In e na ional Jou nal o Cu en Science Resea ch and Re iew
ISSN: 2581-8341
Volume 08 Issue 10 Oc obe 2025
DOI: 10.47191/ijcs /V8-i10-18, Impac Fac o : 8.048
IJCSRR @ 2025
www.ijcs .o g
5048 *Co esponding Au ho : Ousmane Badji Volume 08 Issue 10 Oc obe 2025
A ailable a : www.ijcs .o g
Page No. 5048-5061
Geospa ial Assessmen o Th ee Decades o Sho eline Shi s and Two
Decades o Vege a ion Change in he G and Saloum T ansbounda y
We land Complex, Senegal-The Gambia
Ousmane Badji1*, Adam Ceesay2, Kwame Oppong Hackman3
1WASCAL G adua e Resea ch P og amme on Clima e Change and Land Use, Depa men o Ci il Enginee ing, Kwame
Nk umah Uni e si y o Science and Technology, Kumasi, Ghana
2Ins i u des Sciences de l’En i onnemen , Cheikh An a Diop Uni e si y (UCAD)
3Compe ence Cen e , Wes A ican Science Se ice Cen e on Clima e Change and Adap ed Land Use (WASCAL),
Ouagadougou, Bu kina Faso
ABSTRACT: Coas al we lands a he land–sea in e ace a e on he on line o clima e change, ye in eg a ed e idence on
geomo phic and ecological esponses emains limi ed in Wes A ica. We quan i ied sho eline ajec o ies (1990–2020) and land-
co e dynamics (2000–2020) ac oss he ansbounda y G and Saloum complex (Senegal–The Gambia) using Landsa su ace-
e lec ance ime se ies, spec al indices (NDVI, NDWI, NDBI), and he Digi al Sho eline Analysis Sys em (DSAS). Sho elines we e
ex ac ed om NDWI-based wa e masks, il e ed and ec o ized, hen analyzed in DSAS wi h End Poin Ra e s a is ics. Vege a ion
was mapped in Google Ea h Engine wi h a Random Fo es classi ie (mang o e, o he ege a ion, buil /ba e, wa e ). The coas line
is domina ed by e osion (mean −2.44 m·y ⁻¹) in e spe sed wi h localized acc e ion (mean +1.84 m·y ⁻¹). E osion ho spo s
concen a e in cen al sec o s, whe eas mixed e osion–acc e ion pa e ns occu nea he no he n and sou he n mou hs. Concu en ly,
mang o e co e expanded om 57,867.61 ha in 2000 o 66,840.17 ha in 2020 (~+15.5%), while o he ege a ion declined om
23,483.18 ha o 16,146.11 ha (~−31.3%). Wi hin a 1-km coas al bu e , mang o es emained b oadly s able o sligh ly inc easing
(16.43%→16.81%). These indings depic a dynamic ye esilien sys em whe e mang o e gains coexis wi h he e ogeneous
sho eline e ea and con e sion o non-mang o e co e s o ba e subs a es and wa e . Managemen should sa egua d landwa d
mig a ion co ido s, a ge e osion-p one eaches wi h na u e-based measu es, and ins i u ionalize a ansbounda y moni o ing,
epo ing, and e i ica ion amewo k ha upda es DSAS and sa elli e p oduc s a 2–3-yea in e als while in eg a ing in-si u
ele a ion, salini y, and sedimen da a. Ou wo k low p o ides ans e able, decision- ele an e idence o coas al adap a ion and
blue-ca bon planning in da a-limi ed del as and policy design.
KEYWORDS: Coas al e osion, Sho eline change, Mang o es, Remo e sensing, DSAS, Landsa , Google Ea h Engine, G and
Saloum (Senegal–The Gambia).
1. INTRODUCTION
Coas al we lands a he land–sea in e ace a e on he on line o clima e change. Rising seas, d i en by ocean he mal expansion
and accele a ing land-ice loss, a e al e ing sho eline posi ion, inunda ion egimes, and salini y g adien s ha s uc u e we land
ecosys ems (Chu ch e al., 2001). These physical shi s cascade in o ecological and li elihood impac s o communi ies dependen
on ishe ies, ag icul u e, and coas al p o ec ion; con e sely, slowe ela i e sea-le el ise can expand he window o adap a ion in
del as and low-lying coas s ((IPCC, 2007, 2023).
Mang o e ecosys ems a e bo h clima e sen inels and bu e s: hey seques e ca bon, a enua e wa es, and sus ain coas al
economies, ye hey a e sensi i e o changes in hyd odynamics and salini y (Ellison, 2014). P olonged o mo e equen inunda ion
and salini y shi s can exceed species-speci ic ole ance h esholds, igge ing dieback o communi y eassembly unless sedimen
acc e ion and landwa d mig a ion keep pace (F iess e al., 2012). These esponses a y egionally, unde sco ing he need o si e-
speci ic, decadal moni o ing (Ellison, 2014; F iess e al., 2012).
In Wes A ica, and Senegal in pa icula , coas al isks ha e in ensi ied, h ea ening se lemen s and c i ical habi a s. Along
sec o s o he Senegalese coas , mul i-yea analyses epo sho eline e ea s commonly on he o de o 1–2 m·y ⁻¹ (Diop e al.,
In e na ional Jou nal o Cu en Science Resea ch and Re iew
ISSN: 2581-8341
Volume 08 Issue 10 Oc obe 2025
DOI: 10.47191/ijcs /V8-i10-18, Impac Fac o : 8.048
IJCSRR @ 2025
www.ijcs .o g
5049 *Co esponding Au ho : Ousmane Badji Volume 08 Issue 10 Oc obe 2025
A ailable a : www.ijcs .o g
Page No. 5048-5061
2014; P. W. Bakhoum e al., 2017). A no able mo phological shi in he Saloum sys em was he 1987 s o m-b each o he Sangoma
spi , which econ igu ed connec i i y and sedimen dynamics (MEPN, 2006). Scena io analyses u he sugges subs an ial po en ial
loss o low-lying a eas unde sea-le el ise in he Saloum es ua y and signi ican e osion p essu e along he Senegal–Gambia bo de
(Jallow e al., 1996; Niang e al., 2010).
The G and Saloum, encompassing Senegal’s Saloum Del a Biosphe e Rese e and The Gambia’s Niumi Na ional Pa k, o ms
a ansbounda y Ramsa we land complex o high ecological and economic alue. Al hough nume ous s udies ha e add essed
h ea s, p essu es, and clima e–mang o e linkages in pa s o he sys em (D ame & Sambou, 2013; SIDIBE, 2010; Sow & Ba, 2019),
an in eg a ed, geospa ially explici assessmen ha couples h ee decades o sho eline dynamics wi h wo decades o ege a ion
change emains limi ed. This s udy add esses ha gap by le e aging sa elli e emo e sensing and GIS o quan i y sho eline
ajec o ies (1990–2020) and ege a ion dynamics wi h emphasis on mang o es (2000–2020), p o iding a decision- ele an
geospa ial e idence base o coas al managemen in he G and Saloum (Ceesay e al., 2017).
2. METHODOLOGY
2.1. S udy a ea
The G and Saloum ansbounda y complex (Figu e 1) co e s an a ea o 83,758 ha. I is composed o :
- The Saloum Del a Na ional Pa k (PNDS) in Senegal, which is loca ed be ween la i udes 13.583333 and 13.916667, and
longi udes 16.466667 and 16.083333. This si e was e ec ed by a Dec ee unde he Senegalese law N°76 577 on 28 h Ma ch 1976
and co e ed a o al o 76,000 ha.
- The Niumi Na ional Pa k (NNP) in The Gambia, which is loca ed be ween la i udes 13.516667 and 13.983333 and
longi udes 16.933333 and 16.083333, is a coas al s ip o 7758 ha e ec ed as a Na ional Pa k in 1986 and a RAMSAR Si e in Oc obe
2008. I is he na u al sou he n ex ension o he Saloum Na ional Pa k (PNDS) (WOW, 2015).
2.2. Clima e
The G and Saloum ansbounda y complex is ma ked by a Sudano-Sahelian clima e ype cha ac e ised by ain all alues be ween
400 and 800 mm wi h an a e age empe a u e o 29° C. The ain all is gene ally less in he no he n pa o he complex (Saloum)
and g ea e in he sou he n egion (Niumi). The Cana y cu en coas al in luence is much mo e p ominen on he Senegalese sec ion
o he complex. Two main seasons cha ac e ise he clima e:
- A d y season (cold om No embe o Ma ch, ho om Ma ch o June), whe e he p e ailing winds a e ma i ime ade
winds, esh (in a no h o no h-wes di ec ion)
- A d y con inen al winds (in an eas o no h-eas di ec ion, known as Ha ma an).
- A ho , humid ainy season om July o Oc obe , domina ed by monsoon winds (di ec ion: Wes and sou hwes ). Annual
ain all in he Saloum Del a has declined om a ange o 600-900 mm o he pe iod 1931- 1960 o less han 400-600 mm oday.
The e is a o al o 50-60 days o ain pe yea , wi h maximum ain all in Augus . Recen ly, in Niumi, he e ha e been epo s o
inc eased annual a e age ain all om 2000 o 2010, and his ce ainly migh be he same a he whole complex le el. A e age
annual empe a u es a y be ween 26 and 31° C (WOW, 2015).
In e na ional Jou nal o Cu en Science Resea ch and Re iew
ISSN: 2581-8341
Volume 08 Issue 10 Oc obe 2025
DOI: 10.47191/ijcs /V8-i10-18, Impac Fac o : 8.048
IJCSRR @ 2025
www.ijcs .o g
5050 *Co esponding Au ho : Ousmane Badji Volume 08 Issue 10 Oc obe 2025
A ailable a : www.ijcs .o g
Page No. 5048-5061
Figu e 1: Map o he s udy a ea (Saloum-Nuimi T ansbounda y Ramsa Complex)
2.3. Remo e sensing o he sho eline dynamic
2.3.1. Sa elli e images
Sa elli e images wi h di e en spa ial esolu ions p ocessed wi h a ious change analysis me hods a e e ec i e o quan i ying
changes in he we land (Tou e e al., 2018). Acco dingly, su ace e lec ance images om Landsa 5, 7, and 8 (Table 1) be ween
1990 and 2020 we e accessed and p ocessed in ENVI.
Table 1: Landsa images p ope ies
Da es and ime
o acquisi ion
Pa hs and Rows
Cloud
Co e
Senso s
Da a P o ide :
Bands
1990-12-21
10:47:02
(PATH: 205,
ROW: 50) &
(PATH: 205,
ROW: 51)
0
Landsa 5
USGS
Blue, G een, Red, NIR,
SWIR-1, SWIR-2, NDVI,
NDBI, NDWI
2000-12-08
11:17:48
0
Landsa 7
In e na ional Jou nal o Cu en Science Resea ch and Re iew
ISSN: 2581-8341
Volume 08 Issue 10 Oc obe 2025
DOI: 10.47191/ijcs /V8-i10-18, Impac Fac o : 8.048
IJCSRR @ 2025
www.ijcs .o g
5051 *Co esponding Au ho : Ousmane Badji Volume 08 Issue 10 Oc obe 2025
A ailable a : www.ijcs .o g
Page No. 5048-5061
2.3.2. Image p e-p ocessing
The bands in he su ace e lec ance images we e a mosphe ically co ec ed and o ho ec i ied. A any pixel loca ion, he alue
eco ded on a emo ely sensed image does no e e o he ue g ound-lea ing adiance a ha pa icula poin . One pa o he
b igh ness is due o he a ge o in e es e lec ance and he emainde om he a mosphe e i sel . Thei con ibu ions a e no known
a p io i, so he objec i e o a mosphe ic co ec ion was o quan i y hese wo componen s in o de o use co ec a ge e lec ance
(Themis ocleous e al., 2008). The o ho ec i ica ion is necessa y because o de o ma ions mainly due o came a dis o ions and
acquisi ion geome y.
The e ain- ela ed geome ic dis o ions ha we e emo ed du ing he o ho ec ica ion s age a e ela ed o he image o ma ion
p ocess (e o acking), such as dis o ions caused by he pla o m, and mainly ela ed o he a ia ion o he ellip ic mo emen
a ound he Ea h, ins an aneous ield o iew, opog aphic elie changes, e c. (Chmiel e al., 2004).
2.3.3. Da a analysis and p ocessing
Spec al indices, also known as band ans o ma ions, we e ob ained om he Landsa 5, 7, and 8 su ace e lec ance images by
he ollowing equa ions (Table 2).
Table 2: Fo mulas o he NDVI, NDWI, and NDBI calcula ion
Index Used
Equa ions
NDVI
𝑵𝑫𝑽𝑰 = 𝝆𝑵𝑰𝑹 − 𝝆𝑹𝑬𝑫
𝝆𝑵𝑰𝑹 + 𝝆𝑹𝑬𝑫
NDWI
𝑵𝑫𝑾𝑰 = 𝝆𝑮𝒓𝒆𝒆𝒏 − 𝝆𝑵𝑰𝑹
𝝆𝑮𝒓𝒆𝒆𝒏 + 𝝆𝑵𝑰𝑹
NDBI
𝑵𝑫𝑩𝑰 = 𝝆𝑺𝑾𝑰𝑹𝟏 − 𝝆𝑵𝑰𝑹
𝝆𝑺𝑾𝑰𝑹𝟏 + 𝝆𝑵𝑰𝑹
Wi h: ρ_G een=ToA e lec ance o g een band, ρ_NIR=ToA e lec ance o nea in a ed band.
ρ_NIR=ToA e lec ance o nea in a ed band ρ_SWIR1= sho -wa e in a ed
2.4. Sho eline De ec ion and analysis
2.4.1. Sho eline p ocessing
DSAS is one o he mos e icien and e ec i e as well as less ime-consuming ools in sho eline change analysis compa ed wi h
he many adi ional ools and me hods and p oduces esul s o be e accu acy (Seko ski e al., 2014). I elies on inpu da a such
as he da e and yea and a digi ized geome y (in shape ile o ma ) o he sho eline. A se ies o p ocesses we e ca ied ou o analyse
he changes in he sho eline, as gi en in Figu e 2.
➔Seg ega ion o wa e and non-wa e ea u e using a spec al index
NDWI, as de ined ma hema ically in Table 2, was used o de e mine he wa e and non-wa e ea u es. NDWI alue anges om
−1 o +1. The NDWI image ypically p o ides posi i e esul s o wa e ea u es and nega i e o non-wa e ea u es (McFEETERS,
1996). Only wa e and non-wa e ea u es a e equi ed o delinea e he sepa a ion line as a sho eline, and he e o e a bina y image
classi ica ion, i.e., 0 and 1, was pe o med o depic ing non-wa e and wa e ea u es (Ji e al., 2009).
2010-12-28
11:17:22
0
Landsa 5
2020-12-07
11:27:49
0
Landsa 8
In e na ional Jou nal o Cu en Science Resea ch and Re iew
ISSN: 2581-8341
Volume 08 Issue 10 Oc obe 2025
DOI: 10.47191/ijcs /V8-i10-18, Impac Fac o : 8.048
IJCSRR @ 2025
www.ijcs .o g
5052 *Co esponding Au ho : Ousmane Badji Volume 08 Issue 10 Oc obe 2025
A ailable a : www.ijcs .o g
Page No. 5048-5061
➔Pos -p ocessing o bina y as e image
A 3 × 3 mode il e was applied o he pos -p ocessing ope a ion ha subs i u ed he isola ed pixels o he mos common
neighbo ing class (ei he wa e class o non-wa e class) o decompose he sca e ed and isola ed pixels (Ba uś, 2014). The jagged
bounda ies o he wa e and non-wa e classes we e smoo hened by using QGIS clean ool. The sho eline ec o was hen p oduced
using a as e bina y image, and he abu ing line o wa e and non-wa e class was aced o ex ac he inal sho eline.
Figu e 2: Sho eline de ec ion and analysis p ocess
➔Sho eline gene a ion
A e ha , he di e en ime pe iods sho eline da a was ed o he DSAS o u he compu a ion o sho eline change o 30 yea s
om 1990 o 2020. In he DSAS ool, sho elines posi ions a e compiled wi h i e a ibu e ields which include Objec ID (a unique
numbe assigned o each), shape (polygon), da e (o iginal su ey yea ), and shape leng h, and unce ain y alues. Sho elines o
di e en yea s we e me ged as a single ea u e, which c ea es a single shape ile o he mul iple sho elines. The baseline was
gene a ed o calcula ing he sho eline change by closely digi izing he di ec ion and shape o he ou e sho eline. F om his p ocess,
he a es o sho eline change we e gene a ed.
➔Sho eline change s a is ics
The calcula ion o he sho eline change was done in he o m o End poin Ra e (EPR). The inal decision ma ix was p epa ed
on he basis o he esul s and ou pu . EPR o mula (equa ion 1) was used o p esen he compu a ional esul s. The DSAS ool i sel
chooses he sho eline ansec s, gi es hem dependen and independen a iables, and au oma ically calcula es (EPR) he a es o
e osion and deposi ion. The accu acy le el would be as high as when mo e yea s sa elli e da a se has been inco po a ed (Seko ski
e al. 2014). Fo example, 4 yea s o sa elli e images we e chosen o he sho eline change analysis. A ± 5 m unce ain y and 95%
con idence in e al was se as de aul pa ame e o calcula e he s a is ics.
𝐄𝐏𝐑 =𝐃𝐢𝐬𝐭𝐚𝐧𝐜𝐞 𝐢𝐧 𝐦𝐞𝐭𝐫𝐞𝐬 (𝐦)
𝐭𝐢𝐦𝐞 𝐛𝐞𝐭𝐰𝐞𝐞𝐧 𝐨𝐥𝐝𝐞𝐬𝐭 𝐚𝐧𝐝 𝐦𝐨𝐬𝐭 𝐫𝐞𝐜𝐞𝐧𝐭 𝐬𝐡𝐨𝐫𝐞𝐥𝐢𝐧𝐞(𝐘𝐞𝐚𝐫) Eq. (1)
The EPR alues can ei he be posi i e o nega i e, whe e a posi i e alue ep esen s seawa d o o sho e mo emen , and a nega i e
alue ep esen s landwa d mo emen .
Landsa (5,7,8)
NDWI
Sho eline ex ac ion
Sho eline
Digi iza ion (DSAS)
Baseline (Bu e ing)
T ansec (50 m
in e al)
Sho eline change
s a is ics
Vec o isa ion o
as e da a
EPR
Final Decision Ma ix
In e na ional Jou nal o Cu en Science Resea ch and Re iew
ISSN: 2581-8341
Volume 08 Issue 10 Oc obe 2025
DOI: 10.47191/ijcs /V8-i10-18, Impac Fac o : 8.048
IJCSRR @ 2025
www.ijcs .o g
5053 *Co esponding Au ho : Ousmane Badji Volume 08 Issue 10 Oc obe 2025
A ailable a : www.ijcs .o g
Page No. 5048-5061
3.4. Remo e sensing o he ege a ion dynamic
3.4.1. Image classi ica ion ea u es
Due o he long eco d o con inuous obse a ion and high spa ial esolu ion, he Landsa se ies o sa elli e images a e one o he
mos use ul da a o biodi e si y assessmen (Hackman e al., 2017) and widely used in we land change assessmen s (Ajaj e al.,
2017; Ceesay e al., 2017). The Tie 1 su ace e lec ance images om he Landsa se ies o sa elli es a ailable in GEE we e used
because su ace e lec ance gi es he mos accu a e in o ma ion abou he su ace cha ac e is ics. In addi ion, h ee spec al indices
(NDWI, NDVI, and NDBI) ob ained om he Landsa 5, 7, and 8 su ace e lec ance images (see Table 2) we e used as ea u es.
Because he s udy a ea is a we land, he 30m spa ial esolu ion digi al ele a ion model (DEM) om he NASA Shu le Rada
Topog aphy Mission (SRTM) was added o he ea u e space o dis inguish mang o e om o he ege a ion. Thus, in all he ea u e
space was a 10-band image s ack made up o six su ace e lec ance bands (Blue, G een, Red, Nea in a ed, SWIR-1, and SWIR-
2), h ee spec al indices, and he DEM.
3.4.2. Image p e-p ocessing
P io o hei inges ion in GEE, he su ace e lec ance images om he h ee Landsa senso s we e a mosphe ically co ec ed
using he Landsa Ecosys em Dis u bance Adap i e P ocessing Sys em (LEDAPS) o he Land Su ace Re lec ance Code (LaSRC).
Also, he isible bands we e bands p ocessed o o ho ec i ied su ace e lec ance. The bands om Landsa 8 we e enamed o ma ch
hose in Landsa 5 and 7. I was impossible o ge cloud- ee Landsa images o he s udy a ea. As a esul , he clouds in all a ailable
images we e masked. Finally, o each yea , he comple e collec ion o images om he Landsa senso s was me ged using he
median il e . In his way, clean Landsa composi es we e ob ained o each yea om 2000 o 2020 o use as inpu s o he image
classi ica ion wo k.
3.4.3. T aining and es ing sample collec ion
T aining and es ing samples we e manually collec ed using he high- esolu ion o hopho os on Google Ea h (GE). The sample
collec ion p o ocol was used as he ollowing:
- Gene a e simple andom poin s wi hin he s udy a ea.
- Visually inspec he land use a all poin s wi h a leas 30m adius homogeneous neighbo hood, and accep / ejec based on
local knowledge.
- Spli samples in o aining and es ing se s.
3.4.4. Image classi ica ion
A supe ised classi ie (Random Fo es ) was used o he land-co e classi ica ion on a pixel-by-pixel basis. Apa om i s
a ailabili y in Google Ea h Engine, his classi ie was selec ed because hey a e widely used in land-co e classi ica ion (Jia e al.,
2014; Yu e al., 2013). The classi ica ion wo k low is p o ided in Figu e 3 below.
In o de o make he map o he land-co e classi ica ion o he G and Saloum he classi ied maps ha e been expo ed om GEE
o A cGIS 10.4. Fou (4) classes ha e been aken in o accoun s such as mang o e, o he ege a ion, buil and ba e sand, and wa e .
To access he ege a ion close o he sho eline, a bu e has been manually c ea ed o a dis ance o 1km om he sho eline.
Zooming o he classi ied map along he sho eline has been done o de ec a eas o g ea change.
3.4.5. Accu acy assessmen
The accu acy was es ed using an independen se o samples ha we e andomly selec ed om he aining and es ing samples
and compu ed he con usion ma ix o each classi ied map. The classi ica ion p ocedu e was done in Google Ea h Engine while
es ing p ocedu e we e ca ied ou in A cGIS 10.4. Fo accu acy, 65% o sampling poin s we e used o aining and 35% o es ing.
The accu acy was calcula ed using he ollowing o mula:
𝐀𝐜𝐜𝐮𝐫𝐚𝐜𝐲 (%) = 𝐓𝐨𝐭𝐚𝐥 𝐓𝐫𝐮𝐞 𝐕𝐚𝐥𝐮𝐞 𝐏𝐢𝐱𝐞𝐥𝐬
𝐓𝐨𝐭𝐚𝐥 𝐒𝐚𝐦𝐩𝐥𝐞 𝐕𝐚𝐥𝐮𝐞 𝐒𝐚𝐦𝐩𝐥𝐞𝐬 x 100 Eq(2)
In e na ional Jou nal o Cu en Science Resea ch and Re iew
ISSN: 2581-8341
Volume 08 Issue 10 Oc obe 2025
DOI: 10.47191/ijcs /V8-i10-18, Impac Fac o : 8.048
IJCSRR @ 2025
www.ijcs .o g
5054 *Co esponding Au ho : Ousmane Badji Volume 08 Issue 10 Oc obe 2025
A ailable a : www.ijcs .o g
Page No. 5048-5061
Figu e 3: Landsa image classi ica ion p ocess
4. RESULTS
4.1. Sho eline dynamic
4.1.1. Gene al obse a ion in he sho eline changes
The s udied segmen includes he Dji e coas (Sec o E) and goes as a as Dionewa (Sec o D), Niodio (Sec o C), Be en y
(Sec o B), and Djinack Ba a and Jinack Kaja a Island no he n coas o he Gambia (Sec o A). Be ween 1990 and 2020, e osion
and acc e ion occu ed in some places and he ecosys em is highly domina ed by e osion. Figu e 4 highligh s i e main sec ions
highly dynamic. Sec o A and E show Mode a e o High e osion and acc e ion. The sec o B, C, and D a e cha ac e ised by mode a e
o high e osion a some poin s. Fo his pu pose, an annual a e age e osion a e o 2.44 m is obse ed and an a e age acc e ion a e
o 1.84 m.
Figu e 1: Poin o e osion (Red) and acc e ion (Da k G een) along he G and Saloum Sho eline
Landsa 5, 7
& 8 (SR) TS
NASA SRTM
30m DEM
NDWI,
NDBI, &
NDVI
10-band image
composi e
T aining and
es ing samples
Google Ea h
Classi ica ion
(Random Fo es )
Accu acy
Assessmen
Classi ied Maps
Common aining
samples (65%)
Common es ing
samples (35%)
In e na ional Jou nal o Cu en Science Resea ch and Re iew
ISSN: 2581-8341
Volume 08 Issue 10 Oc obe 2025
DOI: 10.47191/ijcs /V8-i10-18, Impac Fac o : 8.048
IJCSRR @ 2025
www.ijcs .o g
5055 *Co esponding Au ho : Ousmane Badji Volume 08 Issue 10 Oc obe 2025
A ailable a : www.ijcs .o g
Page No. 5048-5061
4.1.2. Sec o ial Analysis
Table 3 shows in o ma ion ela ed o he a e o change in he sho eline occu ing in each sec ion. Sec ions A and E showed a
balance e osion o 4.13 ±0.47 and 1.62 ±0.47 espec i ely and acc e ion o 2.82 ±0.47 o bo h sec ions. Sec ions C and D a e
cha ac e ised by High a e o e osion wi h an a e age o 2.39 ±0.47 and 2.63 ±0.47, espec i ely. The a e age acc e ion o sec ions
C and D ange be ween 1.45 ±0.47 and 1.018 ±0.47, espec i ely. Sec ion B doesn’ show so much dynamic wi h an a e age e osion
and acc e ion o 1.41 ±0.47 and 1.12 ±0.47.
Table 3: Pa ame e s o sho eline dynamics calcula ed in each ansec
4.2. Vege a ion dynamic
4.2.1. Analysis o he changes in he whole ansbounda y we lands
Figu es 5 and 6 show ha om 2000 o 2020 he whole G and Saloum we lands expe ienced an inc ease in mang o e ege a ion
and a dec ease in he o he ege a ion. The igu es show es ima ed mang o e co e ages o 57867.61 ha and 66840.17 ha in 2000
and 2020 espec i ely. The co e age o he o he ege a ion has educed om 2000 o 2020 wi h an es ima ed co e age o 23483.18
ha o 16146.11 ha espec i ely. The accu acies o he classi ica ion a y be ween 97.51 % and 99.37 % .
Figu e 5: Map o he mang o e and o he ege a ion o 2000 and 2020
Region
A
B
C
D
E
T ansec
1-481
482-914
915- 1276
1277 -1391
1392 - 1490
Numbe o ansec
481
433
362
114
98
A e age Acc e ion (m/y )
2.82 ±0.47
1.12 ±0.47
1.45 ±0.47
1.02 ±0.47
2.82 ±0.47
A e age E osion (m/y )
-4.13
±0.47
-1.41
±0.47
-2.39
±0.47
-2.63 ±0.47
-1.62 ±0.47
Max. acc e ion (m/y ) ( ansec )
14.52
±0.47
2.7 ±0.47
2.8 ±0.47
2 ±0.47
4.98 ±0.47
Max. e osion (m/y ) ( ansec )
-47.28
±0.47
-4.53
±0.47
-12.02
±0.47
-9.09 ±0.47
-4.02 ±0.47
In e na ional Jou nal o Cu en Science Resea ch and Re iew
ISSN: 2581-8341
Volume 08 Issue 10 Oc obe 2025
DOI: 10.47191/ijcs /V8-i10-18, Impac Fac o : 8.048
IJCSRR @ 2025
www.ijcs .o g
5056 *Co esponding Au ho : Ousmane Badji Volume 08 Issue 10 Oc obe 2025
A ailable a : www.ijcs .o g
Page No. 5048-5061
Figu e 6: ege a ion dynamics o he G and Saloum
The esul ing map (Figu e 7) om he change de ec ion analysis shows an inc ease o mang o e no hwa d and a dec ease o
he o he ege a ion sou hwa d.
Figu e 7: Change de ec ion analysis o he ege a ion om 2000 o 2020
57868 58365 64097 63161 66840
23483 17931 18386 16580 16146
0
10000
20000
30000
40000
50000
60000
70000
80000
2000 2005 2010 2015 2020
AREA (HECTARE)
YEARS
Mang o e O he Vege a ions
