Multitemporal Remote Sensing for Invasive Prosopis Juliflora Plants
Mapping and Monitoring: Sharjah, UAE
Alya AlMaazmi and Rami Al-Ruzouq
University of Sharjah, University City Road, Sharjah, U.A.E.
Keywords: Prosopis Juliflora, Spatial Analysis, Multitemporal Landsat, NDVI, Support Vector Machine.
Abstract: Prosopis juliflora is one of the 'world's most invasive trees that negatively affects native species and their
ecosystems. The main obstacle for controlling Prosopis juliflora pervasion is to accurately map location as
well as the distribution pattern. Locating Prosopis juliflora is a strategic priority of countries to preserve the
invaded local environment. Recent advances in remote sensing, geographic information system (GIS), and
Machine Learning (ML) techniques provide valuable tools for producing tree distribution maps. In this
research, a supervised classification method with Support Vector Machine (SVM) supported by GIS statistical
analysis was developed to map Prosopis juliflora and their pattern analysis in Sharjah, one of the major cities
in the United Arab. More than 5000-pixel labels taken from Landsat-7 and Landsat-8 imagery were used to
train object-based Support Vector Machine to map Prosopis juliflora. The suggested algorithm resulted in
75% accuracy compared to ground truth samples. Furthermore, multi-temporal detection showed 'that's spatial
clustering pattern of the trees is changing and increasing over time. The approach adopted in this study can
be applied to any other location globally.
1 INTRODUCTION
Invasive species are organisms that are non-native
and have been introduced due to their environmental
and economic benefits, such as providing habitat for
native plants and promote biodiversity due to global
biotic environmental change (Essl et al., 2020).
Nevertheless, the introduction of new species is not
always successful. New species can establish
themselves and dramatically spread in their new
environment and become invasive (Jeschke et al.,
2014). These invasive 'aliens' often reflect negative
impacts, including reduction of grazing areas,
reduction of crop yield, and risk of threat to
biodiversity (Mwangi & Swallow, 2005).
An ideal example of invasive species is Prosopis
juliflora, which is considered one of the world's worst
woody invasive plants (Berhanu & Tesfaye, 2006).
Prosopis juliflora (also known as mesquite) is a
deciduous shrub with spreading cylindrical branches
that can grow from 3-12 meters in height. The trunk
of the tree is usually around 1.2 in diameter with
compound leaves. Florets as usual, greenish-white,
turning light yellow (Burkart, 1976). Prosopis
juliflora is an evergreen tree native to the western
hemisphere, including rangelands in South America,
Central America, and the Caribbean (López-Franco et
al., 2013), and have been introduced into several arid
and semi-arid countries of Africa, Asia, and Australia
over the last century due to their quality wood for
fuel, timber and seeds for human and animal food.
Despite their advantages, these plants are also highly
invasive in new habitats creating dense thorny
thickets (FAO, 2006).
The capability to grow under-stressed, drought
and high-temperature conditions have been made
them the most multipurpose tree in arid and semi-arid
regions (George et al., 2017). Mesquite was also has
been introduced to United Arab Emirates (UAE) in
the 1970s from Central America (Environment
Agency, 2020) on a large scale in the artificial forests
due to its faster growth, soil-binding capacity, and
desertification control. However, this invasive shrub,
locally know as (Ghweif), has aggressively invaded
the UAE environment causing severe damage. It
prevents other species from growing in their vicinity,
causing twofold-allelopathy. Moreover, there is a
noticeable accumulation of litters underneath the
Prosopis juliflora, compared with native species such
as Prosopis cineraria canopies, With a significant
negative influence on seed germination and growth of
understory species (El-Keblawy & Al-Rawai, 2007).
AlMaazmi, A. and Al-Ruzouq, R.
Multitemporal Remote Sensing for Invasive Prosopis Juliflora Plants Mapping and Monitoring: Sharjah, UAE.
DOI: 10.5220/0010440601490156
In Proceedings of the 7th International Conference on Geographical Information Systems Theory, Applications and Management (GISTAM 2021), pages 149-156
ISBN: 978-989-758-503-6
Copyright
c
2021 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
149
The aggressive behavior of Prosopis juliflora has
prompted researchers to establish scientific efforts
utilizing remote sensing data and Geographic
Information System (GIS). Remote sensing data can
cover more extensive areas than single plot studies
(Joshi et al., 2016). Furthermore, remote sensing
offers significant opportunities for providing timely
information on the invasion of non-native species into
native habitats.
Multispectral satellite images aid analyzing the
spectral response of the vegetation. The analysis
pattern of high and low spectral radiances introduced
one of the most effective spectral measurement for
vegetation known as the normalized difference
vegetation index (NDVI), which have been widely
and successfully used to provide surrogate indicators
of canopy greenness (Melesse et al., 2007), in
addition to ecologically map and classify vegetation
(Mehner et al., 2004).
Remote sensing data and spectral analysis could
be beneficial to map and detect Prosopis juliflora.
Several attempts to map Prosopis juliflora in Africa
are well acknowledged. For example, Normalized
Difference Vegetation Index (NDVI) from
multispectral, multi-temporal satellite imagery,
combined supervised classification methods such as
maximum likelihood classification (Rembold et al.,
2015) and random forest classifier is used to quantify
Prosopis juliflora during dry seasons in Somaliland
(Meroni et al., 2017). Another study conducted in
Ethiopia utilized spectral indices such as NDVI and
Enhanced Vegetation Index (EVI) in addition to topo-
climatic variables as input to correlative models to
map both the current and potential distribution of P.
juliflora (Wakie et al., 2014). Other spectral indices
such as Normalized Difference Infrared Index (NDII)
is used to retrieve the mesquite tree distribution area
in Sudan through measuring foliar water content,
where native plants have more water-stressed
growing than the mesquite (Hoshino et al., 2011). In
Kenya, a Land Use Land Cover (LULC)
classification using random forest supervised
classification was used to detect Prosopis juliflora, on
10–15 years study. The study showed that Prosopis
has not stopped or slowed down (Mbaabu et al.,
2019). In Asia, similar studies using remote sensing
data were conducted as well. In India NDVI and
Support Vector Machine (SVM) supervised
classifications were used to map Prosopis juliflora
(Vidhya et al., 2017). NDVI and NDII were also used
to map Prosopis juliflora in Tirunelveli and
Palayamkottai in India (Ragavan et al., 2015). A
study in the United Arab Emirates (UAE) used
onscreen digitizing of multi-temporal aerial images
with the aid of statistical analysis to estimate percent
cover, patch density, patch size, and mean patch
shape index. The study helped to understand the
spread of Prosopis juliflora in two areas in UAE
(Filayah and Khut), where it showed a dramatic
increase during 19 years of time (Issa, S. M. et al.,
2008).
This study focuses on utilizing time serires
analysis using Landsat-8 images with support vector
machiene classification and NDVI to monitor
Prosopis juliflora in Sharjah City, UAE, to obtain
numerical analysis for the covered area every decade
from 2000-2020.
2 STUDY AREA
The UAE is situated in the Middle East, and shares
coastline with Gulf of Oman in the east and the
Arabian Gulf in west, and boarders with Oman and
Saudi Arabia (Figure 1). The UAE is located in an
arid tropical region with an average annual rainfall of
80-140 mm (Sherif et al., 2014). This study
consentrate on Sharjah city, one of the largest cities
in the UAE with an area of 2590 km
2
, located at
25.3463° N, 55.4209° E. The annual rainfall of
Sharjah city is approximately 106.9 mm. Large parts
of the emirate consists of desert areas, profound soil
shaped in eolian sands, in addition to agriculture
areas, and expansions of acacia woodlands. (Al-
Ruzouq et al., 2019).
Figure 1: Study Area Sharjah City – UAE.
GISTAM 2021 - 7th International Conference on Geographical Information Systems Theory, Applications and Management
150
3 MATERIALS AND SATELLITE
DATA
Ten years of satellite imagery were collected from
Landsat 7 and Landsat 8 during the year 2010 to 2020
for yearly analysis. Furthermore, one image was
collected in 2000 for decade analysis during 2000,
2010, and 2020 (Table 1). The images were collected
during the dry season in UAE in August as the
reflection of P. Juliflora appears brighter than other
vegetation. Two scenes were required to cover the full
Sharjah city. In addition, a high-resolution Khalifa-
Sat image was acquired as well for visual inspecting
areas of vegetation.
Table 1: Satellite Imagery Thematic Data Source.
4 METHODOLOGY
The developed methodology to map P.Juliflora trees
in Sharjah is represented in (Figure 2) throughout
three stages. First, historical satellite imagery data is
required to develop the thematic layers of the study
area. The first stage is pre-processing to prepare the
data for major processing in terms of band stacking
and area coverage. Then data analysis using NDVI
and Support Vector MAchiene. Finally mapping
using ArcGIS software to map total Prosopis juliflora
areas
4.1 Pre-processing
With LANDSAT-7, the Scan Line Corrector (SLC)
failed On May 31, 2003. The sensor has acquired and
delivered data with lined gap strips on each band
(Loveland & Irons, 2016). Therefore, nearest
neighbor (NN) resampling was used to fill the gap in
each band for images from 2010-2012 (Figure 3).
Figure 2: Proposed Methodology.
Figure 3: LandSat-7 SLC Using Nearest Neighbor (NN)
Resampling.
Figure 4: Common Pre-Processing Steps.
For each output point, the closest input detector
sample must be identified and selected as the output
image value.
Once Scan lines were corrected, both LANDSAT
7 and LANDSAT 8 were processed with common
steps, including band stacking, scene mosaicking, and
extracting the study area (Figure 4). These steps are
necessary for time series analysis, using WGS 84
Satellite Spatial Resolution
(meter)
Spectral
Resolution (micro meter)
Data
Acquisition
Year
LANDSAT
7
30 m multi-spectral
60 m Thermal
- Band 1 Visible Blue (0.45 - 0.52 µm)
- Band 2 Visible (0.52 - 0.60 µm)
- Band 3 Visible (0.63 - 0.69 µm)
- Band 4 Near-Infrared (0.77 - 0.90 µm)
- Band 5 Short-wave Infrared (1.55 - 1.75 µm)
- Band 6 Thermal (10.40 - 12.50 µm)
2000
2010
2011
2012
LANDSAT
8
30 m multi-spectral
15 m Panchromatic
- Band 1 Coastal aerosol (0.43-0.4530 µm)
- Band 2 Visible Blue (0.45-0.5130 µm)
- Band 3 Visible Green (0.53-0.5930 µm)
- Band 4 Visible Red (0.64-0.6730 µm)
- Band 5 Near Infrared (NIR) (0.85-0.8830 µm)
- Band 6 SWIR (11.57-1.6530 µm)
- Band 7 SWIR (22.11-2.2930 µm)
- Band 8 Panchromatic (0.50-0.6815)
- Band 9 Cirrus (1.36-1.3830 µm))
- Band 10 Thermal Infrared (110.6-11.19100 µm)
- Band 11 Thermal Infrared (211.50-12.51100 µm)
2013
2014
2015
2016
2017
2018
2019
2020
Khalifa-
SAT
0.7 m Panchromatic
2.98 m multi-
spectral
- Panchromatic (0. 55-0.9 µm)
- MS1: Visible Blue (0.450 – 0.520 µm)
- MS2: Visible Green (0.520 – 0.590 µm)
- MS3: Visible Red (0.630 – 0.690 µm)
- MS4: Near Infrared (NIR) (0.77 - 0.889 µm)
2019
Multitemporal Remote Sensing for Invasive Prosopis Juliflora Plants Mapping and Monitoring: Sharjah, UAE
151
zone 40N UTM coordinate system, where areas in
meters can be easily calculated to fulfill the need of
this study. A further step is needed for KhalifaSAT,
which includes geometric correction using 8 well-
distributed ground control points (GCP) and
projective transformation.
4.1 Data Analysis Processing
In the third stage, all the thematic layers were
processed to map vegetation and Prosopis juliflora.
Two approaches were used, Normalized Difference
Vegetation Index (NDVI) and supervised
classification using Support Vector Machine (SVM).
4.1.1 Normalized Difference Vegetation
Index (NDVI)
The Normalized Difference Vegetation Index
(NDVI) synthesizes the information from two
spectral channels: red and near-infrared. Vegetation
tends to have higher reflectance in the red band and
lower reflectance in the near-infrared band. Hence,
this index can be repressed as simple mathematical
expressions represented in equation (1) that combines
spectral measurements used to identify the presence
of vegetation in remotely sensed images (Leprieur et
al., 1996)(Kefalas et al., 2018).
𝑁𝐷𝑉𝐼
𝑁𝐼𝑅  𝑅𝐸𝐷
𝑁𝐼𝑅  𝑅𝐸𝐷
(1)
NDVI was measured for three years (2000, 2010
and 2020), in addition to a yearly measurement from
2010-2020. Further assessment of vegetation analysis
was conducted using change detection over two
periods from 2000-2010 and 2010-2020. Change
detection analysis is conducted through equation (2).
𝐶𝐷 𝐼𝑚𝑎𝑔𝑒1  𝐼𝑚𝑎𝑔𝑒 2
(2)
4.1.2 Object-based-Support Vector Machine
Classification
The Support Vector Machine classifier is promising
in terms of vegetation classification as it can result in
up to 80% overall accuracy assessment with Kappa
0.8 comparing to Maximum Likelihood
Classification and Spectral Angular Mapper (Braun et
al., 2010). P. Juliflora training samples were taken
from the area behind Sharjah airport (ROI1) as shown
in figure 5. The area is located within latitude and
longitude ( 220.61'N, 55° 31.64'E ), with a total
land area around 21.19 km
2
, most of the land is
vegetated with P. juliflora, with medium canopy sizes
(5–10 m diameters) and different densities
distribution[10], (Figure 6).
Figure 5: KhalifaSat image of ROI1 behind Sharjah
Airport.
Figure 6: P. Juliflora as seen from satellite nadir view taken
from Google Earth Image from ROI1.
Two classes were identified in the training samples.
The first one is P. juliflora taken from ROI1. Likewise
other vegetation types, P. juliflora has a high
reflectance from NIR band and low reflectance in the
red band. However, this tree has distinguished
spectral response where it has a minor peak and green
band as well. More than 1000 pixel samples were
taken from each year as training sample (Table 2).
The second class is all other vegetation types taken
from random distributed vegetated areas.
Table 2: Training samples were taken for SVM classifier.
Year Total number of Pixels Total Training Area ( Km
2
)
2000 3,598 3.25
2010 1,307 1.18
2020 5,386 4.86
GISTAM 2021 - 7th International Conference on Geographical Information Systems Theory, Applications and Management
152
4 RESULTS AND DISSCUSSION
4.1 Vegetation Analysis of Sharjah City
The multi-temporal analysis of the NDVI during
(2000,2010,2020) for the vegetated areas in Sharjah
city showed that the vegetation density increased
during the past two decades. The highest density
appeared in 2020 with 67.18km
2,
followed by the year
2010 with a total area of 56.55 km
2
, with increase of
0.188%. Finally, the year 2000 showed the least total
vegetated area with 42.96 km
2
(Figure 7).
Figure 7: Total Vegetated Areas in Sharjah City in (a) 2000,
(b) 2010, (c)2020.
The vegetation index during the period 2010-
2020 showed that variations in vegetated areas per
year were small as they did not exceed 0.25% of net
change. The lowest vegetated areas were found to be
in 2013 with 45.14 km
2
(Figure 8). The vegetation
areas were mapped in (Figure 9)
Figure 8: Bar chart of 10 Years Of Total Vegetation In
Sharjah City From 2010-2020.
Vegetation cover change detection was
implemented over two periods, from 2000-2010, and
2010-2020. In 2000-2010, vegetation cover increased
by a total area of 32.18 km
2
.The vegetation increase
concentrated in city center area, university city area,
and Sharjah-Ajman border farms area. In the same
period, vegetation decreased by 30.43 km
2
, the
decrease mostly found in the farmed eastern part of
Figure 9: Vegetated areas maps in Sharjah City in (a) 2000,
(b) 2010, (c)2020.
the city. In 2010-2020, vegetation cover increased
dramatically, almost double by the total area of 75.28
km
2.
The vegetation increase distributed throughout
the city and concentrated in the city center area,
university city area, and Sharjah-Ajman border farms
area, in addition to the farmed eastern part of the city.
In the same period, vegetation decreased sharply by
total area of 11.74km
2
. The decrease is mostly found
in farms in the eastern and southern parts of the city.
(Figure 10) shows the maps of change detection in
both periods.
Figure 10: Change detection in Sharjah city in period (a)
2000-2010, (b) 2010-2020.
4.2 Vegetation Analysis of Mapped P.
Juliflora
The multi-temporal analysis of the Support Vector
Machine (2000,2010,2020) for the vegetated areas of
P. Juliflora trees in Sharjah (Figure 11) showed that
the highest density appeared in 2020 with 14.13 km
2
followed by the year 2000 with a total area of 11.99
km
2
. Finally, the year 2010 showed the least total
vegetated area with 9.26km
2
as shown in (Figure 12).
The increase between 2010-2020 in the last decade
was 0.525%.
Multitemporal Remote Sensing for Invasive Prosopis Juliflora Plants Mapping and Monitoring: Sharjah, UAE
153
Figure 11: Total P. Juliflora Area In Sharjah City.
Figure 12: Total P. Juliflora Area in Sharjah City in (a)
2000, (b) 2010, (c)2020.
4.3 Vegetation Accuracy Analysis of
Mapped P. Juliflora
Eight sites have been selected (Table 3) for visual
accuracy assessment from high-resolution google
earth (Figure 13). These sites are taken from the
resulted P. Juliflora classification results (Figure 14).
Table 3: Inspection sites ID and their location.
Figure 13: Inspection Sites Visually Inspected (a) 2020_1,
(b) 2020_2, (c) 2020_3, (d) 2020_4, (e) 2020_5, (f) 2020_6,
(g) 2020_7, (h) 2020_8.
The visual inspection showed that all vegetation in
the inspection sites identified as P. Juliflora, except
for site 2000_4 and 2000_5. The inspection sites also
suggest that P. Juliflora density is higher at locations
of lower altitude open fields sand-based areas. The
visual inspection also showed that P. Juliflora showed
minimum existence nearby other vegetation
communities.
Figure 14: Map of Inspection Sites.
5 CONCLUSIONS
Identifying locations of Prosopis Juliflora is a
significant environmental and ecological initiative for
any arid region country. In this study, an algorithm
that combines remote sensing data and GIS along
with ML was developed to locate areas of Prosopis
Juliflora. Eleven thematic layers from satellite
imagery were collected from Landsat 7 and Landsat
8. Two approached were used. The first one is using
vegetation index NDVI, and the second one is a
combination of NDVI and supervised classification.
The NDVI maps all types of vegetation in Sharjah
city using Red and NIR bands. Vegetation in Sharjah
city showed an increment in the past two decades
where it increased from 42.96 km
2
in 2000 to 67.19
km
2
in 2020. The second approach is using a Support
Vector Machine (SVM) over vegetated areas from
NDVI. Over 1000 pixels samples were taken to train
the supervised classification with two classes:
ID Year location correctly mapped
2020_1 2020 55.5587161°E 25.3250445°N yes
2020_2 2020 55.5713399°E 25.4544956°N yes
2020_3 2020 55.5570937°E 25.4715355°N yes
2020_4 2020 55.5529439°E 25.3038846°N no
2020_5 2020 55.9371005°E 25.1879427°N no
2020_6 2020 55.4891707°E 25.2709255°N yes
2020_7 2020 55.5530813°E 25.2568040°N Yes
2020_8 2020 55.5876650°E 25.4609975°N yes
GISTAM 2021 - 7th International Conference on Geographical Information Systems Theory, Applications and Management
154
P.Juliflora and other vegetation. Total area increased
up to 14.13 km
2
in 2020 after it was 9.26 km
2
only in
2010. The vegetation analysis showed that P. Juliflora
increase dramaticlly with 0.525% in the last decade
between 2010 and 2020 comparing with other
vegetation which increased by 0.1888% only. The
visual inspection suggest that P. Juliflora density is
higher at locations of lower altitude open fields sand-
based areas. The visual inspection also showed that P.
Juliflora showed minimum existence nearby other
vegetation communities. Finally, the developed
technique can be generalized and utilized over new
arid-region locations.
ACKOWLEDGEMENTS
The project is funded by the University of Sharjah
(UoS) under grant research project ID:
#20020401163.
REFERENCES
Essl, F., Lenzner, B., Bacher, S., Bailey, S., Capinha, C.,
Daehler, C., Dullinger, S., Genovesi, P., Hui, C.,
Hulme, P. E., Jeschke, J. M., Katsanevakis, S., Kühn,
I., Leung, B., Liebhold, A., Liu, C., MacIsaac, H. J.,
Meyerson, L. A., Nuñez, M. A., Roura-Pascual, N.
(2020). Drivers of future alien species impacts: An
expert-based assessment. Global Change Biology,
26(9). https://doi.org/10.1111/gcb.15199
Jeschke, J. M., Bacher, S., Blackburn, T. M., Dick, J. T. A.,
Essl, F., Evans, T., Gaertner, M., Hulme, P. E., Kühn,
I., Mrugała, A., Pergl, J., Pyšek, P., Rabitsch, W.,
Ricciardi, A., Richardson, D. M., Sendek, A., Vilà, M.,
Winter, M., & Kumschick, S. (2014). Defining the
impact of non-native species. Conservation Biology,
28(5). https://doi.org/10.1111/cobi.12299
Mwangi, E., & Swallow, B. (2005). Invasion of Prosopis
juliflora and local livelihoods. Case study of Lake
Baringo area of Kenya. World Agroforestry Center,
May.
Berhanu, A., & Tesfaye, G. (2006). The Prosopis Dilemma,
impacts on dryland biodiversity and some controlling
methods. Journal of the Drylands1 (2), 1(2).
Burkart, A. (1976). A Monograph Of The Genus Prosopis
(Leguminosae Subfam. Mimosoideae). Journal of the
Arnold Arboretum, 57(3).
López-Franco, Y. L., Cervantes-Montaño, C. I., Martínez-
Robinson, K. G., Lizardi-Mendoza, J., & Robles-
Ozuna, L. E. (2013). Physicochemical characterization
and functional properties of galactomannans from
mesquite seeds (Prosopis spp.). Food Hydrocolloids,
30(2). https://doi.org/10.1016/j.foodhyd.2012.08.012
FAO (2006). Problems posed by the introduction of
Prosopis spp. in selected countries. Plant Production
and Protection Division. FAO, Rome (published).
George, S., Manoharan, D., Li, J., Britton, M., & Parida, A.
(2017). Transcriptomic responses to drought and salt
stress in desert tree Prosopis juliflora. Plant Gene, 12.
https://doi.org/10.1016/j.plgene.2017.09.004
Environment Agency, "Taking action on terrestrial and
freshwater alien species in Abu Dhabi: from prevention
to control" Environment Agency, Abu Dhabi, UAE,
2018. Accessed on: Oct., 18, 2020. [Online]. Available:
https://www.ead.gov.ae/storage/Post/files/460fe11c56
4b7f3674b23b30c8dcc774.pdf
El-Keblawy, A., & Al-Rawai, A. (2007). Impacts of the
invasive exotic Prosopis juliflora (Sw.) D.C. on the
native flora and soils of the UAE. Plant Ecology,
190(1). https://doi.org/10.1007/s11258-006-9188-2
Joshi, N., Baumann, M., Ehammer, A., Fensholt, R.,
Grogan, K., Hostert, P., Jepsen, M. R., Kuemmerle, T.,
Meyfroidt, P., Mitchard, E. T. A., Reiche, J., Ryan, C.
M., & Waske, B. (2016). A review of the application of
optical and radar remote sensing data fusion to land use
mapping and monitoring. In Remote Sensing (Vol. 8,
Issue 1). https://doi.org/10.3390/rs8010070
Melesse, A. M., Weng, Q., Thenkabail, P. S., & Senay, G.
B. (2007). Remote sensing sensors and applications in
environmental resources mapping and modelling. In
Sensors
(Vol. 7, Issue 12). https://doi.org/
10.3390/s7123209
Mehner, H., Cutler, M., Fairbairn, D., & Thompson, G.
(2004). Remote sensing of upland vegetation: The
potential of high spatial resolution satellite sensors.
Global Ecology and Biogeography, 13(4).
https://doi.org/10.1111/j.1466-822X.2004.00096.x
Rembold, F., Leonardi, U., Ng, W.-T., Gadain, H., Meroni,
M., & Atzberger, C. (2015). Mapping areas invaded by
Prosopis juliflora in Somaliland on Landsat 8 imagery.
Remote Sensing for Agriculture, Ecosystems, and
Hydrology XVII, 9637. https://doi.org/10.1117/
12.2193133
Meroni, M., Ng, W. T., Rembold, F., Leonardi, U.,
Atzberger, C., Gadain, H., & Shaiye, M. (2017).
Mapping Prosopis juliflora in West Somaliland with
Landsat 8 Satellite Imagery and Ground Information.
Land Degradation and Development, 28(2).
https://doi.org/10.1002/ldr.2611
Wakie, T. T., Evangelista, P. H., Jarnevich, C. S., & Laituri,
M. (2014). Mapping current and potential distribution
of non-native prosopis juliflorain the Afar region of
Ethiopia. PLoS ONE, 9(11). https://doi.org/10.1371/
journal.pone.0112854
Hoshino, B., Yonemori, M., Manayeva, K., Karamalla, A.,
Yoda, K., Suliman, M., Elgamri, M., Nawata, H., Mori,
Y., Yabuki, S., & Aida, S. (2011). Remote sensing
methods for the evaluation of the mesquite tree
(Prosopis juliflora) environmental adaptation to semi-
arid Africa. International Geoscience and Remote
Sensing Symposium (IGARSS). https://doi.org/10.1109/
IGARSS.2011.6049498
Multitemporal Remote Sensing for Invasive Prosopis Juliflora Plants Mapping and Monitoring: Sharjah, UAE
155
Mbaabu, P. R., Ng, W. T., Schaffner, U., Gichaba, M.,
Olago, D., Choge, S., Oriaso, S., & Eckert, S. (2019).
Spatial evolution of prosopis invasion and its effects on
LULC and livelihoods in Baringo, Kenya. Remote
Sensing, 11(10). https://doi.org/10.3390/rs11101217
Vidhya, R., Vijayasekaran, D., & Ramakrishnan, S. S.
(2017). Mapping invasive plant Prosopis juliflora in
arid land using high resolution remote sensing data and
biophysical parameters. Indian Journal of Geo-Marine
Sciences, 46(6).
K.Ragavan, J. Johnny, "Quantification of Invasive Colonies
of Prosopis Juliflora Using Remote Sensing and GIS
Techniques", International Journal of Engineering and
Technical Research (IJETR), vol. 3, no. 5, May 2015.
S. Issa and B. Dohai , "Gis Analysis of Invasive Prosopis
Juliflora Dynamics in Two Selected Sites From the
United Arab Emirates", Canadian Journal of Pure and
Applied Sciences , vol. 2, no. 1, pp. 235–242, 2008.
Sherif, M., Chowdhury, R., & Shetty, A. (2014). Rainfall
and Intensity-Duration-Frequency (IDF) Curves in the
United Arab Emirates. World Environmental and Water
Resources Congress 2014: Water Without Borders -
Proceedings of the 2014 World Environmental and
Water Resources Congress. https://doi.org/10.1061/
9780784413548.231
Al-Ruzouq, R., Shanableh, A., Yilmaz, A. G., Idris, A. E.,
Mukherjee, S., Khalil, M. A., & Gibril, M. B. A. (2019).
Dam site suitability mapping and analysis using an
integrated GIS and machine learning approach. Water
(Switzerland), 11(9). https://doi.org/10.3390/w11091
880
Loveland, T. R., & Irons, J. R. (2016). Landsat 8: The plans,
the reality, and the legacy. Remote Sensing of
Environment, 185. https://doi.org/10.1016/j.rse.2016.
07.033
Leprieur, C., Kerr, Y. H., & Pichon, J. M. (1996). Critical
assessment of vegetation indices from avhrr in a semi-
arid environment. International Journal of Remote
Sensing, 17(13). https://doi.org/10.1080/01431169608
949092
Kefalas, G., Lattas, P., Xofis, P., Lorilla, R. S., Martinis, A.,
& Poirazidis, K. (2018). The use of vegetation indices
and change detection techniques as a tool for
monitoring ecosystem and biodiversity integrity.
International Journal of Sustainable Agricultural
Management and Informatics, 4(1). https://doi.org/
10.1504/IJSAMI.2018.092411
Braun, A. C., Weidner, U., & Hinz, S. (2010). Support vector
machines for vegetation classification - A revision.
Photogrammetrie, Fernerkundung, Geoinformation,
2010(4). https://doi.org/10.1127/1432-
GISTAM 2021 - 7th International Conference on Geographical Information Systems Theory, Applications and Management
156