Internati
o
nal
Journal of Ele
c
trical
and Computer
Engineering
(IJE
CE)
V
o
l.
6, N
o
. 5
,
O
c
tob
e
r
201
6, p
p
. 2
106
~211
3
I
S
SN
: 208
8-8
7
0
8
,
D
O
I
:
10.115
91
/ij
ece.v6
i
5.1
048
2
2
106
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJECE
Monitoring of Landslides in
Mountainous Regions Based on
FEM Modelling and Rain
Gauge Measurements
Gian Qu
oc Anh
1
, Ng
uye
n
Dinh-
Chin
h
2
, Tran Duc
-
N
g
hia
3
, Tra
n
Duc-Ta
n
2
1
Department of Electronics,
Na
mdinh University
of Techno
log
y
Education
,
Viet
Nam
2
Electron
i
cs
and
Telecomm
unication Faculty
,
VN
U University
o
f
Engineering
and
Technolog
y
,
Viet Nam
3
Institut
e
of
Inf
o
rm
ation T
echn
o
log
y
, Vi
etn
a
m
e
se Acad
em
y
of
Scien
c
e
and
Te
ch
nolog
y, Vi
et
Na
m
Article Info
A
B
STRAC
T
Article histo
r
y:
Received
Mar 14, 2016
Rev
i
sed
Jun
2
,
2
016
Accepted
Jun 20, 2016
Vietnam
is
a
co
untr
y
heav
il
y i
n
fl
uenced b
y
cli
m
a
te chan
ge.
T
h
e eff
ect of
clim
at
e
change
l
eads
to
a
s
e
ri
es
of
dangerous phenomena, such as landslides.
Landslid
es occu
r not on
ly
in
th
e mount
ainous
province, bu
t also in Delta
provinces, wher
e hundreds of landslides
are rep
o
rted annually
in the North-
Western provinces of Vietnam. These
even
ts have catastrophic impact to the
com
m
unit
y
as w
e
ll
as the econo
m
y
. In
m
ountain
ous areas, th
e co
nditions for
lands
lides
to o
ccur ar
e m
e
t frequent
l
y
, es
pe
cia
l
l
y
aft
e
r hea
v
y
ra
ins
or
geological activ
ity
, causing har
m
to th
e community
as well as damaging or
destro
y
i
ng much needed infr
astru
c
ture and key
transport routes. However, in
Vietnam, inv
e
stment in mountainous re
gions has been often lo
wer than in
urban areas
. Th
e m
e
teorolog
y
m
onitoring and forecas
ting s
y
s
t
em
s
are il
l
equipped
and o
v
erload
ed, s
o
th
e
y
cann
o
t de
liv
e
r
earl
i
er
and m
o
re a
ccura
te
forecas
ts
for
co
m
p
lex weath
e
r
e
v
ents
, un
able
to
provide t
i
m
e
l
y
warnings
. I
t
can be seen that
in countries that
lands
lide of
ten o
ccur, r
e
sear
chers
have been
tr
ying to d
e
ve
lo
p low cost and
effic
i
en
t landsl
i
de de
te
ction s
y
stem
. This
paper precisely addressed the probl
ems
mentioned
,
b
y
designing and
im
plem
enting a
n
effici
ent and
reliab
l
e L
a
nds
lide M
onitoring
and Ea
r
l
y
Warning (LMnE) s
y
stem based on the
3G/2G mo
bile communication s
y
stem,
and a rain gauge at the field site
al
ong with a car
efully
FEM (finite elemen
t
method) simulation using th
e r
a
in density
infor
m
ation on th
e s
e
rver. Th
e
s
y
s
t
em
us
es
adv
a
nced p
r
oces
s
i
n
g
algor
ithms combining obtain
e
d data at
the
centr
al s
t
at
ion.
Keyword:
Land
slid
e
M
odel
l
i
n
g
Mo
n
itoring
Rain
ind
u
c
ed
Slo
p
e
stab
ility
Copyright ©
201
6 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
Tran Duc
-
Tan,
El
ect
roni
cs
an
d Tel
e
c
o
m
m
uni
cat
i
on Fac
u
l
t
y
,
Uni
v
ersi
t
y
o
f
En
gi
neeri
n
g
a
n
d Tec
h
nol
ogy
(
U
ET
),
Vi
et
na
m
Nat
i
onal
Un
i
v
ersi
t
y
,
E3 b
u
ildin
g, 1
4
4
X
u
a
n
T
h
uy
,
Cau Giay
, Ha Noi,
Viet Nam
.
Em
a
il: tan
t
d
@
v
n
u
.
edu
.
v
n
1.
INTRODUCTION
Land
slid
e is
on
e
o
f
t
h
e m
o
st d
i
saster
s h
a
ppen
i
ng
ar
oun
d th
e
w
o
r
l
d. Th
is d
i
saster h
a
s catastr
o
ph
ic
i
m
p
act to
th
e co
mm
u
n
ity as well as th
e econ
o
m
y [1
]. Land
slid
e
o
c
cu
rs
by th
e fallen m
o
v
e
m
e
n
t
s o
f
rock
, so
il,
and orga
nic materials unde
r the
gra
v
ity forc
e. The
r
e a
r
e
fou
r
group
s th
at
lead
to
t
h
is h
a
zard
:
rainfall ind
u
ced
l
a
ndsl
i
d
es
, ear
t
h
q
u
ake i
n
d
u
c
e
d l
a
n
d
sl
i
d
es,
end
o
g
enet
i
c
l
a
ndsl
i
des, a
nd
pre
-
exi
s
t
i
n
g l
a
ndsl
i
des.
Am
ong
o
f
these groups, it is about 90%
of land
slides i
s
trigge
red
by rainfall. In
m
ountainous area
s, the conditions for
landslides t
o
occur are m
e
t freque
ntly, espe
cially after
h
eav
y rain
s
o
r
g
e
olo
g
i
cal activ
ity
, cau
sing
h
a
rm to
th
e
com
m
uni
t
y
as wel
l
as dam
a
gi
n
g
or
dest
r
o
y
i
ng m
u
ch ne
eded
i
n
f
r
ast
r
u
c
t
u
re a
n
d
key
t
r
ans
p
ort
r
o
ut
es.
We
i
n
t
e
nd t
o
i
m
plem
ent
a real
-t
im
e l
a
ndsl
i
d
e
m
oni
t
o
ri
ng i
n
Vi
et
nam
,
wh
ere t
h
e a
n
n
u
al
dam
a
ge due
t
o
t
h
e
lan
d
s
lid
e is
v
e
r
y
h
i
g
h
[2
].
For
th
e techno
logy asp
ect,
w
i
r
e
l
e
ss sen
s
or
n
e
t
w
or
k (W
SN)
i
s
th
e m
o
st applicab
le
in these
area
s (i.e. di
ffic
u
lt to
access, a
n
d re
a
l
-tim
e requi
re
ment) [3]-[5].
In
[3], a c
o
m
p
le
te functional s
y
ste
m
Evaluation Warning : The document was created with Spire.PDF for Python.
IJECE
ISS
N
:
2088-8708
Mo
ni
t
o
ri
n
g
of
La
ndsl
i
des i
n
Mo
unt
ai
n
ous
R
e
gi
o
n
B
a
se
d
o
n
FE
M M
odel
l
i
ng
a
n
d
R
a
i
n
..
.. (
G
i
a
n
Q
u
oc
An
h)
2
107
con
s
i
s
t
i
ng o
f
5
0
ge
ol
o
g
i
cal
sens
ors a
nd
20
wi
rel
e
ss sen
s
o
r
n
odes
was d
e
pl
oy
ed i
n
Id
u
kki
, a di
st
ri
ct
in t
h
e
sout
hw
est
e
r
n
r
e
gi
o
n
of
Ke
ral
a
St
at
e, I
n
di
a,
a hi
ghl
y
l
a
ndslide prone
area. It ca
n
be
seen
th
at
th
is kin
d
of
syste
m
co
sts a
lo
t o
f
m
o
n
e
y
an
d
thu
s
, it i
s
d
i
fficu
lt to
ap
p
lied
p
opu
larly. In
[4
], th
e
au
tho
r
s
u
s
e the slip
surface localiz
ation m
e
thod for
detecting la
ndsli
de. T
h
eir
propose
d
network uses de
vi
c
e
s to detect the slope
m
ovem
e
nt
s and t
h
e
n
est
i
m
a
te t
h
e di
spl
ace
m
e
nt
s of se
ns
or
no
des em
bedde
d i
n
t
h
e sl
ope
. Si
m
i
l
a
r to [
3
]
,
i
t
co
sts a lo
t o
f
m
o
n
e
y an
d
th
u
s
, it is d
i
fficu
lt to
ap
p
lied
po
pu
larly. In
[5
], th
ey only p
r
o
p
o
s
ed
wireless
accelerom
eter
network to
de
tect the landslide. Althou
gh it is a
low-c
o
st sol
u
tion
but it can only
detect
l
a
ndsl
i
d
e
but
c
a
n n
o
t
p
r
edi
c
t
t
h
e rai
n
fal
l
i
nduce
d
l
a
n
d
sl
i
d
e. In
[6]
,
t
h
e s
a
t
u
rat
e
d
hy
d
r
a
u
l
i
c
con
d
u
ct
i
v
i
t
y
i
s
considere
d
as
a random
field and once coupled with M
onte-Carlo
sim
u
l
a
tio
n
s
, it is p
o
s
sib
l
e to
d
e
termin
e th
e
failu
re
p
r
ob
ab
i
lity an
d
th
en
d
e
du
ce t
h
e lan
d
s
li
d
e
risk
.
In
[7
], au
tho
r
also
inv
e
stig
at
e th
e in
filtratio
n
o
f
rainwater in
to slo
p
e
s and
thu
s
th
e d
e
p
e
n
d
en
ce of slo
p
e
stab
ility o
n
th
e water in
ltratio
n
.
Howev
e
r, th
ese
si
m
u
latio
n
s
in
[6
],[7
] are
n
o
t
robu
st enou
gh
to
m
o
d
e
l arb
itrary sh
ap
e
of
n
a
tu
ral slop
es.
In this
pa
per,
an efficient and reliable La
ndslid
e M
o
ni
t
o
r
i
ng a
nd Ea
rl
y
War
n
i
n
g (
L
M
n
E) sy
st
em
i
s
d
e
v
e
l
o
p
e
d
to
re
m
o
tely
m
o
n
i
t
o
r an
d
au
to
m
a
te th
e warn
i
n
g ab
ou
t th
e p
o
ssib
ilities o
f
th
e lan
d
s
lid
e. This is a
low-cost sol
u
tion when
a ra
in
ga
uge
ca
n be use
d
to
provi
de the
rai
n
intensity and accum
u
lation. This
in
fo
rm
atio
n
is
p
r
e-pro
cessed
b
y
u
s
i
n
g a m
i
c
r
o
c
on
tro
ller,
an
d th
en
tran
smitted
to
th
e serv
er t
h
rou
g
h
t
h
e 3G/2G
m
obi
l
e
co
m
m
u
n
i
cat
i
on
pr
ot
oc
ol
. The
rai
n
da
t
a
are br
o
u
g
h
t
i
n
t
o
a desi
gne
d FEM
m
odel
i
n
GE
O-
SLO
P
E t
o
ev
alu
a
te th
e facto
r
of safety. Th
is in
fo
rm
ati
o
n
is co
m
p
ared
with
th
e d
e
term
in
ed
th
reshold
s
in
o
r
d
e
r to
mak
e
an
alert ab
ou
t t
h
e
p
o
s
sib
ilities o
f
th
e land
slide. Th
e m
o
n
ito
red
d
a
ta can
b
e
o
b
s
erv
e
d b
y
the d
e
sign
ed
web
s
ite
and the ale
r
t message ca
n
be
receive
d by bot
h feature
and s
m
art
phones.
2.
R
E
SEARC
H M
ETHOD
2.
1.
Rain
induce
d
landslide
Th
e m
a
in
trig
g
e
r
o
f
lan
d
s
li
des is h
eav
y
o
r
p
r
o
l
on
ged
rai
n
fall [6
],[7
].
Fi
g
u
re 1
illu
st
rate o
f
a sl
o
p
e
t
h
at
i
s
unde
r t
h
e ri
sk
of l
a
nd
sl
i
d
e. The sl
o
p
e
can be di
vi
d
e
d i
n
t
o
t
w
o re
gi
o
n
s w
h
i
c
h a
r
e safe an
d p
o
t
e
nt
i
a
l
sl
i
d
e areas
. T
h
eref
ore
,
o
u
r
W
S
N i
s
desi
g
n
e
d
based
o
n
t
w
o area
s. T
h
e s
e
nsi
n
g a
n
d
t
r
a
n
sm
i
t
t
i
ng m
odul
es are
placed i
n
t
h
e
potential slide a
r
ea, a
n
d the
storing m
odul
e a
n
d the
rai
n
gauge are
placed in
the safe
area
.
Fig
u
re
1
.
Illu
st
ratio
n of a
slope th
at is
u
n
d
e
r
th
e risk
o
f
landslid
e
2.
2.
Modeling
the ra
in
induced
l
a
ndslid
e
usin
g
finite eleme
n
t
meth
od
Th
ere are two
k
i
nd
s of sh
ear
stresses ex
isted
in
a slop
e: on
e th
at ho
ld
t
h
e slo
p
e
an
d
one th
at sp
lit th
e
slope
d
o
w
n
(se
e
Fig
u
re
2
)
. T
h
e s
h
ear
stre
ng
th Fs
(i.e.
ho
ldin
g
st
ress) is
determin
ed
b
y
th
e
n
o
rm
al stress FN
and
t
h
e c
o
hesi
on
. T
h
e s
h
ea
r s
t
ress (i
.e
.
do
w
n
-sl
opi
ng
st
res
s
) i
s
det
e
rm
i
n
ed
by
t
h
e a
n
gl
e
of
t
h
e sl
ope
ɵ
and
the
weigh
t
o
f
th
e
p
o
t
en
tial
m
a
ter
i
al. Th
e ratio
of sh
ear streng
t
h
to
sh
ear stress is called
th
e
facto
r
o
f
safety (FS).
Wh
en
th
is ratio
is g
r
eater th
an
1
,
sh
ear streng
th
is g
r
eat
e
r
than s
h
ear stres
s
and the
slop
e is co
n
s
id
ered
stab
le.
Wh
en
th
is rati
o
is clo
s
e to
1, sh
ear streng
th is n
early
equa
l
t
o
shear st
res
s
and t
h
e sl
o
p
e
i
s
unst
a
bl
e (i
.
e
. t
h
e
lan
d
s
lid
e
wou
l
d
o
c
cu
r) [8
]. No
te th
at h
ill slo
p
e
s
(e.g
. the safe reg
i
o
n
i
n
Figu
re
1
)
are
m
o
re stab
le th
an
th
e
ot
he
rs.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 5
,
O
c
tob
e
r
20
16
:
210
6
–
21
13
2
108
Fig
u
re
2
.
Illu
st
ratio
n of th
e sl
o
p
e
stab
ility u
s
in
g
Factor
o
f
Safety (FS), where
W is
h
e
igh
t
of water tab
l
e
Sl
ope
s w
oul
d
be wea
k
e
n
ed
by
seve
ral
rea
s
on
s suc
h
as t
h
e de
f
o
rest
at
i
o
n, t
h
e
weat
he
r
i
ng, a
n
d t
h
e
u
n
d
e
rcu
tting
by riv
e
r fl
o
w
, wav
e
s,
o
r
hu
m
a
n
activ
ities. Howev
e
r, t
h
e m
o
st i
m
p
o
r
tan
t
trigg
e
r
for a land
sl
id
e is
th
e rainfall.
Wh
en
it rai
n
s, the water infiltrates in
to
th
e
g
r
o
und
. Th
e infiltratio
n
p
r
o
cess h
a
s h
a
pp
ened
faster
th
an
t
h
e
d
r
ai
n
p
r
o
cess. Th
us,
th
e pore sp
aces in
t
h
e sl
o
p
e
are filled with
water. It
lead
s to
a h
ydrau
lic u
p
lift
force that re
duces the
bala
nce that we
have m
e
n
tio
ned
prev
iou
s
ly. Obv
i
ou
sly, t
h
e pred
ictio
n o
r
t
h
e
measu
r
em
en
t of rai
n
fall
o
n
an actu
a
l slop
e is v
e
ry im
p
o
r
tant to
d
e
term
in
e t
h
e
p
r
ob
ab
ility o
f
th
e lan
d
s
li
de.
Th
e m
a
ss con
s
erv
a
tion
equ
a
tio
n in
a fo
rm
exten
d
e
d
to un
sat
u
rated
co
nd
ition
s
is
()
(
)
xy
w
m
H
y
HH
kk
Q
m
xx
y
y
t
(1
)
wh
ere H is to
t
a
l h
ead
(m
), k
x
is th
e co
effi
cien
t o
f
p
e
rm
eab
ility with
resp
ect to
th
e
water p
h
ase in
th
e x-
d
i
rection
(m
/s), k
y
is
th
e co
efficien
t o
f
p
e
rm
eab
ility with
resp
ect to
th
e wat
e
r ph
ase in
th
e y-d
i
rection
(m
/
s
), Q
i
s
t
h
e bo
un
dar
y
fl
ux passi
n
g
i
n
or o
u
t
o
f
an
el
em
ent
a
ry
cube (i
n t
h
i
s
case
an el
em
ent
a
ry
squa
re, gi
ven t
h
at
t
h
e
eq
u
a
tion
is in
two-d
i
m
e
n
s
io
ns) (m
2
/m
2
s),
w
m
is the specific water capacity,
m
is th
e u
n
it wei
g
h
t
of water,
an
d
t is tim
e (
s
). Th
e
b
oun
d
a
ry con
d
ition
s
fo
r t
h
e water-head
and
t
h
e fl
ux
are
n
e
ed
ed
t
o
so
lv
e th
is equ
a
tion
[9
].
Furth
e
rm
o
r
e, t
h
e in
itial co
nd
itio
n of t
h
e
water
h
e
ad d
i
stribu
tion
i
s
also
n
e
ed
to so
l
v
e th
is equ
a
tio
n.
Po
sitiv
e and
neg
a
tiv
e po
re
water p
r
essure distrib
u
tion
s
ob
t
a
in
ed
b
y
th
e seep
ag
e an
alysis SEEP/
W
[10] are
th
en
u
s
ed
as inp
u
t
d
a
ta fo
r the stab
ility an
al
ysis (i.e.
th
e determin
atio
n
of FS). Th
e equatio
n
fo
r t
h
e facto
r
of
safety is
(c
o
s
)
t
a
n
sin
w
cg
H
g
W
Shear
St
rengt
h
FS
Shear
St
ress
gH
(2
)
whe
r
e c is cohesion,
is densit
y of regolith,
g
is gravity acceleration,
w
i
s
densi
t
y
of wat
e
r
,
and
i
s
angl
e
of i
n
t
e
r
n
al
f
r
i
c
t
i
on.
Thi
s
c
o
m
put
at
i
on i
s
per
f
o
r
m
e
d
u
s
i
n
g th
e
SLOPE/
W
.
No
te that, bo
th
SEEP/W and
SLOPE/
W
prob
lem
s
were so
lv
ed b
y
th
e
fin
i
t
e
elem
en
t
m
e
th
o
d
(FEM). Fi
gu
re 3 illu
strates of a m
e
sh
ed
m
o
d
e
l
o
f
a
p
o
t
en
tial slid
e slop
e in GEO-SLOPE.
Fi
gu
re
3.
M
e
sh
i
ng
o
f
a
2
D
hy
dr
o
g
eol
ogi
cal
conce
p
t
u
al
m
odel
t
h
at
c
o
nsi
s
t
o
f
t
w
o
di
f
f
ere
n
t
l
a
y
e
rs
F
N
F
S
W
Evaluation Warning : The document was created with Spire.PDF for Python.
IJECE
ISS
N
:
2088-8708
Mo
ni
t
o
ri
n
g
of
La
ndsl
i
des i
n
Mo
unt
ai
n
ous
R
e
gi
o
n
B
a
se
d
o
n
FE
M M
odel
l
i
ng
a
n
d
R
a
i
n
..
.. (
G
i
a
n
Q
u
oc
An
h)
2
109
3.
PROP
OSE
D
SYSTE
M
3.
1.
Integ
r
ation system
a
n
d its w
o
rking
princi
ple
Fi
gu
re 4 sh
o
w
s t
h
e di
agram
of t
h
e p
r
op
ose
d
sy
st
em
appl
i
e
d fo
r a pot
e
n
t
i
a
l
sl
i
d
e sl
ope. Aft
e
r a fi
el
d
in
v
e
stig
ation
,
a real m
o
d
e
l o
f
t
h
e slop
e is
bu
ilt for
SEEP/
W
an
d SLOPE/
W si
m
u
latio
n
.
M
u
ltip
le scen
ario
s are
run
with
d
i
fferen
t
v
a
lu
e of rain
in
ten
s
ities. It will h
e
lp
u
s
t
o
d
e
term
in
e th
e th
resho
l
d
o
f
rain
in
ten
s
ities th
at
lead
to
lan
d
s
li
d
e
. Howev
e
r,
in
real ap
p
licatio
n
,
th
e rai
n
in
ten
s
ity
m
a
y
b
e
ch
an
g
e
d
du
e to
ti
m
e
. T
h
u
s
, a
p
r
ed
ictio
n and
m
o
n
ito
rin
g
are propo
sed as
follo
ws:
Fi
gu
re
4.
The
di
ag
ram
of t
h
e
pr
o
pose
d
sy
st
e
m
Case 1
(no
rai
n
o
r
ligh
t
rain
): th
e
v
a
lu
e
o
f
th
e rai
n
g
a
ug
e
will d
e
term
in
e th
is situ
atio
n.
In
th
is case,
t
h
e dat
a
fr
om
t
h
e rai
n
gau
g
e
are acqui
re
d
and f
o
r
w
ar
d t
o
a web dat
a
base. T
h
ere i
s
no nee
d
t
o
execut
e
SEEP/
W
or
SLOPE/
W in th
is
case.
Case 2 (rain
i
n
ten
s
ity is b
e
low th
e alert
regio
n
):
Not only the
data
from
the rain ga
uge are
ac
qui
re
d
th
en
fo
rward
t
o
a web
d
a
tabase, bu
t also
the SEEP/
W
or
SLOPE/
W
is execute
d eve
r
y five hours
. Note that,
th
e ou
tpu
t
of th
e prev
iou
s
si
m
u
la
tio
n
ti
m
e
is th
e in
itiatio
n
p
a
ram
e
ters fo
r t
h
e n
e
x
t
run
with
t
h
e real-ti
m
e
u
p
d
a
ted rai
n
inten
s
ity fro
m
th
e rain g
a
ug
e.
Case 3
(rain
i
n
ten
s
ity is in
th
e alert reg
i
o
n
): No
t
o
n
l
y th
e
data fro
m
th
e rain
g
a
u
g
e
are acq
u
i
red
with
h
i
gh
er data rate th
en
forward to
a web
d
a
tab
a
se, bu
t also
the SEEP/
W
or SLOPE/
W
is
executed e
v
ery one
h
ours to
p
r
ed
ict th
e ti
m
e
th
at l
a
n
d
slid
e m
a
y o
ccu
r. Th
e
ou
tpu
t
of th
e
p
r
ev
io
u
s
sim
u
lat
i
o
n
ti
m
e
is th
e in
iti
atio
n
p
a
ram
e
ters
for th
e
n
e
x
t
run
wi
th
th
e real-tim
e
up
d
a
ted
rain
i
n
ten
s
ity
fro
m
th
e rain
g
a
u
g
e
.
3.
2.
Weat
her sta
t
i
o
n
WS-30
00
Weathe
r Station
W
S
-3000 is a kit that co
mprises
three sensors: wind ga
uge
, anem
o
m
e
t
er and wi
nd
vane
(see Figure 5). It is a rel
i
able and acc
urate weathe
r
ki
t
avai
l
a
bl
e fo
r ju
st
a fract
i
o
n
of t
h
e
pri
ce
of
ot
he
r
st
anda
rd
weat
h
e
r st
at
i
ons
[1
1]
. Som
e
expe
ri
m
e
nt
s are pe
rf
orm
e
d t
o
t
e
st
t
h
e pe
rf
o
r
m
a
nce of t
h
e Li
bel
i
um
W
S
-
30
0
0
agai
nst
o
n
e o
f
t
h
e
W
eat
her St
at
i
o
ns f
r
o
m
Davi
s and
a pl
u
v
i
o
m
e
t
e
r
fr
om
R
a
i
n
’o
’
M
at
i
c
(bot
h
de
vi
ces are
kn
o
w
n
f
o
r
bei
n
g acc
urat
e,
reli
ab
le, an
d exp
e
n
s
iv
e).
Fig
u
r
e
5
.
Th
e w
eath
e
r
station W
S
-3
000
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 5
,
O
c
tob
e
r
20
16
:
210
6
–
21
13
2
110
3.
3.
Micropr
ocess
o
r
and GSM/
GPRS m
o
dule
In t
h
i
s
resea
r
c
h
, t
h
e At
m
e
ga32
8 m
i
crop
roc
e
ssor
i
s
a m
e
m
b
er of the
hi
gh-perform
ance Atm
e
l 8-bit
AVR R
I
SC-ba
s
ed m
i
crocontrollers [12]. It i
s
use
d
to
recei
ve a
nd
proces
s
observe
d
data
from
W
S
-3000. T
h
e
At
m
e
ga328
ha
ve m
a
ny
special
feat
ures s
u
i
t
a
bl
e fo
r b
u
i
l
d
i
ng a
WS
N, su
ch as hi
g
h
p
e
r
f
o
rm
ance, l
o
w
po
we
r,
ad
v
a
n
c
ed
RISC ar
ch
itectur
e
u
p
to
20
MIPS
th
ro
ugh
pu
t at
20
MH
z an
d 23
p
r
og
r
a
mm
ab
le I
/
O
lin
es [13
]
.
In the case
of
giving ale
r
t, the syste
m
automa
tically sends an SMS m
e
ssage t
h
rough
GSM/GPR
S
[14] to t
h
e
responsi
v
e c
o
ntact to act in tim
e (see Figure
6).
Fi
gu
re
6.
The
ph
ot
o
o
f
t
h
e
G
S
M
/
GPR
S
:
S
I
M
9
0
0
m
odel
4.
RESULTS
4.
1.
Estima
tion
o
f
FS usin
g
FEM
In t
h
e ce
nt
ral
com
put
er, t
h
e G
e
o Sl
o
p
e so
ft
w
a
re i
s
set
up. Fi
gu
re 7 s
h
o
w
s t
h
e di
st
ri
b
u
t
i
o
n m
a
p of t
h
e
p
o
re water pressu
re with
t
h
e rain
fall d
e
n
s
ity o
f
12
.6
mm/h
.
In
o
r
d
e
r to
un
d
e
rstand th
e stab
ility
o
f
t
h
e
lan
d
s
lid
e site un
d
e
r d
i
fferen
t
rain
fall in
ten
s
it
y were in
troduced
to
th
e m
o
d
e
l to
co
n
d
u
c
t th
e cou
p
l
ed
stab
ility
analysis. The
factor
of sa
fety was co
m
p
u
t
ed
an
d
can
b
e
used
to
estab
lish
th
e relation
s
h
i
p b
e
tween
th
e
rain
fall
in
ten
s
ity and
sl
o
p
e
stab
ility.
Fi
gu
re 7.
Di
st
ri
but
i
o
n
m
a
p
o
f
t
h
e po
re wat
e
r press
u
re wi
t
h
t
h
e rai
n
fal
l
de
n
s
i
t
y
of 12
.6
m
m
/
h
Fi
gu
re 8 s
h
ow
s t
h
e rel
a
t
i
ons
h
i
p bet
w
ee
n t
h
e
rai
n
fal
l
i
n
t
e
nsi
t
y
and t
h
e fact
or
of sa
fet
y
(F
S). It
ca
n be
seen t
h
at the
sl
ope
stability decreased as
the
rainfall
am
ount
inc
r
ease
d
. T
h
e value
of FS
significantly
reduce
d
in
a lin
ear tren
d wh
en
th
e
rain
fall in
ten
s
it
y in
creased
fro
m
3
to
6 mm/h
ou
r. Howev
e
r, th
e p
ace of
d
eclin
e
eased significa
n
tly after th
e rainfall intensity reached
to
6.7 mm/hour. As the above se
ction desc
ribe
d, the
g
r
ou
nd
water at th
e lan
d
s
lid
e site is
m
a
in
ly lo
cated
in
t
h
e
fract
u
r
e
d
wea
t
here
d be
dr
oc
k
.
M
o
st
rai
n
fal
l
m
a
y
flow downward to th
e
hillside through t
h
e fractures a
n
d di
fficu
lt to
rem
a
in in the pore space.
As we
men
tio
n
e
d in
Sectio
n
3
,
th
e rainfall in
ten
s
ity fro
m
3
to
6
mm/h
o
u
r is
co
rresp
ond
s to th
e case
2
,
an
d the
rain
fall i
n
ten
s
ity eq
u
a
l
or larger th
an
6
.
7
mm/h
o
u
r
is corresp
ond
s t
o
th
e case 3.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Mo
ni
t
o
ri
n
g
of
La
ndsl
i
des i
n
Mo
unt
ai
n
ous
R
e
gi
o
n
B
a
se
d
o
n
FE
M M
odel
l
i
ng
a
n
d
R
a
i
n
..
.. (
G
i
a
n
Q
u
oc
An
h)
2
111
Fig
u
re
8
.
Th
e relatio
n
s
h
i
p
b
e
tween rai
n
fall co
nd
itio
ns an
d facto
r
o
f
safety, th
e
rain
p
a
ttern
is
un
iform
.
4.
2.
D
a
t
a
mo
n
i
to
r
i
n
g
a
n
d
wa
rn
in
g
me
c
h
a
n
is
m
4.
2.
1.
Real
-time data
m
o
nitorin
g
using
web
interfa
c
e and
a
mo
bile applicatio
n
Th
e
d
a
ta acqu
i
red
i
n
th
e
field
site will tran
smit wirelessly
fro
m
W
S
-300
0 to
th
e
g
a
teway
,
and
th
en
t
o
th
e d
a
ta lo
gg
er. Con
s
equ
e
n
tly, th
e d
a
ta is
au
to
m
a
tical
ly
u
p
l
o
a
d
e
d
to
a MySQL d
a
tabase o
n
a web
serv
er.
Fin
a
lly, th
e web
app
licatio
n
fo
r
rem
o
te
m
o
n
ito
ring
is
bu
ilt u
s
ing
web
services (see Figu
re 9
)
. After th
at
, th
e
dat
a
i
s
neede
d
t
o
p
r
ocess t
o
gi
ve al
ert
or
n
o
t
[1
5]
.
Fi
gu
re 9.
R
eal
-
t
im
e
dat
a
m
oni
t
o
ri
n
g
usi
n
g w
e
b
i
n
t
e
rface
4.
2.
2.
Warni
n
g SM
S
t
o
any
ph
one
To
g
i
v
e
a warn
ing
abou
t th
e lev
e
ls o
f
land
slid
e
to
an
y p
hon
e (
i
.e. smar
t
or feathe
r phones),
a
GSM
/
GPR
S
m
odul
e i
s
con
n
ect
ed t
o
t
h
e
cent
r
al
com
p
u
t
er. I
n
t
h
e cas
e of
war
n
i
n
g,
an SM
S
wo
ul
d sen
d
au
to
m
a
tical
ly t
o
th
e d
e
sign
ed
p
hon
e (
s
ee
Fi
gu
r
e
10
).
5.
CO
NCL
USI
O
NS
In
th
is p
a
p
e
r, a si
m
p
le an
d
effectiv
e system is
d
e
v
e
lop
e
d
to
rem
o
tely
m
o
n
ito
r an
d
au
to
m
a
te
th
e
warn
ing
ab
ou
t
th
e po
ssib
ilities o
f
t
h
e land
slid
e. Th
is is
a low-co
st so
l
u
tion
wh
en
a rain
g
a
ug
e can
b
e
used
to
provide the rai
n
intensity and accum
u
la
tion. This inform
ation is e
x
tracted
and
brought to
a designed m
odel in
GEO-SLOPE to
ev
al
u
a
te th
e
facto
r
of safety
. Th
is info
rm
at
io
n
is co
m
p
ared
with
t
h
e
d
e
term
in
ed
th
resh
old
s
i
n
o
r
d
e
r to
m
a
k
e
an
alert ab
ou
t th
e p
o
s
sib
ilitie
s o
f
th
e land
slid
e. Th
is m
o
n
ito
ri
n
g
system i
s
co
m
p
leted
,
si
m
p
le,
low-cost, and e
ffective
co
m
p
ared
t
o
ot
he
r
wo
rks
as s
h
ow
n i
n
[3]
-
[
7]
.
2
4
6
8
10
12
14
0.
5
0.
6
0.
7
0.
8
0.
9
1
1.
1
1.
2
1.
3
1.
4
1.
5
R
a
in
in
t
e
n
s
it
y
(
m
m/
h
)
F
a
c
t
or
of
s
a
f
e
t
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 5
,
O
c
tob
e
r
20
16
:
210
6
–
21
13
2
112
Figure
10. Rea
l
-tim
e alert to the
phone
ACKNOWLE
DGE
M
ENTS
Thi
s
wo
rk
ha
s
bee
n
s
u
p
p
o
rt
e
d
by
Vi
et
nam
Nat
i
onal
U
n
i
v
ersi
t
y
, Ha
noi
(
V
N
U
)
u
n
d
er
P
r
o
j
ect
No
.
QG
. 14. 05.
REFERE
NC
ES
[1]
Glade,
et al.
, “Landslide hazard
and risk,”
John
Wiley
& Sons, 2
006.
[2]
Petley
D. N., “The global occu
rr
en
ce of fatal lan
d
slides in 2007,”
Geophysical research abstracts
, vol. 10, pp. 3
,
2008.
[3]
Ramesh M. V., “Design, develo
pment,
and deplo
y
ment of a wir
e
less sensor
network for detection of landslides,”
Ad Hoc N
e
twork
s
, vol. 13, pp. 2-
18, 2014
.
[4]
Terz
is
A
.,
et a
l
.
,
“
S
lip s
u
rface
lo
cal
iza
tion
in wir
e
les
s
s
e
ns
or net
w
orks
for lands
l
i
de pred
ic
tion
,
”
in
Proceedings of
the 5
t
h in
ternational con
f
eren
ce
on Info
rmation p
r
ocessing in sen
s
or networks
, vo
l. 5
,
pp
. 109-116
, 2006
.
[5]
Kotta H.
Z.
,
et al.
, “Wireless sensor network for landslid
e m
onitoring in Nu
sa Tenggar
a
Timur
,
”
Indonesia
n
Journal of Electr
ical Engineerin
g
, vol/issue: 9(1)
, pp
. 9-18
, 2011
.
[6]
Ali A.,
et al.
, “
S
im
plified quant
itat
i
ve risk assessm
ent of rainfa
ll-induced landslides modele
d b
y
infinite slopes,”
Engineering Geology
, vol. 179
,
pp.102-116, 201
4.
[7]
Collins B. D.
and Znid
arci
c D., “Stabilit
y
an
al
yses of rainf
a
ll
i
nduced
landslid
es
,
”
Journal
o
f
Geotechn
i
cal
an
d
Geoenvironmental Eng
i
neering
,
vol/issue:
130(4)
, pp
. 362-372
, 2
004.
[8]
Fredlund,
et al.
, “Soil
mechanics for
unsaturated
soils,” John W
iley
& Sons, 1993.
[9]
Mei-hai,
et a
l
.
,
“
R
es
earch on
S
liding M
e
chan
is
m
and Tre
a
tm
e
n
t Measure of
Slope with
Thick Accumulation
Horizon under
R
a
infall Cond
ition
,
”
Journal o
f
Co
nvergence In
formation Techno
lo
gy
, vo
l/issue: 8(
8), 2013
.
[10]
T. D. Fang, “
A
Slope Stabilit
y
Anal
y
s
is Metho
d
Based
on Unsatura
ted Seepag
e of Sl
ope and its Com
p
arison with
Geo-Seep Softw
a
re,”
Applied M
e
chanics and
Materials
, vo
l. 540,
2014.
[11]
Libel
i
um
websit
e. [Onlin
e]
, Avaiab
le: h
ttp:
//l
ib
eliu
m.
c
o
m We
at
he
r St
at
i
ons Comp
arative –
W
eather M
onit
o
r
(Davis) vs. WS-3000, T
echn
i
cal report,
2014.
[12]
N.
D.
Chinh,
et al.
, “
M
ulti-sensors integration f
o
r landslide m
onitoring appl
ic
at
ion,”
VNU Jour
nal of Science –
Natural Science
and Technolog
y
, vol/issue: 30(6S
-B), pp
. 202-210
, 2014
.
[13]
K. A. Hoshang and L.
Tang
, “ATmeg
a and XBee-
b
ased wireless sensing,”
in Pro
c
.
IEEE
Inter
natio
nal Conferen
ce
on Automation,
Robotics
and Ap
plications
, pp
. 3
51-356, 2011
.
[14]
Chang,
et al.
, “Application of
wireless sensor
network and gpr
s technolog
y
in
de
velopment o
f
remote monitoring
sy
s
t
e
m
,
”
Indonesian Journal o
f
Electrica
l
En
g
i
n
eering and
Computer Science
, v
o
l/issue: 1
3
(1), p
p
. 151-158
, 201
5.
[15]
H
u
W
.,
et al.
, “Gross Error De
noising Method
for Slope
Monitoring Data at
H
y
dropower Station,”
Indon
esia
n
Journal of Electr
ical Engineerin
g
and Computer
Scien
c
e
, vol/issue: 11(10)
, pp
. 55
45-5552, 2013
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Mo
ni
t
o
ri
n
g
of
La
ndsl
i
des i
n
Mo
unt
ai
n
ous
R
e
gi
o
n
B
a
se
d
o
n
FE
M M
odel
l
i
ng
a
n
d
R
a
i
n
..
.. (
G
i
a
n
Q
u
oc
An
h)
2
113
BIOGRAP
HI
ES OF
AUTH
ORS
Gian Quoc An
h
was born in
1981. He r
e
ceived th
e
B.S. d
e
gree in Ph
y
s
ics from VNU-
University
of
Science in
200
3 and M.S.
de
gree
in E
l
e
c
tro
n
ics and
Te
le
c
o
m
m
unication
techno
log
y
fro
m VNU-
University
of En
gineering and
Techn
o
log
y
(UET)
in 2010. He is
currently
working towards the Ph.
D
.
degree
in
Electronic Engineer
ing at
VNU-UET.
His
research
in
terest
s are
appl
ica
tion
s
of di
gital signal processing and
embedded s
y
s
t
ems.
Nguy
en Dinh Chinh
was bor
n in 1992. He r
eceived th
e bachelor degr
ee
in
Electronics and
Telecommunication at Univ
ersity
of
Engineer
in
g and Technolo
g
y
(UET) in 2
014. He is
a
research
er at University
of Eng
i
neering
and T
echnolog
y
.
He aw
arded th
e Best Paper Award at
the Organi
zing
Com
m
ittee of t
h
e Intern
ation
a
l
Conference on
Green and Human Information
Techno
log
y
(IC
GHIT 2015). He is au
thor and
co
-author
of several p
a
pers on
MEMS based
sensors, their
ap
plic
ations
and a
pplic
ation o
f
m
u
lti-sensor in
tegr
a
t
ed s
y
s
t
em
s base
d on W
i
rel
e
ss
Sensor Network
techn
i
que.
Tran Duc Nghi
a
was born in 1986. He receiv
ed his B.
Sc and
M.Sc degrees r
e
spectiv
ely
in
2010, and 2015 at the Vietn
a
m
Maritim
e Universit
y
(VI
M
ARU), and the Universit
y
of
Engineering and
Technolog
y
(U
ET), Vietnam
Na
tional Univer
sity
– Hano
i (VNUH). He is
current
l
y
a r
e
s
e
a
r
cher a
t
th
e Ins
t
i
t
ute of Inform
a
t
i
on Techno
log
y
(
I
OIT), Vie
t
nam
Academ
y
of
Science and
Technolog
y
(VAST). He h
a
s been pa
rticipating
in the implemen
tation
of some
projects in information processing.
He is
the aut
hor and coautho
r
of some papers on Biomedical
engineering and
Bioinformatics. His present res
earch in
teres
t
i
s
in m
a
them
ati
c
s
and s
i
gnal
processing app
l
ications.
Tran Duc-Tan
received h
i
s B.Sc, M.Sc,
and PhD. degr
ees respectively
in 2002, 2
005, and 2010
at th
e University of Engineering
and Technol
og
y (UET), Vietnam National University
– Hanoi,
Vietnam (VNUH), where h
e
has been a lectur
er
si
nce 2006
. He
was the recip
i
ent of the Vietn
a
m
National Univer
sity
, Hano
i, Vietnam Young S
c
ie
n
tific Award
in 2008
. He is currently
an
As
s
o
ciate P
r
ofe
s
s
o
r with the Facul
t
y
of
Ele
c
tr
onics and Telecommunications, University
of
Engineering
and
Technolog
y
,
Viet
nam National
Univ
ersity
, Han
o
i, Vietnam. H
e
is the autho
r
and co
author o
f
50 papers on
MEMS based sens
or
s
and th
eir
ap
plic
ation
.
His
pr
es
ent r
e
s
ear
ch
inter
e
st is
in DS
P appli
cat
ions.
Evaluation Warning : The document was created with Spire.PDF for Python.