as (Gebrehiwot et al. 2016;
27 years. The central parts of the study area were dominated by Gidey et al. 2018 and Senamaw et al. 2021). Analysis of VHI
high land surface temperature from 1990 to 2020 (Fig. 5). is crucial for decision and policy makers to design appropriate
mitigation and adaptation strategies to minimize the potential
13.4 Analysis of Vegetation Health Index (1990– impacts of drought.
2020)

Unlike VCI, no extreme drought has been detected by VHI. 14 Conclusions

However, very small areas experienced severe drought in the
year 2020. The results indicated that severe drought struck a 6 By using four drought monitoring indices (NDVI, VCI, TCI
­km2 in 2020 (0.1%). Furthermore, moderate drought increased and VHI), agricultural drought in the study area was analyzed,
by 3.3 k­ m2, 22.2 k­ m2 and 1051.1 k­ m2 in 1990, 2000 and 2020, which will be used by the farming communities and policy
respectively (Table 4). Results show that moderate and mild makers. The results show that severe drought was experienced
drought has been detected by VHI. Moderate drought has been in 2020 in the study area. The removal of vegetation cover is
increased from almost zero in 1990 and later increased to 0.2% the primary cause of the increase in agricultural drought. The

13
Multiple indices‑based assessment of agricultural drought: A case study in Gilgel Gibe… 463

Table 4  Table 4 VHI Decadal expansion of agricultural drought-prone Decadal changes of agricul-
based agricultural drought area tural drought-prone area [
classification compared to in 1990]

VHI-based agri- 1990 2000 2020 2000 2020

cultural drought (km2) (%) (km2) (%) (km2) (%) (km2) (%) (km2) (%)
class
Severe drought 0.0 0.0 0.0 0.0 6.0 0.1 0 0 6 0.1
Moderate drought 3.3 0.0 22.2 0.2 1051.1 11.8 18.9 0.2 1028.9 11.6
Mild drought 581.2 6.5 316.6 3.6 4480.1 50.3 -264.6 -2.9 4163.5 46.7
No drought 8315.3 93.4 8560.8 96.2 3362.4 37.8 245.5 2.8 -5198.4 -58.4

result revealed that the extreme drought has been increased year 2000 and 2020, respectively, while the severe drought
from 0.3 k­ m2 in the year 1990 to 0.4 k­ m2 and 0.7 km 2 in the increased from 70.7 k­ m2 in 1990 to 159.2 k­ m2 and 829 k­ m2

Fig. 6  VHI map of the study area

13
464 M. B. Moisa et al.

by the year 2000 and 2020, respectively. In addition, moder- derived Vegetation Condition Index (VCI) and Standardized Pre-
ate drought was increased by area of 3.3 ­km2, 22.2 ­km2 and cipitation Index (SPI). The Egyptian Journal of Remote Sensing and
Space Sciences 18(1):53–63
1051.1 ­km2 in 1990, 2000 and 2020, respectively. From veg- Fu B, Burgher I (2015) Riparian vegetation NDVI dynamics and its rela-
etation condition index, extreme drought was increased from tionship with climate, surface water and groundwater. J Arid Environ
0.3 ­km2, 0.4 k­ m2 and 0.7 k­ m2 in 1990, 2000 and 2020, respec- 113:59–68. [Link]
tively. Consequently, severe drought was expanded from 70.7 Gebrechorkos SH, Hulsmann S, Bernhofer C (2018) Changes in tempera-
ture and precipitation extremes in Ethiopia, Kenya, and Tanzania.
­km2 to 829 ­km2 in 1990 and 2020. We recommend that con- Int J Climatol 39:18–30
servation of natural resources like vegetation cover is the best Gebrehiwot T, van der Veen A, Maathuis B (2011) Spatial and temporal
option to against agricultural drought in the study area. assessment of drought in the Northern highlands of Ethiopia. Int J
Appl Earth Obs 13:30932
Acknowledgements The authors would like to acknowledge Wollega Gebrehiwot T, Veen Avdand Maathuis B (2016) Governing Agricul-
University Shambu Campus Faculty of Technology and Oda Bultum tural Drought: Monitoring Using The Vegetation Condition Index.
University Institute of Land Administration and Jimma University Col- Ethiopian Journal of Environmental Studies & Management 9(3):
lege of Agriculture and Veterinary Medicine for the existing facilities 354 – 371
to conduct this research. Gidey E, Dikinya O, Sebego R, Segosebe E and Zenebe A (2018) Analy-
sis of the long-term agricultural drought onset, cessation, duration,
frequency severity and spatial extent using Vegetation Health Index
Author contributions MBM involved in research design, data collec- (VHI) in Raya and its environs, Northern Ethiopia. Environ Syst Res
tion, data analysis and draft manuscript. BBM involved in data analysis. 7:13 [Link]
DOG participated in methodology, data analysis and manuscript edi- Gonzalez-Orozco CE, Porcel M, Velasquez DFA, Orduz-Rodriguez O
tion. All authors read and approved the final manuscript. (2020) Extreme climate variability weakens a major tropical agri-
cultural hub. Ecological Indicators 111:106015
Funding No funding received for this research. Guo J, Ren H, Zheng Y, Lu S, Dong J (2020) Evaluation of Land Sur-
face Temperature Retrieval from Landsat 8/TIRS Images before and
Declarations after Stray Light Correction Using the SURFRAD Dataset. Remote
Sensing 12(6):1023
Kogan FN (1995) Application of vegetation index and brightness tem-
Ethics approval N/A perature for drought detection. Adv Space Res 15(11):91–100
Kogan FN (2002) World droughts in the new millennium from
Consent for publication The authors agreed to publish on Theoretical AVHRR-based Vegetation Health Indices. Eos Transaction of
and Applied Climatology. American Geophysical Union 83(48):562–563
Randell H, Gray C (2016) Climate variability and educational attainment:
Code availability N/A. Evidence from rural Ethiopia. Global Environ Change 41:111–123
Rojas O, Vrieling A, Rembold F (2011) Assessing drought probability for
Consent to participate N/A agricultural areas in Africa with coarse resolution remote sensing
imagery. Remote Sens Environ 115:343–352
Data availability Available in this manuscript. Senamaw A, Addisu S, Suryabhagavan KV (2021) Mapping the spatial
and temporal variation of agricultural and meteorological drought
Competing interest The authors declared no competing interest. using geospatial techniques. Ethiopia Environ Syst Res 10:15.
[Link]
Sholihah RI, Trisasongko BH, Shiddiq D, La Ode S, Iman SK, Manijo
References DR, Panuju (2016) Identification of agricultural drought extent
based on vegetation health indices of Landsat data: case of Subang
Aik DHJ, Ismail MH, Muharam FM (2020) Land Use/land Cover and Karawang, Indonesia. Procedia Environ Sci 33:14–20
Changes and the Relationship with Land Surface Temperature Using Sobrino JA, Jimenez-Munoz JC, Paolini L (2004) Land surface tem-
Landsat and MODIS Imageries in Cameron Highlands. Malaysia perature retrieval from LANDSAT TM5. Remote Sens Environ
Land 9:372. [Link] 90:434–440
Alemu MM (2019) Analysis of Spatio-temporal Land Surface Tempera- Sultana MSH, Gazi Md, Yousuf MM, Bodruddoza (2021) Multiple indi-
ture and Normalized Difference Vegetation Index Changes in the ces based agricultural drought assessment in the northwestern part
Andassa Watershed, Blue Nile Basin. Ethiopia Journal of Resources of Bangladesh using geospatial techniques. Environmental Chal-
and Ecology 10(1):77–85 lenges. [Link]
Bustos E, Meza FJ (2015) A method to estimate maximum and mini- Tafesse B, Suryabhagavan KV (2019) Systematic modeling of impacts
mum air temperature using MODIS surface temperature and veg- of land-use and land-cover changes on land surface temperature
etation data: application to the Maipo Basin. Chile Theoretical and in Adama Zuria District. Ethiopia Modeling Earth Syst Environ
Applied Climatology 120(1–2):211–226. [Link] 5:805–817. [Link]
s00704-​014-​1167-2 Temam D, Uddameri V, Mohammadi G, Hernandez EA, Ekwaro-Osire
Cheng CH, Nnadi F, Liou YA (2015) A regional land use drought index S (2019) Long-Term Drought Trends in Ethiopia with Implications
for Florida. Remote Sens 7(12):17149–17167. [Link] for Dryland Agriculture. Water 11:257
3390/​rs712​15879 Tomar V, Kumar P, Rani M, Gupta G, Singh J (2013) A satellite-based
Dener E, Alves L (2016) Seasonal and spatial variation of surface urban biodiversity dynamics capability in tropical Vegetation. Electron
heat island intensity in a small urban agglomerate in Brazil. Climate J Geotech Eng 18:1171–1180
4(4):61. [Link]
Dutta D, Kundu A, Patel NR, Saha SK, Siddiquid AR (2015) Assessment Publisher's note Springer Nature remains neutral with regard to
of agricultural drought in Rajasthan (India) using remote sensing jurisdictional claims in published maps and institutional affiliations.

13
Theoretical & Applied Climatology is a copyright of Springer, 2022. All Rights Reserved.