Analysis of Drought in the Tigrigra Plain (Azrou Region, Middle Atlas):

Integrated Approach of AHP and GIS for Evaluating Remote Sensing and
Meteorological Indices
1 1
FangourAnass* ,SelamiMehdi
1
Departement of Geology, Faculty of Sciences Ain Chock, University Hassan II,
Casablanca 20100, Morocco.
1
* Fangouranass10@[Link]

ABSTRACT: This study aims to harness spatial remote sensing (RS) technologies, particularly by integrating the
Analytical Hierarchy Process (AHP) with a Geographic Information System (GIS) and various influential parameters, for
drought monitoring. We assessed the spatial extent of drought in the Tigrigra watershed using Landsat 5, Landsat 7, and
Landsat 8 datasets for each season over a 20-year period (from 2000 to 2022). Additionally, we aimed to characterize the
spatiotemporal model of drought conditions using the Standardized Precipitation Index (SPI), Vegetation Condition Index
(VCI), Temperature Condition Index (TCI), and Vegetation Health Index (VHI). The results reveal episodes of drought ranging
from severe to extreme in 2003, 2006, 2007, 2008, 2012, 2016, 2019, and 2021, with specific stations experiencing
intensified drought conditions at different time scales. Particularly, the VCI indicates that only 20% of the total area is free
from drought, while the TCI highlights an incidence of drought affecting 50% of the northeast part. Similarly, the VHI
indicates an incidence of drought affecting 70% of the northeast zone. However, after processing all the indices, it is
necessary to determine a reliable value that aligns with all the selected indices. To this end, we used the AHP method to
obtain reliable values aligned with the selected indices. The southeastern and southwestern regions exhibit extreme and
severe drought, contrasting with the absence of drought in the northeast and northwest zones.

KEYWORDS: Analytic Hierarchy Process; Geographic Information System; Remote Sensing; Drought

INTRODUCTION
One of the most severe meteorological (Shafer, 1982), the Streamflow Drought
disasters is drought, which severely impacts Index (Nalbantis et Tsakiris, 2009), the
the economy, agriculture, natural resources, Standardized Hydrological Index (Sharma et
society, and the environment.[1]Drought is Panu, 2010), the Standardized
a complex issue involving a significant Evapotranspiration Index (Vicente-Serrano
reduction in available water resources due et al., 2010), the Agricultural Reference
to prolonged lack of Drought Index (Woli et al., 2012), among
precipitation.[4]Droughts can be quantified others. Several indices have been used for
using several hydrometeorological drought characterizing hydrological droughts and
indices, the most popular of which include: generally require extensive data and
the Palmer Index (Palmer, 1965), the calculations, unlike very simple and effective
Standardized Precipitation Index (Mckee et meteorological drought indices such as the
al., 1993), the Surface Water Supply Index
Standardized Precipitation Index (Nalbantis et Tigrigra watershed for each four-year season
Tsakiris, 2009).Due to the scattered (from 2000 to 2022) using Landsat 5, Landsat
distribution of in situ meteorological stations 7, and Landsat 8 datasets; (2) characterizing
per unit area and the resulting uncertainties, it the spatiotemporal pattern of drought
is challenging to accurately estimate drought, conditions using SPI, VCI, TCI, and VHI
particularly in arid and hyper-arid regions. indices. This research will be valuable for
Natural barriers such as mountains and sand urban planners and environmental scientists,
dunes fields can also contribute to this issue enabling them to make informed decisions
(Hereher et al., 2020). However, by the end of and implement policies aimed at mitigating
the 20th century, advancements in remote drought in the Tigrigra watershed.
sensing technologies and earth observation,
Site description
such as the launch of the Landsat series by the
National Aeronautics Space Administration The watershed of the Tigrigra River is delineated
(NASA) in 1972, have led to advancements in on a topographic background. It is located at the
drought monitoring (West et al., 2019). confluence of the Central Middle Atlas with the
Furthermore, the increasing temperatures due central plateau. It extends between parallels
33°36 and 33°12 N and longitudes 5°34 and 5°06
to climate change have sparked growing
W. The Tigrigra watershed has an area of 1109
interest and better understanding of its
km2 and a perimeter of 211 km. It is situated at
impacts. This has led geospatial science the confluence of the Central Middle Atlas with
experts to consider remote sensing as an the central plateau. The watershed presents
essential tool for providing the necessary data varied relief, with altitudes ranging between 785
for assessing ecosystem conditions and m and 2196 m. Its elongated shape follows a
monitoring extreme climate changes both NW-SE orientation, and it is delimited by:
spatially and temporally (Sarkar et al., 2020).
• To the Northwest by the causse of Agouray and
The Analytical Hierarchy Process (AHP) is a Adarouch, which runs eastwards towards the
theoretical method that establishes priority confluence of Tigrigra with the Beht River;
scales using pairwise comparisons, expert • To the Northeast by the causse of El Hajeb and
opinions, and previous research. It is an the causse of Ifrane;
effective method for decision-making that
• To the East by Michlifène, Jbel Hebri;
considers multiple criteria. The Analytical
Hierarchy Process (AHP) method is more • To the South by the causse of Ain Leuh and the
effective when combined with GIS Ifrane River, which runs towards its confluence
(Geographic Information Systems).[1] with the Tigrigra River before pouring into the
Beht [Link].1
In order to address the lack of data and
limited use of RS and GIS technologies in
drought monitoring, this study aims to create
a drought risk map by integrating influential
factors using the AHP approach and GIS
techniques. The objectives include: (1)
assessing the spatial extent of drought in the
 Weather station
Tigrigra watershed

Fig 1. Location of the site.

Methods
Five conditioning parameters were indices. The methodology of this study
used to determine drought in the study can be divided into several stages, such
areas: Normalized Difference as acquiring meteorological data,
Vegetation Index (NDVI), land surface collecting satellite data, and
temperature (LST), Temperature monitoring meteorological drought
Condition Index (TCI), Vegetation using the SPI index. Additionally,
Condition Index (VCI), and Vegetation monitoring the state of drought is
Health Index (VHI). The Analytical done using drought indices derived
Hierarchy Process (AHP) was employed from remote sensing such as VCI, TCI,
to perform this process. To make the and [Link] 2
raw data usable, preprocessing is
1. Analytical hierarchy process
conducted before generating drought
A pairwise matrix of conditional variables was The relative weight matrix was divided by the sum
created using the Analytical Hierarchy Process of column values to compute the estimation of
(AHP), a well-known method for solving complex the normalized eigenvector based on the
problems related to drought vulnerability. To comparison matrix. The normalized eigenvector
enable pairwise comparison, each parameter was obtained by averaging the rows, and then
was hierarchically organized, and its relative the effect percentages for each thematic layer
importance was assessed on a scale from 1 to 9. were calculated using this vector. The
The value of each unit, with elements equal to 1, consistency ratio (CR) was used to determine the
is compared in a 6 x 6 diagonal pairwise accuracy of the relationship. A ratio less than or
reference matrix. The relative importance of the equal to 0.1 indicates an appropriate reciprocal
remaining variables is represented by the matrix, while a ratio greater than 0.1 indicates
remaining values in each row. The normalized that the matrix should be modified. It is
eigenvector, pairwise comparison matrix, and determined as follows: CR = CI/RI (1), where CI is
relative weight matrix were used to determine the consistency index and RI is the random
the rank of each parameter in Table 1. consistency index.

Multi-temporal imagery (2000-2022) Daily meteorological data (2000-2022)

Data Acquisition
Landsat 5 Landsat 7 ETM+ Landsat 8

Pre-processing Atmospheric correction and clipping

NDVI LST

Indexes calculation VCI TCI

VHI SPI

Preparation of Thematic Layer

Drought Mapping
Analytical Hierachy Process

Drought Mapping

Fig 2: Methodology adopted in this study

 Table 1 : Ranking of Parameters in Drought Vulnerability Assessment
Parameters NDVI LST VCI TCI VHI SPI

NDVI 1 4 2 3 0,5 0,25

LST 0,25 1 0,33 0,5 0,17 0,12

VCI 0,5 3 1 2 0,25 0,17

TCI 0,33 2 0.5 1 0,33 0,14

VHI 2 6 4 3 1 0 ,5

SPI 4 8 6 7 2 1

Number of comparisons = 15,Consistency Ratio CR = 2.2%,Principal eigen value = 6.137

𝐶𝐼 = where, k max is the principal eigen value

and n is the number of comparisons.

1.1 Acquisition of meteorological data past two decades (2000-2022) at 4-year

This study used monthly precipitation data intervals (i.e., 2000, 2004, ..., and 2022).

from the Sebou Basin Hydraulic Agency Individual satellite images were collected

(ABHS) and global meteorological data from for each season from 2000 to 2022.

the Soil and Water Assessment Tool. (Global Twenty-eight Landsat images from Path

Weather Data for Swat) [2]. The data is from 201 and Row 37 were downloaded at Level
1 via the USGS Earth Explorer website.
eight meteorological stations. (Agouai; Ain
([Link] The
Louh, Ait Aissa, Azrou, El Hajeb, Ifrane, Sidi El
details of the data collected from the USGS
Moukhfi and Souk Had Oued Ifrane) The data
archive for the years 2000, 2004, 2008,
from eight meteorological stations located in
2012, 2016, 2020, and 2022 are presented
the Tigrigra plain were collected over a period
in Table 2. Landsat data, including
spanning from 1982 to 2020. (Table 2).
Thematic Mapper (TM), Enhanced
1.2 Acquisition of satellite data Thematic Mapper (ETM+), and Thermal
This study includes a multi-step analysis to
determine drought conditions over the
Infrared Sensor (TIRS), were obtained from techniques such as radiometric and
the USGS. The Universal Transverse Mercator atmospheric correction were applied to the
(UTM) projection system for Zone 30 North images to improve their quality and precision
and the WGS-1984 reference system were before analysis. These Landsat 5, Landsat 7,
applied to the Landsat data. and Landsat 8 images are corrected to remove

The indices used in this study include the atmospheric effects and normalize the data.

Normalized Difference Vegetation Index Subsequently, the indices are calculated to

(NDVI), Land Surface Temperature (LST), assess vegetation density, changes in plant

Vegetation Condition Index (VCI), Temperature health, and to monitor drought conditions

Condition Index (TCI), and Vegetation Health using satellite imagery.

Index (VHI). Advanced preprocessing

Table 2 : Satellite images used to calculate the various indices.

Satellite image sensor spatial resolution Acquisition date

31-03-2000
06-08-2000
2000
23-09-2000
26-11-2000
Landsat 5 ETM 30 m
03-04-2004
25-08-2004
2004
13-11-2004
31-12-2004
17-01-2008
29-03-2008
2008
20-08-2008
24-11-2008
16-03-2012
23-08-2012
2012
08-09-2012
29-12-2012
Landsat 7 ETM+ 30 m
14-05-2016
17-07-2016
2016
21-10-2016
24-12-2016
07-04-2020
13-08-2020
2020
16-10-2020
17-11-2020
29-04-2022
19-08-2022
Landsat 8 OLI+ 30 m 2022
14-10-2022
25-12-2022
 1.3 Standardized Precipitation Index (SPI) The Standardized Precipitation Index (SPI) is a
very simple index created by Mckee et al.,

A monthly rainfall database was established for (1993). It is a reference index recommended by

calculating the SPI index in the study area. The the World Meteorological Organization (WMO)

database used in this work covers a period of thirty- in 2009 to facilitate drought monitoring and

five years, from 2000 to 2022 (Table 2). climate risk management. The SPI is a normalized
monthly indicator based on the probability of
The majority of the collected data comes from the
precipitation (P) occurring, regardless of the time
Sebou Basin Hydraulic Agency (ABHS), while the
frame considered. It is mathematically expressed
remaining data is extracted from global
as follows:
meteorological data provided by the Soil and Water
Assessment Tool. (Global Weather Data for Swat) [2]. Pi  Pm
SPI 

For this study, we considered eight (08) weather
With Pi representing the rainfall of the month or
stations, namely Agouai, Ain Louh, Ait Aissa, Azrou,
year i, Pm representing the average rainfall of
El Hajeb, Ifrane, Sidi El Moukhfi, and Souk Had Oued
the series over the considered time scale, and σ
Ifrane (Fig 3)
representing the standard deviation of the series
over the considered time scale.

The SPI can be estimated at different time

scales (1, 3, 6, and 12 months) for each
station. SPI-1 is calculated using monthly
precipitation data. SPI-3 is calculated by
taking the average over three months
(moving average of precipitation over three
months). Similarly, other indices for 6 and 12
months can be calculated. The SPI-1 index is
useful for studying short-term variations in
Tigrigra watershed
Precipitation for the period 2000-2022 (mm)
the frequency and severity of droughts, SPI-3
and SPI-6 indices are generally used to
monitor seasonal variations in drought, while

Fig 3: Average yearly rainfall (2000-2022) at studied stations. the SPI-12 index is useful for studying the
annual trend of drought. The severity of a and SPI-6 indices are generally used to
drought can be estimated using the index, monitor seasonal variations in drought, while
which compares actual precipitation to the the SPI-12 index is useful for studying the
amount of water lost through evaporation annual trend of drought. The severity of a
and transpiration over a given period drought can be estimated using the index,

The SPI can be estimated at different time which compares actual precipitation to the

scales (1, 3, 6, and 12 months) for each amount of water lost through evaporation

station. SPI-1 is calculated using monthly and transpiration over a given period.

precipitation data. SPI-3 is calculated by Drought occurs when the SPI is consecutively
taking the average over three months negative and its value reaches an intensity of
(moving average of precipitation over three -1 or lower, and it ends when SPI becomes
months). Similarly, other indices for 6 and 12 positive. A classification of drought is
months can be calculated. The SPI-1 index is performed based on the values of this index
useful for studying short-term variations in (Table 4).
the frequency and severity of droughts, SPI-3

Table 1 : Classification of drought based on the Standardized Precipitation Index (SPI)

value (Mckee et al. (1993)):

SPI class Drought severity level

SPI ≥ 2 Extremely Wet

1,5 ≤ SPI ≤ 1,99 Very Wet
1 ≤ SPI ≤ 1,49 Moderately Wet
−0,99 ≤ SPI ≤ 0,99 Near Normal
−1,49 ≤ SPI ≤ −1 Moderately Dry
−1,99 ≤ SPI ≤ −1,5 Very Dry
SPI ≤ −2 Extremely Dry
 Remote Sensing Derived Indices to any adverse temperature change. The
1.3.1 Vegetation Condition Index (VCI) following formula illustrates the calculation
The Vegetation Condition Index (VCI) is an of TCI (Kogan, 1995):
index used in remote sensing to assess the LST − LST
TCI (i) = × 100
health and vigor of vegetation in a given area. LST − LST

It is primarily employed to monitor and a. The minimum values LSTmin and LSTmax indicate

evaluate vegetation conditions, particularly the minimum and maximum values of the LST (Land

during periods of drought or water stress Surface Temperature) in the region. TCI values
range from 0 to 100, indicating either stress or good
(Kogan, 2001). This index is calculated using
thermal conditions for vegetation, as shown in Table
the equation
5.
()
VCI (i) = × 100
b. The following steps are used to calculate the LST
(Land Surface Temperature) (Wukelic et al., 1989).
The values of VCI range from 0 to 100, where
Lλ = M × Q +A
VCI values below 40 indicate drought
c.
conditions in the region (as presented in
Where Lλ is the spectral radiance at the top
Table 5) [45, 55-57]. The NDVI is calculated
of the atmosphere (TOA) in
using red and near-infrared (NIR) bands and
Wm−2sr−1mm−1, ML is the band-specific
is given by (Tucker, 1979).
multiplicative rescaling factor from metadata
NIR − RED
NDVI = (RADIANCE_MULT_BAND_x, where x is the
NIR + RED
band number), AL is the band-specific
Temperature Condition Index (TCI)
additive rescaling factor from metadata
The Temperature Condition Index (TCI) is a
(RADIANCE_ADD_BAND_x, where x is the
drought index derived from satellite imagery
band number), and QCAL is the quantized
that is used to assess plant stress caused by
temperatures and excessive humidity. TCI is and calibrated standard product of pixel

an important index for monitoring drought values (DN). All these variables can be

as it provides information on weather extracted from the metadata file of satellite

conditions that can affect plant health and images.
agricultural production (Singh et al., 2003). It
d. The conversion of TOA temperature to
is related to the vegetation's responsiveness
radiance (Brightness Temperature) can be
e. calculated using the following equation Planck constant (6.626 x 10^−34 Js); and c

(Cao et al., 2008; Carlson et al., 1997): is the speed of light in a vacuum (2.998 ×
10^8 m/s).
K
BT = − 273,15 1.3.2 Indice de santé de la végétation (VHI)
K
ln L + 1

[Link] K1 is the band-specific thermal The following equation is used to calculate

conversion constant from metadata the Vegetation Health Index (VHI), which is

(K1_CONSTANT_BAND_x, where x is the the weighted sum of the Vegetation

thermal band number); K2 is the band-specific Condition Index (VCI) and the Temperature
Condition Index (TCI), providing valuable
thermal conversion constant from metadata
information regarding vegetation stress due
(K2_CONSTANT_BAND_x, where x is the
to droughts. Droughts based on VHI are
thermal band number).
classified into five categories, following the
g. The vegetation proportion (Pv) is calculated recommendations of (Kogan et al., 2001).
using the following equation (Sobrino et al., 2004) Table 5 shows the different drought
conditions based on VCI, TCI, and VHI values.
NDVI − NDVI
P =
NDVI − NDVI VCI + TCI
VIH =
2
h. The emissivity is calculated using the
following equation (Weng et al., 2004).
Tableau 2 : Different drought categories for
ε = 0,004 × P + 0,986 the indices used

i. Finally, the LST (Land Surface Temperature) was Class VHI/VCI/TCI Degree of Drought
calculated using the following equation (Kogan et al.,
0 à 10 Extreme Drought
1997).
10 à 20 Severe Drought
BT 20 à 30 Moderate Drought
LST = 30 à 40 Mild Drought
λ × BT
1+ ρ × lnε
Plus de 40 No Drought
Where λ is the effective wavelength (10.9 μm for
band 10 in Landsat 8 data) and ε is the emissivity;

ρ: h x c/s=1.4388 x 10−2 m K ; s: Boltzmann

constant (1.38 x 10^−23 J/K); h is the
 Results and discussion
months) at the eight studied stations from
2000 to 2022 in the Tigrigra watershed. Fig 4
1.4 Evaluation of drought indices
shows that the most severe and extreme
The meteorological drought index (SPI) at
different time scales, namely 1, 3, 6, and 12 drought episodes are observed by SPI-12,
months, along with remote sensing-derived followed by SPI-6, while SPI-3 and SPI-1
indices (i.e., VCI, TCI, and VHI), are utilized to
indicate drought-free conditions. However,
analyze the severity/condition of drought
both on the spatial and temporal scale from significant fluctuations are observed in the
2000 to 2022 in the Tigrigra watershed. time series of SPI-1 and SPI-3 compared to
SPI-6 and SPI-12. Furthermore, the analysis of
1.5 Standardized Precipitation Index (SPI)
Fig 4 and Fig 5 depict the time series of SPI at the chronological series shows severe to

different time scales (i.e., 1, 3, 6, and 12 extreme drought episodes in 2000, 2003,
2007, 2009, 2014, 2019, and 2021 (Fig 4).

Fig 4 : Analysis of the time series of stations in the Tigrigra watershed at different temporal
scales of SPI (1, 3, 6, and 12 months)
 Overall, the time series curve presented in 2008, and 2019. Similarly, severe to extreme
Figure 10 shows more severe to extreme drought episodes have been observed at the
drought episodes in 2003, 2006, 2007, 2008, Azrou station in 2008, 2019, and 2021. Overall,
2012, 2016, 2019, and 2021. Additionally, the results (Fig 5) demonstrate a significant
drought at the Ain Louh station is more severe number of extreme and severe drought
at the 1, 12, and 6-month scales, followed by episodes in 2003, 2007, 2008, 2019, and 2021 in
SPI-3. Extreme and severe drought episodes at the Tigrigra watershed.
the Ait Aissa station are observed in 2000, 2006,

Fig 5 : Analysis of the time series of stations in the Tigrigra watershed at different temporal
scales of SPI (1, 3, 6, and 12 months)

1.6 Vegetation Condition Index (VCI) conditions during the study period. In
contrast, wetter conditions (and thus
Maps A to J provide the annual average of
relatively less drought) are observed in
maps 1 to 28, illustrating in Fig 6 the
2000, 2016, and 2022. Furthermore, it is
spatial and temporal distribution of the
noteworthy that drought intensified
Normalized Difference Vegetation Index
relatively more in the southern part of
(NDVI) from 2000 to 2022 in the Tigrigra
the watershed than in its northwest part.
watershed. These calculations are based
on the NDVI presented in Annex A. These
maps reveal that the years 2004, 2008,
2012, and 2020 were marked by drought
1) 2) 3) 4) A)
( ( ( ( (

5) 6) 7) 8) B)
( ( ( ( (

9) 10) 11) 12) C)

( ( (

13) 14) 15) 16) D)

( ( ( ( (

17) 18) 19) 20) E)

( ( ( ( (

21) 22) 23) 24) F)

( ( ( ( (

25) 26) 27) 28) J)

( ( ( ( (

Fig 6: Temporal and spatial distribution of the Vegetation Condition Index.

 Fig 7: Average VCI from 2000 to 2022

According to Fig 7, which represents the temperature exceeds 40°C and where
average of all VCI maps from 2000 to annual precipitation is generally reduced.
2022, we find that only 20% of the total TCI shows a spatial distribution trend less
area shows a lack of drought. similar to that of VCI, meaning TCI is more
severe in the northwest part than in the
1.7 Temperature Condition Index (TCI)
south, except for a few years (2000, 2012,
and 2016).
Maps A to J provide the annual average of
maps 1 to 28, illustrating in Fig 8 the The spatial distributions of VCI and TCI

spatial distribution of the Temperature differ for certain periods, such as the years

Condition Index (TCI) in the Tigrigra 2000, 2012, and 2016, which is mainly

watershed over a 4-year interval from explained by the different perspectives of

2000 to 2022, calculated from satellite these indices in describing drought. VCI is

images. The minimum TCI values are calculated from NDVI and only considers

observed in 2000, 2008, 2012, 2016, and the vegetation factor, while TCI is

2020, precisely in the northwest part of calculated from LST and is an integrated

the study area. TCI has become more result of many factors, including

severe from 2008 to 2016, especially in vegetation, precipitation, topography,

the northwest part, where the maximum altitude, soil, and meteorology.
1) 2) 3) 4) A)
( ( ( ( (

5) 6) 7) 8) B)
( ( ( ( (

9) 10) 11) 12) C)

( ( (

13) 14) 15) 16) D)

( ( ( ( (

17) 18) 19) 20) E)

( ( ( ( (

21) 22) 23) 24) F)

( ( ( ( (

25) 26) 27) 28) J)

( ( ( ( (

Fig 8: Temporal and spatial distribution of the Temperature Condition Index

 Fig 9: Average TCI from 2000 to 2022

According to Fig 9, which represents the drought years in the studied watershed.
average of all TCI maps from 2000 to However, it is worth noting that drought
2022, we find that the northeast part is
estimates using VHI show that drought
affected by drought, with 50% of the
area impacted. has become severe in recent years.
Furthermore, the results also show that
drought is severe in the northwest part.
1.8 Vegetation Health Index (VHI)
VHI is a more comprehensive drought

Maps A to J provide the annual average of index than VCI and TCI. VHI indicates that

maps 1 to 28, illustrating in Fig 10 The the areas in the northwest part of the

spatial distribution of the Vegetation watershed are more vulnerable to

Health Index (VHI) in the Tigrigra drought. Therefore, it is advisable to

watershed, obtained respectively from design robust mitigation plans to cope

Landsat satellite images, is presented in with the negative impacts of drought on

Figures 25 through 31. Analysis of the available water resources and agriculture

results shows that 2000, 2008, 2012, in the northwest part of the Tigrigra

2016, 2020, and 2022 are the most severe watershed.

1) 2) 3) 4) A)
( ( ( ( (

5) 6) 7) 8) B)
( ( ( ( (

9) 10) 11) C)
( ( (

13) 14) 15) 16) D)

( ( ( ( (

17) 18) 19) 20) E)

( ( ( ( (

21) 22) 23) 24) F)

( ( ( ( (

25) 26) 27) 28) J)

( ( ( ( (

Fig 10: Temporal and spatial distribution of the Vegetation Health Index
 Fig 11: Average VHI from 2000 to 2022

According to Fig 11, which depicts the method presented in Table 1, which
average of all VHI maps from 2000 to represents the intensity of each index,
2022, we find that the northeast part is and we find the result appearing in
affected by drought, with 70% of the Figure 12. In the southern-eastern and
area impacted. However, after southern-western parts, we observe
processing all the indices, we need to extreme and severe drought, while in
determine a reliable value that adapts the northeastern and northwestern
to all the selected indices. For this zones, there is no drought
purpose, we proceed to use the AHP
.
igra:: 2000-2022 – AHP Analysis
Drought Intensity in Tigrigra

Fig 12: Drought Intensity in Tigrigra from 2000 to 2022: AHP Analysis
In conclusion, the analysis results This study evaluated the spatial and

reveal a temporal variety of drought temporal distribution of droughts,

with periods of varying intensity. After including meteorological drought

estimated by SPI (SPI-1, SPI-3, SPI-6, and
employing the Analytical Hierarchy
SPI-12) and drought indices retrieved by
Process (AHP) method to obtain a
remote sensing, including VCI, TCI, and
reliable overall value, we found that
VHI. Additionally, analyses were
the southeastern and southwestern
conducted on the Tigrigra watershed
regions exhibited a predominance of
with a limited number of in situ
extreme and severe drought, while meteorological stations from 2000 to
the northeastern and northwestern 2022.
regions had no drought. However,
Our main conclusions are presented
despite these variations, there is a below.
general trend towards improvement, (1) The results of the SPI index showed that
with a decrease in the proportion of a significant number of severe and extreme
areas affected by moderate drought drought episodes were observed in 2000,
and an increase in the proportion of 2003, 2007, 2009, 2014, 2019, and 2021.
drought-free areas. These findings Significant fluctuations in drought severity
suggest an improvement in are observed at long-term time scales (i.e.,
vegetation health in the studied SPI-6 and SPI-12) compared to SPI-1 and
region over time. Therefore, it is SPI-3. However, interesting observations
crucial to continue monitoring and are made at the Ain Louh station where
implementing appropriate measures drought is more severe at shorter time
to maintain this positive trend and scales (SPI-1) than at longer time scales.
address drought variations in the Overall, the results have shown significant
region. regional variations in drought severity due
to climate change.
(2) According to the results of the VCI index,
CONCLUSION
the Tigrigra watershed experienced
 droughts in 2004, 2008, 2012, and 2020. In conclusion, droughts have been
The droughts were more intense, with monitored using both in situ data
minimum VCI values (ranging from 0 to 30) obtained through remote sensing
frequently observed in the northwest part techniques. Contrasting results are
of the area. observed for VCI and TCI, indicating

(3) According to the TCI index, drought more severe droughts in the

became more severe from 2008 to 2016, southeast than in the southwest.

especially in the northwest part. TCI values Furthermore, remote sensing

are more severe in the northwest part than technology has proven to be very

in the south, except for a few years (2000, useful in regions where

2012, and 2016). meteorological stations and data

availability are limited. Therefore, the
The Vegetation Health Index (VHI) has shown
results of this study are essential for
that the northwest part of the Tigrigra
understanding the nature of droughts
watershed is prone to drought, especially in
2000, 2008, 2012, 2016, 2020, and 2022. The in the studied region. The study

spatial distribution of the drought index has results provide useful information for

highlighted episodes of moderate drought in addressing local and regional drought

2000, 2008, 2012, 2016, and 2020. This index issues and for designing drought
varies from 10 to 30, indicating episodes of mitigation plans.
moderate drought.
References
[1] Koem, C., Nusit, K., & Tantanee, S. (2022). "Spatial Distribution of Drought Hazard
Mapping Based on AHP and GIS in Kampong Speu Province.", 442-450.
Cao, L., Li, P., Zhang, L., & Chen, T. (2008). Remote sensing image-based analysis of the
relationship between urban heat island and vegetation fraction. The International
Archives of the Photogrammetry, Remote Sensing and Spatial Information
Sciences, 37, 1379-1384.
Carlson, T. N., & Ripley, D. A. (1997). On the relation between NDVI, fractional vegetation
cover, and leaf area index. Remote sensing of Environment, 62(3), 241-252.
Hereher, M., & El Kenawy, A. M. (2020). Exploring the potential of solar, tidal, and wind
energy resources in Oman using an integrated climatic-socioeconomic
approach. Renewable Energy, 161, 662-67.
Kogan, F. N. (1995). Application of vegetation index and brightness temperature for drought
detection. Advances in space research, 15(11), 91-100.
Kogan, F. N. (1997). Global drought watch from space. Bulletin of the American
Meteorological Society, 78(4), 621-636.
Kogan, F. N. (2001). Operational space technology for global vegetation assessment. Bulletin
of the American meteorological society, 82(9), 1949-1964.
Kogan, F. N. (2001). Operational space technology for global vegetation assessment. Bulletin
of the American meteorological society, 82(9), 1949-1964.
McKee, T. B., Doesken, N. J., & Kleist, J. (1993, January). The relationship of drought
frequency and duration to time scales. In Proceedings of the 8th Conference on
Applied Climatology (Vol. 17, No. 22, pp. 179-183).
Nalbantis, I., & Tsakiris, G. (2009). Assessment of hydrological drought revisited. Water
resources management, 23, 881-897.
Palmer, W. C. (1965). Meteorological drought (Vol. 30). US Department of Commerce,
Weather Bureau.
Sarkar, A., Kumar, V., Jasrotia, A. S., Taloor, A. K., Kumar, R., Sharma, R., ... & Roy, S. (2020).
Spatial analysis and mapping of malaria risk in dehradun city India: A geospatial
technology-based decision-making tool for planning and management. Geoecology of
Landscape Dynamics, 207-221.
Shafer, B. A. (1982). Developemnet of a surface water supply index (SWSI) to assess the
severity of drought conditions in snowpack runoff areas. In Proceedings of the 50th
Annual Western Snow Conference, Colorado State University, Fort Collins, 1982.
Sharma, T. C., & Panu, U. S. (2010). Analytical procedures for weekly hydrological droughts: a
case of Canadian rivers. Hydrological Sciences Journal–Journal des Sciences
Hydrologiques, 55(1), 79-92.
Singh, R. P., Roy, S., & Kogan, F. (2003). Vegetation and temperature condition indices from
NOAA AVHRR data for drought monitoring over India. International journal of remote
sensing, 24(22), 4393-4402.
Sobrino, J. A., Jiménez-Muñoz, J. C., & Paolini, L. (2004). Land surface temperature retrieval
from LANDSAT TM 5. Remote Sensing of environment, 90(4), 434-440.
Tucker, C. J. (1979). Red and photographic infrared linear combinations for monitoring
vegetation. Remote sensing of Environment, 8(2), 127-150.
Vicente-Serrano, S. M., Beguería, S., & López-Moreno, J. I. (2010). A multiscalar drought
index sensitive to global warming: the standardized precipitation evapotranspiration
index. Journal of climate, 23(7), 1696-1718.
Weng, Q., Lu, D., & Schubring, J. (2004). Estimation of land surface temperature–vegetation
abundance relationship for urban heat island studies. Remote sensing of
Environment, 89(4), 467-483.
West, H., Quinn, N., & Horswell, M. (2019). Remote sensing for drought monitoring & impact
assessment: Progress, past challenges and future opportunities. Remote Sensing of
Environment, 232, 111291.
Woli, P., Jones, J. W., Ingram, K. T., & Fraisse, C. W. (2012). Agricultural reference index for
drought (ARID). Agronomy Journal, 104(2), 287-300.
Wukelic, G. E., Gibbons, D. E., Martucci, L. M., & Foote, H. P. (1989). Radiometric calibration
of Landsat Thematic Mapper thermal band. Remote sensing of environment, 28, 339-
347.
USGS: [Link]
USGS, 2016. Landsat 4-5 TM and Landsat 7 ETM+ bands and their uses:
[Link]
uses
[2] : [Link]