Indonesi
an
Journa
l
of El
ect
ri
cal Engineer
ing
an
d
Comp
ut
er
Scie
nce
Vo
l.
1
3
,
No.
3
,
Ma
rch
201
9
, p
p.
1
259
~
1
2
6
6
IS
S
N: 25
02
-
4752, DO
I: 10
.11
591/ijeecs
.v1
3
.i
3
.pp
1
259
-
1
2
6
6
1259
Journ
al h
om
e
page
:
http:
//
ia
es
core.c
om/j
ourn
als/i
ndex.
ph
p/ij
eecs
Rainf
all
–
l
andsli
de
e
arly
w
arnin
g
system (RLEWS
) using
TRMM
p
recipita
tion
e
sti
mates
No
rs
uz
il
a
Y
a’acob
1
, No
rais
yah Taju
din
2
,
Az
ie
an
Moh
d
Az
i
z
e
3
1
,2
Facul
t
y
of Electrical E
ng
ine
er
i
ng,
Univer
si
ti T
e
knologi
MA
RA, 404500 Shah
Al
am Sel
angor
,
M
al
a
y
si
a
1
W
ire
le
ss
Com
muni
cation Techn
olog
y
(W
iCoT
)
,
Facul
t
y
of Electr
ic
a
l
Eng
ineeri
ng
,
Univer
si
ti
T
ekn
ologi
MA
RA,
404500
Shah
Al
am Sel
angor
,
M
al
a
y
si
a
3
Univer
siti
Te
kn
ika
l
Malay
si
a
M
el
ak
a, 76100
Du
ria
n
Tungga
l
,
M
el
ak
a
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Sep
15
, 201
8
Re
vised
N
ov
3
0,
2018
Accepte
d Dec
2
0
, 201
8
Thi
s
pap
er
pre
s
ent
s
Rai
nf
all
–
L
andsli
de
Ea
r
l
y
W
arn
ing
S
y
st
e
m
(RLE
W
S)
using
Tropica
l
Rai
nfa
ll
Me
asuring
Miss
ion
(TR
MM
)
pre
ci
pi
ta
t
i
on
esti
m
ates
to
noti
f
y
the
war
ning
le
v
el
for
th
e
poss
ibi
li
t
y
of
l
andsli
de
o
cc
urr
e
nce
s
in
Ulu
Kela
ng,
Sela
ng
or.
In
th
is
stud
y
,
RL
EWS
is
deve
lop
ed
to
m
onit
or
the
poss
ibi
li
t
y
of
r
ainfal
l
-
induc
ed
lands
li
de
oc
cur
ren
ce
s
b
y
compari
n
g
rea
l
ti
m
e
TRMM
rai
nfa
ll
dat
a
with
a
l
andsli
de
r
ai
nfa
ll
thr
e
shold.
The
l
ands
li
de
r
ai
nfa
ll
thre
shold
is
co
nstruct
ed
b
y
using
the
accum
ula
t
ed
ra
infa
l
l
-
a
cc
um
ula
t
ed
rai
nfa
ll
(E
-
E)
di
agr
am
m
et
hod.
The
warni
ng
le
v
el
s
of
rai
nfa
l
l
thre
shold
ar
e
cl
assifi
e
d
int
o
t
hre
e
l
evels;
hig
h,
m
oder
ate
an
d
low.
Th
e
an
aly
s
is
an
d
noti
ficat
ion
are
updat
ing
eve
r
y
24
hours
to
p
rovide
the
initi
al
pote
nt
ial
la
ndslide
info
r
m
at
ion
signal.
The
ra
infall
thr
eshold
anal
y
s
is
was
abl
e
to
det
e
ct
th
e
e
arly
signal
of
in
it
i
al
pote
ntial
la
ndsli
de
occ
urr
ences.
The
a
ims
of
thi
s
stud
y
ar
e
to
deve
lop
a
low
-
cost,
sus
ta
in
able
ea
rl
y
w
arn
ing
s
y
stem
an
d
web
base
appl
i
c
at
ion
to
send
no
t
ifi
c
at
ion
and
awa
ren
ess
for
r
esid
ent
i
al
are
a
s
in
Ulu
Ke
la
ng
,
S
el
angor
.
Ke
yw
or
d
s
:
Early
w
a
rn
i
ng
s
yst
e
m
Lan
ds
li
de
Ra
infall
t
hr
e
shold
Rem
ote
s
ensin
g
TRM
M
Copyright
©
201
9
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
Nors
uzila
Y
a
’a
cob,
Faculty
of Elec
tric
al
Engineer
ing
,
Un
i
ver
sit
i Te
knol
og
i M
ARA
,
404500
Sh
a
h Alam
Selang
or
, Mal
ay
sia
.
Em
a
il
: no
rs
uzila
@salam
.u
itm
.
edu.m
y
1.
INTROD
U
CTION
Lan
ds
li
des
a
re
def
i
ned
as
th
e
m
ov
e
m
ent
of
a
m
ass
of
r
ock,
de
br
is
or
eart
h
dow
n
slo
pe
i
nf
l
uen
ce
by
gr
a
vity
.
The
oft
en
tri
gg
e
rin
g
facto
rs
of
la
ndsli
de
i
nclu
ding
inte
ns
e
or
pr
olong
e
d
rain
fa
ll
,
earth
qu
a
kes
,
ra
pid
sn
ow
m
el
ti
ng
,
volc
anic
act
i
vity
and
var
ie
ty
of
ant
hropo
gen
ic
a
ct
ivit
ie
s.
La
ndsli
de
di
saste
r
m
ay
aff
ect
on
hu
m
an
li
fe,
pr
op
e
rty
an
d
the
env
i
ronm
ent
[1
]
-
[
3]
.
I
n
Ma
la
ysi
a,
m
os
t
of
la
nd
sli
des
or
slop
e
fail
ur
e
i
ncide
nt
occurre
d
due
t
o
inte
ns
e
rainfal
l
and
urba
n
dev
el
op
m
ent
over
hill
side
ar
eas
[
4
]
.
O
n
6t
h
De
cem
ber
2008,
a
la
nd
sli
de
was
occurre
d
at
Ta
m
an
Buk
it
n
M
ewah,
Bu
kit
A
ntara
bangsa,
U
lu
Kelan
g,
Sel
ango
r.
T
his
traged
y
cl
aim
ed
the
li
ve
s
of
fou
r
people
wit
h
15
ot
her
s
in
jure
d.
It
was
est
im
a
te
d
that
101,5
00
cub
ic
m
et
ers
of
eart
h
had
tra
ns
la
te
d,
rep
re
sents
one
of
the
m
ajo
r
la
nd
sli
de
phe
no
m
ena
that
oc
curred
i
n
Ulu
Kelang
Sela
ngor.
An
early
war
ni
ng
syst
e
m
is
ca
pab
le
to
re
duc
e
or
m
ini
m
iz
e
the
i
m
pact
of
la
nd
sli
de
oc
cur
e
nces
on
hum
an,
pro
per
ty
d
am
age a
nd loss
of l
ive [5]
,
[6
]
.
Ra
infall
thr
es
holds are
w
idely
u
sed
in
t
he
de
velo
pm
ent o
f
la
ndsli
de
Early
W
ar
ning Syste
m
(
E
W
S)
at
reg
i
on
al
scal
e.
This
m
et
ho
d
is
able
to
pr
ov
i
de
bette
r
r
esults
tha
n
t
he
physi
cal
ly
ba
sed
m
et
ho
d
[
7
]
-
[
9].
Em
pirical
thresh
ol
ds
for
crit
ic
al
rainf
al
l,
ei
ther
daily
or
ho
ur
ly
an
d
a
ntecedent
rain
fall
,
w
hich
is
tri
gg
e
ring
t
he
la
nd
sli
de
is
de
velo
ped
f
ro
m
acce
ssing
data
of
date,
ti
m
e
a
nd
rain
fall
dat
a
for
the
pri
or
la
nd
sli
de
e
ve
nt
s.
Th
e
te
rm
“thr
esh
old
” is def
i
ned
a
s the
m
ini
m
u
m
o
r
m
axi
m
u
m
l
evel o
f
a
nu
m
ber
o
f q
uan
ti
ti
es n
eede
d
f
or
a pro
ce
s
s
of
ta
ke
place
or
a
sta
te
of
c
hange.
F
or
rai
nf
al
l
-
in
duce
d
la
ndsli
des,
a
ra
infall
thres
ho
l
d
is
ref
e
rr
in
g
to
the
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
3
,
Ma
rc
h
201
9
:
1
2
5
9
–
1
2
6
6
1260
a
m
ou
nt
of
rain
fall
that,
wh
e
n
reache
d
or
e
xc
eeded,
is
li
ke
ly
to
trigg
er
la
nd
sli
des.
Ra
in
f
al
l
threshold
c
an
be
cl
assifi
ed
into
five
cat
eg
or
ie
s
including
inte
ns
it
y
-
durati
on
(ID
diag
ram
),
accum
ulate
d
rainf
al
l
-
du
rati
on
(R
-
D
diag
ram
),
Acc
um
ulate
d
rainfal
l,
intensit
y
-
accum
ulate
d
rainf
al
l
(
I
-
R
diagr
am
)
an
d
acc
um
ulate
d
rainfal
l
–
accum
ulate
d
ra
infall
(
R
-
R
d
ia
gr
am
)
[
10
]
-
[
12
]
.
Ther
e
are
m
a
ny
ap
proac
hes
f
or
rain
fall
thres
h
old
a
naly
sis,
it
s
inclu
di
ng
the
us
e
of
ga
ug
e
-
base
d
rainf
al
l
data
to
exam
ine
the
relat
ion
sh
i
p
bet
ween
rain
fall
and
la
nd
sli
de
occ
urre
nce.
H
ow
e
ve
r,
this
m
et
ho
d
pro
vid
es
a
li
m
i
te
d
rai
nf
al
l
data
beca
us
e
of
th
e
lim
it
ed
nu
m
ber
an
d
acce
ss
ibil
it
y
of
the
ra
in
ga
uge
sta
ti
on
a
nd
so
m
e
rainf
al
l
data
is
based
on
m
on
thly
accum
ula
ti
on
read
i
ng
[
13
]
-
[
16]
.
Ther
e
f
or
e,
the
appr
oach
es
us
i
ng
th
e
rem
ote
sensing
i
m
agine
is
m
or
e
sig
nifica
nt
to
asses
s
the
ch
aracte
risti
cs
of
rain
fall
prece
di
ng
la
ndsli
de
e
ven
ts
recorde
d
i
n
histor
ic
al
in
ven
t
or
y.
Acc
ordin
g
to
[
17
]
t
he
est
ablishe
d
s
yst
e
m
can
us
e
both
real
ti
m
e
an
d
forecast
in
g
rainf
al
l
data
a
nd
can
i
den
ti
fy
t
he
m
os
t
hazar
dous
r
ai
nf
al
l
of
each
rai
n
e
ve
nt.
[18]
def
i
ne
d
t
hat
rem
ote
sensing
sat
el
li
te
-
base
d
obse
rvat
ion
data
can
pro
vi
de
a
s
olu
ti
on
for
s
patia
l
sam
pl
ing
li
m
it
ation
s
of
gauge
-
base
d
ga
ug
e
net
w
ork.
[
19
]
s
uggeste
d
that
sat
el
li
te
data
ca
n
be
use
d
for
foreca
sti
ng
la
nd
sli
de
s,
on
ly
perform
ing
a
local
scal
ing
be
tween
sat
el
li
te
and
groun
d
data.
The
i
ncrea
sed
avail
abil
it
y,
con
sist
enc
y
and
high
-
s
patiote
m
poral
-
reso
l
utio
n
of
data
is
us
ef
ul
in
la
ndsli
de
haza
rd
a
ssessm
ent
fo
r
broa
d
-
scal
e,
global
,
reg
i
on
al
,
c
ount
ry
and
cat
ch
m
ent
-
le
vel
an
d
are
no
t
li
m
i
te
d
by
inacce
s
sibil
it
y
du
e
to
te
rr
ai
n,
c
ultu
re
a
nd
po
li
ti
cs.
In
this
st
ud
y,
Ulu
Kelan
g,
S
el
angor
was
us
ed
as
a
case
stud
y
t
o
dev
el
op
a
RLE
WS
base
d
on
the
relat
ion
s
hip
be
tween
rai
nf
al
l
and
la
ndsli
de
occure
nces.
T
he
dev
el
ope
d
m
et
hod
is
based
on
r
eal
tim
e
T
RM
M
sat
el
li
te
data
and
rainf
al
l
thr
e
sh
ol
d
analy
sis,
wh
ic
h
is
able
to
retrieve
a
nd
store
real
tim
e
rainf
al
l
data
f
or
the
la
nd
sli
de
ea
rly
warnin
g
syst
em
.
The
obj
ect
ive
of
this
rese
arch
was
de
velop
i
ng
the
lo
w
-
cost
an
d
su
sta
i
nab
le
early
warnin
g
syst
e
m
;
and
web
base
d
app
li
cat
io
n
to
prov
i
de
noti
ficat
ion
,
a
wareness
a
nd
la
nd
sli
de
inf
or
m
at
ion
for reside
ntial
are
as in
Ulu Kela
ng, S
el
a
ngor.
2.
METHO
D
OL
OGY
2.1.
S
tu
d
y Are
a a
nd Da
ta
RLE
W
S
is
de
velo
ped
f
or
Ulu
Kelan
g,
Sela
ngor
with
co
ve
r
an
area
of
600
hectares.
U
lu
Kelan
g
is
locat
ed
at
the
la
ti
tud
e
of
3°
12’
30’
’N
a
nd
101°
45’
28’’E
with
a
ppr
ox
i
m
at
e
distance
of
5
km
fr
om
Ku
al
a
Lum
p
ur
ci
ty
centre
as
s
hows
in
F
igure
1.
Ulu
K
el
an
g
is
known
as
a
n
urba
n
area
w
hi
ch
ha
ving
ve
r
y
hig
h
dem
and
for
it
s
la
nd
,
pro
per
t
y
and
ho
us
in
g
dev
el
op
m
ent
in
the
hill
side
area.
T
he
cl
im
at
e
of
this
area
is
com
m
on
ly
h
ot
and hum
id as it i
s locate
d i
n
t
he
tropica
l m
on
so
on
re
gion.
The
rain
fall
da
ta
wh
ic
h
are
use
d
in
the
rain
fall
threshold
a
naly
sis
is
ob
ta
ined
f
r
om
TR
MM
sat
el
l
ite
pr
eci
pitat
ion
in
gr
i
dded
point
loca
te
d
at
the
la
ti
tud
e
of
3.1
25°
N
a
nd
101.875°
E.
I
n
this
stud
y,
TRM
M
Dat
a
pro
du
ct
Dail
y
Ra
infall
(TR
MM
_3
B4
2_Da
il
y
v7
)
is
a
ppli
ed
in
t
he
rainf
al
l
a
naly
sis.
Fou
rteen
hist
or
ic
a
l
la
nd
sli
de
locat
i
on
is
sel
ect
ed
to
const
ru
ct
a
r
ai
nf
al
l
thres
ho
l
d
analy
sis
in
U
lu
Kelan
g
Sela
ngor.
The
histo
ri
c
al
la
nd
sli
de
ev
e
nt
is o
cc
urr
ing
be
tween yea
rs 1
999
t
o 201
2.
Figure
1
.
L
ocat
ion
of
Ulu Kel
ang, Sela
ngor
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Rain
f
all
–
la
nds
li
de
early w
arn
ing
syste
m
(
RL
EWS)
u
sin
g
T
R
MM p
r
eci
pitati
on esti
m
ates (N
or
suzi
la Y
a’a
cob
)
1261
2.2.
Ra
in
fa
ll
th
res
ho
ld
analysis
The
rai
nf
al
l
th
reshold
was
de
velo
ped
t
o
de
te
rm
ine
the
am
ou
nt
of
rain
f
al
l
that,
wh
e
n
reache
d
or
exceede
d,
is
li
kely
to
trig
ge
r
la
ndsli
des.
The
th
res
ho
l
d
al
so
inter
pret
ed
as
a
n
a
ppr
ox
im
at
e
lower
-
bo
und
thres
ho
l
d.
Where
is
w
hen
be
low
the
sp
eci
fi
ed
le
vel,
t
he
r
a
infall
induce
d
la
nd
sli
de
act
ivi
ty
do
es
no
t
oc
cur,
or
rar
el
y
occ
urs,
and
ab
ove
w
hi
ch
it
m
a
y
occur
un
der
ce
rtai
n
conditi
on.
In
this
stu
dy,
the
E3
-
E3
0
diag
r
a
m
is
plo
tt
ed
to
dete
rm
ine
the
rain
fall
thres
hold
for
Ulu
Kela
ng,
Sela
ng
or
.
T
he
rain
fall
thr
esh
old
wa
s
an
al
yz
ed
us
in
g
TRM
M
Sate
ll
it
e d
at
a ba
sed o
n
the
h
is
torical
lan
ds
li
de
o
cc
urren
c
e in
form
ation
si
nc
e ye
ar
1999
–
2012.
The
pro
po
se
d
rainf
al
l
th
res
hold
f
or
la
ndsli
de
occ
ur
e
nces
i
n
Ul
u
Kelan
g
i
s
co
ns
tr
ucted
by
plo
tt
ed
3
-
day
an
d
30
-
da
y
cum
ulati
ve
rain
fall
for
f
ourteen
sel
ect
ed
la
ndsli
de
e
ven
ts
.
T
her
e
a
re
tw
o
cat
eg
ori
es
of
la
nd
sli
de
eve
nt
s
;
m
ajo
r
la
nds
li
de
an
d
m
ino
r
la
ndsli
de
in
vo
lve
in
t
he
a
nal
ysi
s
to
dete
rm
i
ne
the
wa
r
ning
le
vel
of
po
te
ntial
lan
ds
li
de
occ
ur
e
nces.
T
he
plo
t
te
d
gr
a
ph
was
fo
rm
ed
of
tw
o
lim
i
ta
ti
on
threshold
li
ne
w
hich
i
s
div
ide
d
the
wa
rn
i
ng
le
vel
i
nto
t
hr
ee
sta
te
s,
nam
ely
as
lo
w
,
m
od
erate
a
nd
hi
gh
le
vel
of
pontentia
l
la
nd
sli
de
occura
nces.
Th
e
rain
fall
th
res
ho
l
d
li
ne
in
F
i
gure
2
is
s
how
s
the
f
or
m
ed
of
propose
d
rain
fall
thres
hold
f
or
Ul
u
Kelan
g,
Sela
n
gor
. T
he rai
nf
al
l
thr
es
hold
li
m
itati
on
li
ne
as
fol
lowing:
Fo
r
m
ajo
r
la
ndsli
de:
3
=
161
.
71
−
0
.
607
30
(1)
Fo
r
m
ino
r
la
nd
sli
de:
3
=
110
.
02
−
0
.
607
30
(2)
Figure
2
.
The
Pr
op
os
e
d
Ra
in
fall
(
E
3
-
E
30) Thre
shold
Cha
rt for
U
l
u Kela
ng, S
el
a
ngor
2.3.
RLEWS
De
ve
lopmen
t
RLE
W
S
is
a
syst
e
m
wh
ic
h
involves
t
he
a
naly
sis
of
rain
fall
data
obta
ined
from
TRM
M
sat
el
lite
pr
eci
pitat
ion
e
stim
at
es.
On
e
gr
i
dd
e
d
po
i
nt
of
TRM
M
data
is
require
d
t
o
re
pr
e
sent
th
e
rain
fall
a
m
ou
nt
of
la
nd
sli
de
pro
ne
area
in
Ul
u
Kelan
g,
Sela
ngor.
T
he
rai
nf
al
l
data
is
a
vaila
ble
on
th
e
NASA
Ea
rth
Data
Netw
ork.
T
he
real
tim
e
daily
rainf
al
l
data
is
autom
at
ic
ally
r
et
rieved
from
t
he
N
ASA
data
base
to
the
RL
E
WS
Web
Se
rv
e
r.
T
hen
the
TRM
M
rain
fall
data
are
a
naly
zed
into
t
hr
ee
pa
ra
m
et
ers;
these
a
re
daily
rainf
al
l,
3
day
accum
ulate
d
rainf
al
l
and
30
da
y
accu
m
ulate
d
rain
fall
.
All
these
pa
ram
et
er
s
are
plo
tt
ed
on
a
gr
a
ph
known
as
rainf
al
l
patte
rn
analy
sis.
T
he
3
day
acc
umulat
ed
rain
fall
an
d
30
day
a
ccum
ulate
d
rainf
al
l
par
am
et
e
rs
a
r
e
require
d
to
i
de
ntify
the
warnin
g
le
vel
of
l
andsl
ide
occ
ur
ren
ces
by
us
in
g
rai
nf
al
l
th
re
sh
ol
d
a
naly
sis.
This
analy
sis
is
a
con
ti
nuin
g
pro
cess
is
wh
e
re
the
syst
em
is
updati
ng
the
daily
rain
fall
data
e
ver
y
da
y.
Th
e
flo
wch
a
rt
dia
gram
fo
r
RL
E
WS
is
s
how
n
in
F
igure
3.
The
r
e
are
var
i
ous
i
nfor
m
at
ion
ha
ve
is
a
vaila
ble
to
vie
w
for
purpose
of
early
warni
ng
sign
al
for
la
ndsli
de
occurr
e
nces
a
nd
as
w
el
l
as
the
la
nd
sli
de
m
on
it
or
ing
to
ol.
The We
b
-
base
d
inte
rf
ace
can
be display
ed
in
d
es
ktop a
nd m
ob
il
e
dev
ic
es
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
3
,
Ma
rc
h
201
9
:
1
2
5
9
–
1
2
6
6
1262
Figure
3
.
Fl
owcha
rt fo
r
RL
E
W
S
2.4.
R
LE
WS
Web
-
based I
nt
er
face
RLE
W
S
is
a
w
eb
-
base
d
inter
f
ace
of
so
m
e
relat
ed
com
po
ne
nts.
The
RLE
WS
W
e
b
-
base
d
I
nterf
ace
is
com
bin
at
ion
of
rainf
al
l
analy
s
is,
com
pu
te
r
program
m
ing
an
d
interc
onnect
ed
data
base.
T
he
PHP
Java
s
cript
is
us
e
d
to
retriev
e
and
extr
act
da
ta
from
the
TRM
M
sat
el
lite
database
,
ide
nt
ify
the
warnin
g
le
vel
by
anal
yz
ing
rainf
al
l
th
res
hold
a
nd
to
prov
i
de
ot
hers
a
vaila
ble
real
t
i
m
e
info
rm
at
i
on.
T
he
w
hole
syst
e
m
is
s
how
n
in
F
igure
4.
Figure
4.
The
RLE
W
S
b
l
ock
diag
ram
3.
RESU
LT
S
A
ND AN
ALYSIS
The
Web
-
ba
se
d
I
nter
face
of
RLE
W
S
is
div
ide
d
into
f
our
inf
orm
ative
sect
ion
s,
know
n
as
hom
e,
rainf
al
l
patte
rn,
data
m
anag
e
m
ent
and
a
bout
us
.
Eac
h
sect
ion
pro
vid
es
t
he
in
form
at
ion
and
a
naly
sis,
wh
ic
h
includi
ng
warn
ing
le
vel,
rai
nfal
l
data,
gr
ap
h,
ta
ble
and
ex
planati
on.
All
the
inform
ation
pro
vid
e
d
by
RL
E
WS
is use
d
as
early
w
a
rn
i
ng sig
na
l and m
on
it
or
i
ng to
ol for p
ot
entia
l l
and
sli
de
.
Ho
m
e
is
a
fr
ont
pag
e
w
hich
i
s
viewe
d
the
w
arn
i
ng
si
gn
al
le
vel
of
po
te
ntial
la
nd
sli
de
oc
currence
s
i
n
Ulu
Kelan
g,
S
el
angor.
I
n
thi
s
sect
ion
,
t
he
r
ai
nf
al
l
am
ou
nt
of
daily
rainfal
l,
accum
ulated
3
-
day
rain
f
al
l
an
d
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Rain
f
all
–
la
nds
li
de
early w
arn
ing
syste
m
(
RL
EWS)
u
sin
g
T
R
MM p
r
eci
pitati
on esti
m
ates (N
or
suzi
la Y
a’a
cob
)
1263
accum
ulate
d
30
-
days
rain
fall
are
real
tim
e
rainf
al
l
inf
orm
at
ion
to
iden
ti
fy
the
war
ni
ng
si
gn
al
le
ve
l.
Thi
s
inf
or
m
at
ion
can
be
vie
wed
a
s
show
n
i
n
F
i
gure
5
.
Fig
ur
e
5
s
hows
t
he
r
a
infall
inf
orm
ation
on
4
th
Sept
e
m
ber
2018.
Wh
e
re
da
il
y
rainf
al
l
is
3.99
m
m
,
total
3
day
rai
nf
al
l
is
22.08
m
m
an
d
total
30
day
r
ai
nf
al
l
is
44.16
m
m
.
Wh
il
e th
e wa
r
ning level
sho
ws
a l
ow pote
nt
ia
l of
lan
ds
li
de
o
cc
ur
s
Figure
5.
Re
al
-
tim
e inf
or
m
at
i
on
:
Dai
ly
r
ai
nf
al
l, 3
-
day rai
nfal
l, 3
0
-
day rai
nfal
l an
d war
ning level
The
plo
tt
ed
gr
aph
of
rain
fall
thres
ho
l
d
a
nal
ysi
s
of
previ
ous
30
-
day
is
s
ho
wn
in
F
ig
ur
e
6
.
T
he
gr
a
ph
can
be
us
e
d
to
m
on
it
or
pot
entia
l
la
nd
sli
de
occ
urren
ce
s
base
d
on
rain
f
al
l
data.
T
he
r
ai
nf
al
l
thr
esh
ol
ds
ar
e
div
ide
d
int
o
3
areas
of
gra
ph
;
wh
ic
h
are
high,
m
od
erate
a
nd
lo
w.
T
he
bo
r
der
of
eac
h
are
a
le
vel
is
assigned
by
the
gree
n
li
ne
f
or
l
ow
-
m
od
e
rate
lim
it
and
re
d
li
ne
f
or
m
od
erate
-
hi
gh
lim
it
.
The
gr
aph
s
hows
30
points
represe
nt
as
a
30
days
rain
fa
ll
threshold
an
al
ysi
s,
wh
e
re
29
points
is
l
oc
at
ed
in
t
he
lo
wer
le
vel
area
an
d
1
po
i
nt
is
fall
into
the
m
od
erate
le
vel
of
pote
ntial
la
nd
sli
de
occurre
nces
.
The
inf
orm
ati
on
a
bout
date,
3
-
day
accum
ulate
d
rainf
al
l
and
30
-
da
y
accu
m
ulated
rain
fall
fo
r
e
ach
rain
fall
threshold
point
can
be
ap
pear
i
ng
whe
n
cur
s
or is a
ppoi
nted
t
o
the
po
i
nt.
Figure
6.
Ra
inf
al
l t
h
reshol
d
a
naly
sis view
in
te
rf
ace
The
ta
ble
of
ra
infall
analy
sis
in
pr
e
vious
30
days
is
pr
ovide
d
to
view
the
va
lue
of
pa
ram
e
te
r
su
ch
a
s
date,
daily
rainf
al
l,
3
-
day
ac
cum
ulate
d
rainf
al
l,
30
-
day
accum
ulate
d
r
ai
nf
al
l
an
d
wa
rn
i
ng
le
vel.
Fi
gure
7
sh
ows
t
he
sam
ple
7
-
days
of
r
ai
nf
al
l
data
a
na
ly
sis
wh
ere
al
l
day
are
rec
ord
ed
at
lo
w
warn
ing
le
vel
of
po
t
entia
l
la
nd
sli
de
o
cc
ur
ren
ces
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
3
,
Ma
rc
h
201
9
:
1
2
5
9
–
1
2
6
6
1264
Figure
7.
Ra
inf
al
l t
hr
esh
ol
d
a
naly
sis t
able
view
inte
rf
ace
In
rain
fall
patt
ern
sect
ion,
th
r
ee
rain
fall
patte
rn
s
w
e
re
plo
tt
ed
us
in
g
real
ti
m
e
TRM
M
satell
it
e
data
in
pr
e
vious
30
-
da
ys.
This
in
form
ation
is
us
e
fu
l
f
or
la
ndsli
de
m
on
it
or
in
g
based
on
rai
nf
al
l
patte
r
n
of
daily
rainf
al
l,
acc
umulat
ed
3
-
day
r
ai
nf
al
l
an
d
acc
um
ulate
d
30
-
da
y
rainf
al
l.
Wh
en
t
he
li
ne
gr
aph
is
inc
reas
i
ng,
the
po
te
ntial
la
nd
s
li
de
occurre
nc
es
will
be
increased
a
nd
ne
ed
highly
m
on
it
or
e
d
by
res
pons
i
ble
agency
and
reside
nt
area
.
The
real
tim
e
rain
fall
patte
rn
i
nterfac
e
f
or
30
days
in
betwee
n
25
th
A
ugus
t
2018
to
3
rd
Septem
ber
2018 is sho
wn
i
n
F
igure
8
a
nd
9
.
I
n
F
ig
ure
8
(a)
t
he
m
axi
m
u
m
d
ai
ly
r
ai
nf
al
l i
s r
eco
rd
e
d
as
14.
1mm
wh
il
e
m
ini
m
u
m
is
0
m
m
.
Th
e
gr
a
ph
s
how
s
that
21
days
without
rai
n
and
t
he
rem
ai
nin
g
days
are
rainy
day.
Accor
ding
to
[
20
]
,
w
he
n
t
he
daily
rain
fall
a
m
ou
nt
is
m
or
e
than
40m
m
,
the
pote
ntial
la
ndsl
ide
occ
urre
nces
a
re
increase
d.
In
acc
um
ulated
3
-
day
rainf
a
ll
patte
rn
sho
wn
i
n
F
ig
ur
e
8
(
b),
the
m
axim
u
m
accu
m
ulate
d
3
day
rainf
al
l
is
24
.
72m
m
.
The
m
in
i
m
u
m
value
is
0mm
and
this
value
sho
wn
t
hat
the
pr
e
viou
s
are
3
days
w
it
ho
ut
rain.
The
accu
m
ula
te
d
3
-
day
rain
fall
patte
r
n
is
a
ble
to
m
on
it
or
th
e
pot
entia
l
la
nd
sli
de
occ
ur
s
beca
us
e
of
heav
y
raini
ng
i
n
s
hort
durati
on
rai
nf
al
l
pe
rio
d.
Ba
se
d
on
t
he
stud
y,
t
he
la
nd
sli
de
occ
urr
ences
wer
e
tri
gg
e
re
d
wh
e
n
the
am
ou
nt
of
accum
ulate
d
3
-
day
rain
fall
is
reached
110mm
[2
0].
The
rai
nf
al
l
pa
tt
ern
show
n
in
F
igure
9
is
plo
tt
ed
to
m
on
it
or
the
po
te
ntial
la
nd
sli
de
occ
urred
bec
ause
of
prol
on
ged
co
ntin
uous
rain
fall
.
T
he
r
ai
nf
al
l
thres
ho
l
d
am
ou
nt
wh
ic
h
rea
ches
300mm
t
o
45
0mm
can
be
trig
ge
red
t
he
la
nd
sli
de
oc
currence
s.
178m
m
is
recorde
d
as
a
m
axi
m
u
m
a
m
o
un
t
of
accum
ulate
d
30
day
r
ai
nf
al
l,
w
hile
20
m
m
is
a
m
i
nim
u
m
a
m
ou
nt.
Tha
t
m
eans each am
ount is
recor
de
d unde
r
the
lim
it
at
ion
of rai
nfal
l t
hr
esh
ol
d of p
otentia
l l
an
dsl
ide o
cc
urre
nce
s.
(a)
(b)
Figure
8
(a) R
ai
nf
al
l Pat
te
r
n
i
nterf
ace
for Da
il
y rainfall
(b)
Ra
infall
Patt
er
n
inte
rf
ace
for
3
-
day
accum
ulati
ve
rainf
al
l
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Rain
f
all
–
la
nds
li
de
early w
arn
ing
syste
m
(
RL
EWS)
u
sin
g
T
R
MM p
r
eci
pitati
on esti
m
ates (N
or
suzi
la Y
a’a
cob
)
1265
Figure
9.
Ra
inf
al
l Pat
te
rn
inte
rf
ace
f
or
30
-
da
y ac
cum
ulati
ve
rain
fall
4.
CONCL
US
I
O
N
RLE
W
S
is
i
m
plem
ented
a
real
-
tim
e
data
t
o
obta
in
m
or
e
accurate
anal
ysi
s
relat
ed
to
rainf
al
l
and
la
nd
sli
de
occ
urences
.
T
he
m
ai
n
pur
pose
of
RLE
WS
is
pro
vid
in
g
a
n
early
warnin
g
sign
al
f
or
po
te
ntial
la
nd
sli
de
occ
urences
.
Be
side
s
that,
the
inf
or
m
at
ion
an
d
analy
s
is
in
R
LE
W
S
W
e
b
-
ba
sed
inte
rcafe
can
be
app
li
ed
as
la
ndsli
de
m
on
it
or
ing
t
oo
ls
f
or
Ulu
Kela
ng,
S
el
angor.
T
hr
ee
noti
ficat
ion
le
vel
of
wa
rn
i
ng
sig
nal
was
intr
oduce
d;
high,
m
od
a
r
at
e
and
l
ow.
T
he
ide
ntific
at
io
n
of
warnin
g
l
evel
is
base
d
on
rain
fall
thr
esh
old
ana
ly
sis
of
TRM
M
sat
ellit
e
data.
The
dev
el
op
m
ent
of
RLE
W
S
is
su
cces
sfu
l
an
d
the
an
al
ysi
s
info
rm
ation
i
s
updatin
g
i
n
24
hours
wh
ic
h
is
be
nef
it
t
o
c
omm
un
it
y
of
resi
den
ti
al
area
in
Ulu
Kela
ng,
lo
cal
auth
or
it
y
s
uch
as
MPAJ; a
nd urba
n dev
el
op
e
r
i
n pro
vid
in
g
no
ti
fi
cat
ion
, a
wareness a
nd lan
dsl
ide info
rm
ati
on.
ACKN
OWLE
DGE
MENTS
The
a
uthors
w
ou
l
d
li
ke
to
t
ha
nk
Faculty
of
Ele
ct
rical
En
gin
ee
rin
g,
Un
i
ver
sit
i
Tek
nolog
i
M
ARA
(U
iTM
)
f
or
th
ei
r
val
uab
le
s
upport.
T
his
r
esearch
is
par
t
ly
fund
e
d
by
t
he
Ma
la
ysi
an
Gove
rn
m
ent
thr
ou
gh
UiTM
un
der
600
-
RM
I/D
A
N
A5
/
3/BEST
ARI
(
122/
2018).
We
are
grat
ef
u
l
to
N
ASA
T
MPA
f
or
prov
i
ding
the
TRM
M pro
du
c
t ver
si
on 7 3B
42
(7) dat
a.
REFERE
NCE
S
[1]
K.
Zha
ng
,
e
t
al
.,
“
The
assess
me
nt
of
l
andsli
d
e
sus
ce
pti
bi
li
t
y
m
appi
ng
using
ran
dom
fore
st
a
nd
dec
ision
tree
m
et
hods i
n
th
e T
hre
e
Gorg
es
Res
erv
ior area, Chin
a,
”
En
vi
ron
E
arth S
ci.,
2017
.
[2]
K
y
ungji
n
A
.,
et
al
.
,
“
Deve
lopi
n
g
an
Acc
essible
La
ndslid
e
Sus
ce
pti
bi
li
t
y
Mod
el
Us
ing
Open
-
Source
Resour
ce
s,
”
Sustainabi
lity
,
v
ol.
10
,
pp
.
293
,
2
018
.
[3]
F
.
Ferrigno
,
e
t
al
.
,
“
GB
-
InSA
R
m
onit
oring
and
observa
ti
on
al
m
et
hod
for
la
nd
slide
emerge
n
c
y
m
ana
gement:
th
e
Montagut
o
ea
rth
flow
(AV
,
Ita
l
y
),
”
Nat
.
Haz
ards
Earth
Syst
.
S
ci
.
,
vol.
17
,
pp
.
845
–
860,
2017
.
[4]
I.
E.
Sam
y
,
e
t
a
l
.
,
“
L
andsli
de
Modell
ing
and
An
aly
s
is
using
Re
m
ote
Sensing
an
d
GIS
:
A
ca
se
stud
y
o
f
Camero
n
Highla
nd,
Mal
a
ysia
,
”
Journal
of
Geomatic
s
ISG
,
v
ol
/i
ss
ue:
8
(
2
)
,
2
014
.
[5]
Y.
Artha
and
E.
S.
Julia
n,
“
La
nd
slide
ea
r
l
y
warn
i
ng
sy
st
em prot
oty
pe
with
GIS
ana
l
y
s
is i
ndic
a
te
s b
y
soil
m
ovement
and
ra
infall,
”
IO
P
Conf
.
S
erie
s:
Earth
and
Env
ir
onmental
S
ci
en
c
e
,
pp
.
106
,
2018
.
[6]
Z.
Liao
,
e
t
al
.
,
“
Protot
y
ping
an
expe
rimental
e
arly
w
arn
ing
s
y
stem
for
rai
nfa
ll
-
induc
ed
l
andslides
in
Indone
sia
using sa
tellite
re
m
ote
sensing
an
d
geospa
t
ia
l
da
tas
et
s,
”
Lands
li
d
es
,
v
ol
/i
ss
ue:
7(3),
pp
.
317
-
324
,
20
10
.
[7]
A.
Rosi
,
et al
.
,
“
Updati
ng
EWS
rai
nfa
ll
thre
shold
s fo
r
the t
r
igge
r
i
ng
o
f
l
andsli
d
es
,
”
Nat
Haz
ard
s
,
2
015
.
[8]
A
.
Agos
ti
ni
,
et
al
.
,
“
A
new
appr
ai
sal
of
the
Ancona
la
ndslid
e
base
d
on
geot
ec
h
nic
a
l
inve
stig
a
tions
and
stabi
lit
y
m
odel
ing
,
”
Jour
nal
of
Engi
n
ee
ri
ng
Geology and H
ydroge
ology
,
v
ol.
47
,
pp
.
29
-
43
,
2013
.
[9]
B
.
B.
Mirus
,
et
al
.
,
“
Deve
lop
in
g
H
y
dro
-
Me
te
or
ologi
c
al
Thre
sh
olds
for
Shal
lo
w
La
ndslid
e
Ini
ti
ation
and
E
ar
l
y
W
arn
ing,
”
Water
,
vol.
10,
pp.
12
74,
2018
.
[10]
C.
F.
Lee
,
et
a
l
.
,
“
Com
bini
ng
rai
nfa
ll
par
amet
er
and
la
ndslid
e
sus
ce
pti
bil
i
t
y
t
o
fore
c
ast
shal
l
ow
la
ndslide
in
Ta
iwan
,
”
Geot
echnic
al
Engi
n
ee
ri
ng
Journal
o
f th
e
SE
AGS
&
AGS
SEA
,
v
ol
/
issue:
47
(
2
),
2016
.
[11]
E
.
Mons
ie
urs
,
et
al
.
,
“
A
sus
ce
pti
bil
ity
-
b
ase
d
rai
n
fal
l
thr
eshold
ap
proa
ch
for
la
nds
li
de
oc
cur
ren
ce
,
”
Nat.
Haz
ards
Earth
Syst
.
S
ci
.
Discuss
,
2018
.
[12]
T
.
Vaz
,
e
t
al
.,
“
Regi
onal
r
ai
nf
all
thre
sholds
for
la
ndslide
o
cc
urr
enc
e
using
a
cente
nar
y
d
at
ab
ase
,
”
Na
t.
Haz
ards
Earth
Syst
.
S
ci
.
,
vol.
18
,
pp
.
1037
–
1054,
2018
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
3
,
Ma
rc
h
201
9
:
1
2
5
9
–
1
2
6
6
1266
[13]
A.
Kass
im,
et
al
.
,
“
Modelling
of
sucti
on
dist
ribut
ions
in
an
unsatura
t
ed
he
t
ero
gen
eous
resi
dual
soil
slope
,
”
Engi
ne
ering
Ge
ology
,
pp.
131
-
1
32,
2012
.
[14]
D.
Kirschba
um
and
T
.
Stanle
y
,
“
Sate
llite
Based
As
sessment
of
Rai
nfa
ll
-
Tr
igge
r
e
d
La
ndslid
e
H
a
z
ard
for
Situ
ation
al
Aw
are
ness
,
”
Ea
rth's
Fut
ure
,
vol
.
6,
pp
.
505
-
523
,
2018
.
[15]
M.
T.
Brun
et
t
i
,
et
al
.,
“
How
far
are
we
from
th
e
use
of
satel
lite
r
ai
nfa
ll
produ
ct
s
in
la
ndsl
ide
for
e
ca
sting
?
”
Re
mot
e
Sensing
of
En
vi
ronm
ent
,
v
ol
.
210
,
pp
.
65
-
75
,
201
8.
[16]
B
.
G
.
Ch
ae
,
e
t
al.
,
“
La
ndslid
e
pre
diction,
m
onit
oring
and
ea
rl
y
wa
rning:
ac
o
nci
se
r
evi
ew
o
f
state
-
of
-
the
-
art,
”
Geosci
ences J
ou
rnal
,
v
ol
/i
ss
ue:
21
(
6
)
,
p
p
.
1033
−1
070,
2017
.
[17]
Rosi
A.
,
e
t
al
.,
“
Defi
nit
ion
of
a
fully
func
ti
ona
l
EWS
base
d
on
rai
nfa
ll
th
resho
lds,
the
ca
se
of
stud
y
of
Tusca
n
y
Regi
on
,”
in
Mikoš
M
.
,
et
al.
,
“
Advanc
ing
cul
t
ure
of
li
ving
with
la
ndslide
s
,
”
Springer
Inte
rna
ti
onal
Publishin
g,
Sw
it
ze
r
l
and,
vol
.
3
,
2017
.
[18]
J.
C
.
Robbins
,
“
A
proba
bil
isti
c
f
or
assess
ing
la
n
dslide
-
tri
gg
eri
ng
eve
nt
r
ai
nf
al
l
in
Papua
New
Guine,
using
TRM
M
Sate
llite Prec
ipi
t
at
ion
esti
m
ates
,”
Journal
o
f
Hydr
ology
,
vol
.
541
,
pp
.
296
-
309
,
20
16
.
[19]
M
.
Ross
i
,
et
al
.
,
“
Com
par
ison
o
f
S
at
el
l
ite
Rai
nf
al
l
Est
imate
s
an
d
Rai
n
Gauge
Mea
surem
ent
s
i
n
Ita
l
y
,
and
Im
p
act
onLa
ndslid
e
Mo
del
ing
,
”
Cli
mat
e
,
vol
.
5
,
pp
.
90
,
2
017.
[20]
N
.
Ya’a
cob
,
et
a
l
.
,
“
Rai
nfa
l
l
thr
e
sholds
for
poss
ibl
e
l
andsli
de
oc
cur
ren
c
e
in
Ulu
Kela
ng,
Sela
ngo
r,
Mal
a
y
s
ia
usin
g
TRMM sat
el
li
t
e prec
ip
it
a
ti
on
estim
at
es
,
”
IOP
Co
nf.
S
er.
: Ear
th E
nvi
ron.
S
ci.
,
pp.
169
,
2018
.
Evaluation Warning : The document was created with Spire.PDF for Python.