TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol. 12, No. 8, August 201
4, pp. 5938 ~ 5945
DOI: 10.115
9
1
/telkomni
ka.
v
12i8.626
9
5938
Re
cei
v
ed Ap
ril 15, 2014; Revi
sed Ma
y 2
5
, 2014; Acce
pted Ju
ne 10,
2014
Misidentification of Type of Lightning Flashes in
Malaysia
Puteri Nur S
uhaila Ab Rahman*
1
, Zikri Abadi Bharudin
2
, Nur Hida
y
u
Abdul
Rahim
3
F
a
cult
y
of Elec
trical Eng
i
ne
eri
ng, Univ
ersiti T
e
knik
al Mal
a
ysi
a
Melak
a
(UT
e
M),
Han
g
T
uah Ja
ya, 7610
0 Dur
i
a
n
T
unggal, Mel
a
ka, Mala
ys
ia, Ph.: +
606-55
5
226
7
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: suell
abp
90
@
g
mail.c
o
m
1
, zikri@utem.e
du.
m
y
2
, are
y
u
pea
ce@
y
a
hoo.co
m
3
A
b
st
r
a
ct
The Lightning
Detection Syst
em (LDS)
system
m
o
nitored by
Malaysian Meteorological
Servic
e
(MMS) and Te
nag
a Nasi
on
al
Berha
d
(TNB) are not rel
i
a
b
le
that m
a
y du
e
to unreso
l
ve
d site errors. From
infor
m
al
obser
vation, Ba
haru
d
in a
nd co-w
or
kers found th
e data recor
d
e
d
by T
N
B w
a
s not identic
al to the
i
r
data.
In this
p
aper, w
e
obs
e
r
ved
abo
ut L
i
g
h
tnin
g F
l
as
hes
in M
a
l
a
ysia
n
early
21
dat
a
match
i
n
g
the
ti
me
and
date
of o
u
r
data w
i
th d
a
t
a
fro
m
T
N
B R
e
ports that
the
li
ghtni
ng
fl
ash p
o
larity betw
een
both data has a
neg
ative
pol
ari
t
y, but there
a
r
e diffe
re
nces
on
mu
ltipl
i
city. Reg
a
rdi
ng
da
ta from T
N
B
R
eports, they
al
s
o
noted th
at ther
e w
a
s prese
n
ted a tota
l of 1
0
pos
itive cl
ou
d-to-gro
und fl
a
s
h data, b
u
t w
e
di
d not fin
d
any
data
p
o
sitiv
e
clou
d-to-gro
un
d
flash data. How
e
ver,
the
data w
e
h
a
ve
found
a total
of 191
data
w
e
re
prese
n
ted fro
m
cloud flas
h but
T
N
B did not re
cord any cl
ou
d flash.
Ke
y
w
ords
: lig
htnin
g
flash
e
s, pol
arity, mu
ltipl
i
city, cloud-to-
g
roun
d flash
e
s
Copy
right
©
201
4 In
stitu
t
e
o
f
Ad
van
ced
En
g
i
n
eerin
g an
d
Scien
ce. All righ
ts reser
ved
.
1. Introduc
tion
It is beli
e
ved that d
o
wn
ward n
egativ
e lightnin
g
f
l
ash
e
s a
c
co
unt for
abo
ut 90%.
Typically, the
tropi
cal
cou
n
try su
ch
as
in Mala
y
s
ia
wa
s foun
d to
have o
n
ly n
egative cl
oud
-to-
grou
nd lig
htn
i
ng flashe
s a
s
re
po
rted b
y
many in
vestigators. Mal
a
ysian M
e
terologi
cal Servi
c
e
(MMS) ha
s
been respon
sible in
colle
cting data
o
n
thunde
rsto
rm event (in
c
ludi
ng lightn
i
ng
grou
nd fla
s
h
and
clo
ud
discha
rge
)
day l
e
vel for mo
re
than
30
years. Th
e lig
htni
ng g
r
ou
nd
fla
s
h
den
sity Ng i
s
defined
a
s
the nu
mbe
r
o
f
cloud
-t
o-gro
und fla
s
he
s i
n
km
-2 yr-1,
is a
c
tually a
n
importa
nt me
teorolo
g
ical
data that is
use
d
in
cal
c
ulating the
ri
sk
of lightni
ng st
rike
s to
a
st
ru
ct
ur
e,
av
ionic
sy
st
em,
flight activities a
nd a
n
y sen
s
itive d
e
vice
s. Furth
e
rmo
r
e T
ena
ga
Na
sion
al Berhad (TNB
) al
so play
role f
o
r mo
nitorin
g
this a
c
tivity since
1995. It is believe th
at the
MMS or T
N
B
use
d
the wel
l
-kn
o
wn type
of lightni
ng m
easure
m
ent f
o
r Lig
h
tning
L
o
catin
g
sy
stem
(LLS)
su
ch a
s
mag
netic di
rectio
n findin
g
(MDF),
time of arrival (TOA) an
d interferometry. Th
e
arrang
ement can be
eith
er
individual or combi
nat
ion
of all types. I
n
200
3, Ha
rt
ono [1] repo
rt
ed
about the
availability of lightni
ng
data
re
corded
by MMS or T
N
B.
He cl
aime
d that the d
a
ta
colle
cted
eith
er by
MMS o
r
T
N
B p
r
e
s
en
ted site
e
rro
r
in their ma
ppi
ng
system. T
hey sugg
este
d
that this e
r
ro
r may d
ue to
unresolved
site e
rro
rs in
here
n
t in the
existing
Lig
h
tning
Dete
ction
System (LDS
). Mo
re
over,
Pinto and
co-worke
rs
[2] al
so
claim
ed th
e same i
s
sue
as poi
nted
b
y
Harto
no th
at the avail
abil
i
ty data of L
D
S in
Ma
lay
s
ia
are
not
reliable. In
20
09, the
wo
rk of
Baharudin
a
n
d
co
-wo
r
kers has repo
rted and
characte
ri
ze
d lightnin
g
activities ([
3-5]) in M
a
lay
s
ia
su
ccessfully. Ho
wever, fro
m
their info
rmal co
m
p
a
r
a
t
ive study betwee
n
their d
a
ta and the
data
recorded
by
TNB
we
re fo
und to
be
not
identi
c
al
to
e
a
ch
othe
r. M
o
reove
r
th
ey
found
a p
o
siti
ve
grou
nd flashe
s we
re regi
stered in T
N
B
system
wh
ile
in their mea
s
urem
ent, that type of flashes
never
appe
ared in thei
r
system. Th
e
po
sitive ground
flashe
s i
s
a
c
tually co
nsi
d
e
r
ed
as
a uni
q
ue
flashe
s that
only occu
rre
d in tempe
r
a
t
e regio
n
(lat
itude above
30 deg
re
e).
This i
s
sue re
ally
raised a big d
oubt on the reliability of LDS for MMS or TNB.
In princi
ple, “negative clo
u
d
-t
o-gro
und li
ghtning flash
”
or
“ne
gative grou
nd flash” may
con
s
i
s
t
of
a
si
ngle
st
ro
ke
o
r
sev
e
ral
st
r
o
k
e
s.
E
a
ch
st
r
o
ke i
n
v
o
lv
es
a
combi
nat
ion
of
a d
o
w
n
w
a
r
d
leade
r a
nd a
n
up
wa
rd
ret
u
rn
stro
ke,
o
r
bette
r kno
w
n a
s
th
e le
ader-return-stroke
seq
uen
ce.
Furthe
rmo
r
e, these ch
aract
e
risti
c
s
combi
ned with
the relation
ship b
e
twee
n
the season,
lo
cati
on
and sto
r
m type, are b
egi
nning to be
of intere
st
to research
ers for t
he purp
o
se of
weath
e
r
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Miside
ntificati
on of Typ
e
of Lightnin
g
Fla
s
he
s in Malaysia (Puteri Nur Suhaila Ab
Rahm
an)
5939
forecast,
clim
atology a
nd t
he d
e
si
gning
of lightni
n
g
p
r
otection
sy
stems. F
o
r exa
m
ple
Da
rveni
za
and co
-worke
r [7] showed that the failure mode
s of
surge
-
p
r
ote
c
tive device
s
de
ployed in po
wer
system
s dep
end on the n
u
m
ber of st
ro
kes pe
r flash a
nd interstro
k
e
intervals.
This stu
d
y no
tes that the numbe
r of stro
ke
s per fla
s
h
and the interstro
k
e inte
rva
l
s are
very impo
rtan
t para
m
eters
to co
nsi
der in
co
-o
rd
in
ating
the
circuit
breakers in
po
wer di
strib
u
tion
sy
st
em
s.
The
char
act
e
ri
st
ics of
the neg
ative cloud
-to
-
groun
d light
ning flash
e
s,
as rep
o
rted
by
Ra
kov and
Huffine
s [6], can b
e
co
nsid
ere
d
a
s
a reliabl
e ben
chma
rk in re
cog
n
izin
g the
“accu
r
ate
-
st
roke
-count
” studie
s
,
which is
b
a
sed
on t
w
o type
s
of
measurement
tech
niqu
es:
(1)
measurement
s ba
sed o
n
the co
rrelat
ed
electri
c
field
reco
rdi
ng on
the electri
c
fields g
ene
rat
ed
by the whole
flash
with
hi
gh level
s
of
tempor
al resolution
(pe
r
fo
rmed
by Coo
r
ay an
d Pe
re
z
(199
4) i
n
Sweden
and
Co
oray, and
Jayaratne
(1
994
)
in Sri La
nka); (2) a
co
mbi
nation of
(1)
and
high sp
eed
p
hotographi
c reco
rd
s,
a
s
utilized
by
Kita
gawa et
al. (1962
) i
n
Ne
w Mexico, Rakov
and
Uma
n
(1
990),
and
Ra
kov et al.
(19
94) i
n
Flo
r
ida
.
Prior to thi
s
study, Co
ora
y
and
Jayaratne
(199
4)
co
mp
ared
their ob
servatio
ns wi
th the dat
a
set in Swe
den
by Co
oray a
nd Perez (1
9
94)
and in
Flori
d
a
by Thottappil
lil et al. (19
9
4
)
. T
hey repo
rted a
rema
rka
b
ly good
agre
e
ment
with th
e
cha
r
a
c
teri
stics of negative
cloud
-t
o-gro
und flash, su
ch a
s
the nu
mber of st
rokes count (m
e
an,
percenta
ge
o
f
singl
e-stro
ke flashe
s),
a
m
plitude
dist
ribution
of th
e SRS, a
nd
the inte
rstroke
intervals. [6]
empha
si
zed t
hat the pe
rce
n
tage of
“a
ccurate
-
st
roke-cou
nt” st
udie
s
of
single
-
stroke
flashe
s (1
4 to 21%) in Ne
w Mexico, Fl
orida, Swede
n and Sri La
nka
were fairl
y
similar to e
a
ch
other. In addi
tion, [6] explained a
bout t
he diffe
re
nce
betwee
n
the
“accu
r
ate
-
stroke
-count
” a
n
d
the Lightni
ng
Locating Sy
stem
(LLS) studies
su
ch
as the
U.S.
Nation
al Lig
h
tning Detecti
o
n
Network (NL
D
N) for Flo
r
id
a and New M
e
xico, and A
u
stria
n
Lightn
i
ng Dete
ction
and Informat
ion
System (AL
D
IS) net
wo
rk. First, in m
any multiple
-stro
k
e
gro
u
n
d
flash
e
s, o
n
ly one
stro
ke i
s
recorded
by the lightni
ng d
e
tecti
on
networks
whi
c
h
consi
dered tha
t
many SRS f
a
iled to ex
ce
ed
the sy
stem’s
trigge
r thresh
old level i
n
NLDN
an
d AL
DIS networks. Furthe
rmo
r
e
,
the pe
rcent
age
of peak
current measure
d
by using NLDN and
AL
DIS were found to be very low. The other
con
s
id
eratio
n
is th
at the fi
rst
stro
ke
is
usu
a
lly
larger than S
R
S which
is likely
to be th
e first
recorded
stroke. Se
con
d
,
the stro
ke-grou
p
ing
alg
o
rithm of th
e lightning
d
e
tection n
e
twork
(NL
D
N) d
e
fines on
e stro
ke
from a flash
and a
ssi
gn
s it
to a separate
flash. This is
the ca
se eith
er
whe
n
two st
roke
s in a fla
s
h have long
e
r
interstro
k
e i
n
tervals
by more tha
n
50
0 ms o
r
wh
e
n
a
stro
ke termi
n
ates on g
r
ou
n
d
by more tha
n
10 km from
the first stro
ke of the flash. Third, som
e
of
the si
ngle
-
stroke
flashe
s
reporte
d
by NLDN a
nd A
L
DIS were
mi
ssed to
b
e
id
entified a
s
cl
oud
flashe
s. The
lightning de
tection net
work
system
s assume tha
t
a cloud flash is unli
k
el
y to
prod
uce mo
re than
one
p
u
lse. T
h
e
r
efo
r
e, the
pul
se
is a
c
cepted
b
y
the syste
m
s a
s
a
cl
oud
-to
-
grou
nd lightni
ng RS pul
se.
Overall, [6]
con
c
lu
ded
th
at the percen
t
age of singl
e-st
ro
ke flash
e
s
repo
rted
by the
NL
DN i
s
a
factor of 2
to
3 time
s hi
gh
er tha
n
the
“a
ccurate
-
stro
ke-count
”
studi
es
in Florid
a and
is a factor of
3 to 4 times h
i
gher in
Ne
w Mexico.
In this p
a
p
e
r
we valid
ate the d
a
ta from
TNB Re
sea
r
ch
t identifica
t
ion data
corre
s
p
ondin
g
to alm
o
st
21
sam
p
le
data
we
re
take
n at t
he
same
time bet
wee
n
our
measurement
and T
N
B Rese
arche
r
. B
a
se
d on th
e
st
atement
ab
ove reg
a
rdin
g the multipli
city
feature
s
, we
reexami
ne o
u
r existing
sample
s sh
ows that there wa
s no po
sitive lightning from
2004 to
20
07
in Penin
s
ul
ar Mala
sysia
by denie
d
the
th
e data f
r
om T
N
B Research
of Abdulla
h
N
[9]. TNB
Re
searche
r
ment
ioned
that th
ey have
dete
c
ted
th
e po
si
tive lightning in Peninsula
r
Malaysia.
2. Res
earc
h
Method
2.1. Summary
of TNB Re
searc
h
on Lightning Flas
hes de
tails w
i
thin 5km r
a
dius
Refer from li
ghtning d
a
ta
analysi
s
rep
o
rt of
TNB Rese
arch of T
able 1(a) o
r
Table 1
belo
w
of su
mmary of lig
htning fla
s
h
details
within
5km
radi
us;
wa
s the
r
e
were no
det
ail
explanation
f
o
r th
e type
of flash
e
s ei
ther
cl
o
ud fl
ash
e
s o
r
g
r
ound
flashe
s. There'
s o
n
l
y
mentione
d a
bout pola
r
ity of the flashe
s eithe
r
Po
sit
i
ve (P) or
Ne
gative (N). While from Ta
b
l
e
1(b
)
from lig
htning data
analysi
s
re
p
o
rt of TN
B
Re
sea
r
ch of
Lightning fl
ash
e
s d
e
tails; th
e
feature
s
of m
u
ltiplicity
which
al
so kn
own
as numb
e
r of
stro
ke i
n
g
r
o
u
nd fla
s
he
s
we
re m
ention
e
d
.
That mean
s all the sampl
e
that given from t
he light
ning data an
alysis of TNB
Rese
arch were
con
s
id
ere
d
a
s
gro
und fla
s
hes.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 8, August 2014: 593
8 –
5945
5940
Table 1. TNB
Research
su
mmary of ligh
t
ning flash d
e
t
ails within 5
k
m radiu
s
T
y
pe
of Flash vs Return St
roke
Flashes
(all polarity)
Flashes
(N)
Flashes (P)
Number of
flashes
94
84
10
Min first return st
roke current
(kA)
8.8245
8.8245
12.4135
Max first retu
rn s
t
roke current
(kA)
113.7195
113.7195
22.8845
Average first retu
rn stroke curre
nt
(kA)
33.6277
35.6061
17.0089
2.2. Measure
met Us
e
The m
e
a
s
urements of e
l
ectri
c
field
s
gen
erate
d
by neg
ative grou
nd fla
s
h
e
s
we
re
recorded fro
m
April to Ju
ne 200
9 du
ri
ng the south
w
e
s
t mon
s
oo
n peri
od in th
e Joh
o
r
state
at th
e
south
e
rn p
a
rt of Peninsula Malaysi
a
, in cl
ose
proximity to the equato
r
(Latitud
e:1
°
N,
Longitu
de:10
3°E). Our measurement
station
alti
tude i
s
13
2m a
bove
sea
level
a
nd
approximatel
y 30km away from Teb
r
au
Straits.
We ha
d used
three pa
rallel
plate antenn
as to se
ns
e fast ele
c
tric fie
l
d, radiation fi
eld at 3
and 30M
Hz signal
s,
while a
whip anten
na wa
s
u
s
e
d
to
se
nse
the
slo
w
el
ectri
c
field sign
al.
T
he
antenn
a syst
em for the fast and
slo
w
electri
c
fi
eld
electro
n
ic b
u
ffer circuit
s
used for b
o
th
measuri
ng si
tes are ide
n
tical to the lit
eratu
r
e
s
de
scrib
ed in
Co
oray an
d Lu
ndqui
st (1
98
2),
Galvan a
nd F
e
rna
ndo
(2
00
0) an
d Coo
r
a
y
(2003
). Th
e
whip
anten
n
a
co
nsi
s
ts
of a lower m
e
tal
lic
rod, an u
ppe
r metallic rod
(3.3m)
and a
n
insul
a
tor. T
he lower met
a
llic rod is b
u
r
ied a
bout 0.
5m
in the grou
nd
while the other en
d is ab
out 1.
5m abo
ve the groun
d level.The u
pper a
nd lo
wer
metallic rods were insulat
ed fr
om each other by usi
ng an insu
lat
o
r of 0.05m t
h
ickness whi
c
h
have the cap
a
citan
c
e
s
of
58pF. The
ph
ysical h
e
igh
t,
insul
a
tor thi
c
kness an
d dia
m
eter of pa
ra
llel
flat plate antenna is 1.5, 0.05 and
0.45
m, resp
ective
ly. All three flat plate antenna we
re pl
a
c
ed
side by
side
with a di
stan
ce of 1m fro
m
eac
h othe
r. The whi
p
a
n
tenna
wa
s
placed 3m a
w
ay
from the flat plate antenn
a system. Th
ey were
lo
ca
ted 8m from the control room wh
ere t
he
recording
sy
stem was set
up. Th
e pl
an
e of
anten
na
is
orie
nted
p
a
rallel
po
sitio
n
to
the
grou
nd,
whi
c
h to en
su
re the ho
rizon
t
al electri
c
field effect can b
e
eliminated.
A 60cm lo
ng
co
axial
cabl
e (RG5
8) wa
s u
s
e
d
to
co
nne
ct the
ant
enna
to the
electroni
c
buffer circuit for
sl
ow
an
d fast
ele
c
tri
c
fi
elds.
T
he ze
ro-to-pea
k rise
time
of
th
e output wa
s
le
ss
than 3
0
n
s
when th
e
step
input p
u
lse is appli
ed to
th
e fast
ele
c
tric fields ante
n
na
system. T
he
decay time consta
nt for the fast and slo
w
ele
c
tric
fiel
d circuit is det
ermin
ed by the RC
circuit. We
tune the
de
ca
y time co
nsta
nt for the
ce
rt
ainty fast
an
d
slo
w
el
ectri
c
field ci
rcuit to
the 15
ms
an
d
1s, re
sp
ectiv
e
ly.The de
ca
y time consta
nt for the fast
electri
c
field
wa
s found to
be suffici
ent for
faithful repro
ductio
n
of micro
-
se
con
d
scale
while th
e value for th
e slo
w
ele
c
tri
c
field wa
s lo
ng
enou
gh to all
o
w in
our
an
alysis. Th
e tu
ne ci
rcuit
at 3
M
Hz i
s
a
co
mbination
of passive ele
m
ents
whe
r
e the i
n
ducta
nce (4
7
H) i
s
conn
ected in serie
s
with the a
n
t
enna (58p
F) and 50
terminatio
n formin
g a sim
p
le RL
C ci
rcuit. The t
une circuit at 30M
Hz
wa
s co
nst
r
ucte
d by usi
n
g
an a
c
tive ba
n
dpa
ss topolo
g
y co
nsi
s
ts o
f
LMH6
559
(spe
ed
buffer) and
LM
H66
09 (volta
ge f
eed
back o
peratio
nal amplifie
r).
The b
and
wid
t
h of a
tune
ci
rcuit i
s
at 3
a
nd 30
MHz a
r
e 264
kHz
an
d
2 MHz
re
spe
c
t
i
v
e
ly
.
Signals from
all antenna
were fed by 10m l
engt
h coaxial
cabl
e
s
(RG-58
) in
to a 4-
cha
nnel
12-b
i
t digital tran
sient
recorde
r
(Yo
k
og
awa
SL1000
eq
uippe
d with
DAQ mo
dule
s
7202
10) thro
ugh p
r
o
per t
e
rmin
ation
(5
0
termin
ation
)
. The
sam
p
li
ng
rate was
s
e
t to 20 or
100MS/s
with the total length re
co
rde
d
being eith
e
r
0.25 or 1
s
. The transi
e
nt reco
rd
er
wa
s
operated eith
er in 125
or 3
00ms in
pre
-
t
r
igge
r
mod
e
. The trigg
e
r
setting of the oscillo
scope
wa
s
set su
ch that
for the sign
al
s of
both pola
r
ities could b
e
captu
r
ed.
The cl
ose di
stance
of the negativ
e g
r
ou
nd flashes
wit
h
a dista
n
ce l
e
ss than 1
6
km we
re
cal
c
ulate
d
using the thun
der rangi
ng
method. In
this meth
od, the elap
se
d time between
the
arrival
s
of the
electri
c
pul
ses an
d acou
stic sign
al
s i
s
divided by the spe
ed of th
e sou
nd to arrive
at the di
stan
ce of the
fla
s
h
.
There i
s
an
approxim
ate 1s delay of
a
po
ssi
ble erro
r when
the
first
electri
c
p
u
lse
is di
splayed
in the scop
e
.
The
trigg
e
r
level is set in a ra
nge
of 500mV-
2V, to
ensure o
n
ly clo
s
e flash
could be re
co
rded. Alt
houg
h the thunde
rs for the cl
o
s
e flashes
were
audibl
e for l
e
ss than
20
km, we
only li
mit our
sel
e
ction to le
ss than 1
6
km to
preve
n
t fro
m
an
uncertainty in
time corre
s
p
ond to an un
certainty in distance.
The characteristic
of t
he fast elect
r
ic fiel
d waveform is
crucial si
nce it has good
reliability
to indicate a
clea
r sele
ctio
n of processe
s p
r
e
c
edi
n
g
the first neg
ative return-
stroke. In a
dditi
on,
the sl
ow ele
c
tric fiel
d
wav
e
form i
s
com
pulsory to
id
entify the
sta
r
ting
po
sition
of el
ectri
c
field
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Miside
ntificati
on of Typ
e
of Lightnin
g
Fla
s
he
s in Malaysia (Puteri Nur Suhaila Ab
Rahm
an)
5941
cha
nge
s
co
rrectly th
at p
r
ecede
the
first
neg
at
iv
e retu
rn
str
o
k
e
. Furth
e
rm
o
r
e, the
u
s
e
of
narro
wba
nd system (HF
radiation
at 3
M
Hz) featu
r
e
allows
de
cr
easi
ng u
n
certainty espe
ci
ally
detect to the first prelimina
r
y breakdo
wn
pulse.
2.3. Cloud Flashes v
s
Cloud-to-gr
oun
d Flashes
The differe
nces bet
wee
n
Cloud Fla
s
h an
d Clou
d-to
-ground Fla
s
h
chara
c
te
risti
c
s.
2.3.1. Cloud
Flash
If the discha
rges hap
pen
i
n
sid
e
a th
un
der
clou
d o
r
betwe
en thu
nderclo
u
d
s
, the terms
intracl
oud
fla
s
he
s
or cl
ou
d flashe
s
(ICs) are typi
cal
l
y use
d
a
s
shown in
Fig
u
re
1. Cl
oud
dis-
cha
r
ge i
s
the
most commo
n of all types
of lightni
ng
s [10]. ICs u
s
ual
ly occurs bet
wee
n
the cen
t
er
and th
e
cent
er of
po
sitive ch
arg
e
o
n
t
he lo
we
r n
e
g
a
tive ch
arg
e
.Compl
ete di
scha
rge
s
ha
s
the
orde
r of 0.2
se
con
d
s d
u
ri
ng whi
c
h tim
e
a co
nt
inuo
us lumi
nosity
obse
r
ved in
the cloud. It is
thought that
durin
g the
s
e
perio
ds the
leade
rs
sp
read na
rrowi
n
g the gap b
e
twee
n the two
cente
r
s resp
onsi
b
le. Ove
r
lap in
co
ntin
uou
s fla
s
he
s som
e
q
u
ite
brig
ht light
pulses that ti
me
perio
d a
bout
1mse
c. It wo
uld ap
pea
r from the m
e
a
s
urem
ents
of the ele
c
tri
c
fie
l
d that pul
se
s of
light is rel
a
tiv
e
ly
wea
k
retu
rn st
ro
ke that
oc
curs when
passing
acq
uaintan
ce p
o
c
kets lea
d
e
r
is
respon
sibl
e for the pola
r
o
ppo
site of a leade
r.The a
m
ount ch
arg
ed may be re
leased intra
c
l
oud
neutrali
ze
d in
the sam
e
o
r
der of m
agnit
ude a
s
t
he
subje
c
t land i
s
transfe
rred i
n
the cl
oud t
o
be
released. IC lightning
still in the clouds and is
the most common type of discharge [11].
Figure 1. Clo
ud Flash
2.3.2. Cloud-to
-Gr
ound
Flash
Approximatel
y one third of the lightning discha
rge tra
v
els betwe
en
charge
s in the clou
d
and the
g
r
ou
nd. We will
u
s
e th
e term
cl
oud-to
-g
rou
n
d
a
s
sho
w
n i
n
Figu
re
2 through
as we
shall
see some qu
ite rare flash emission fro
m
the
groun
d
and travel up towards th
e relea
s
e cl
o
ud.
Most clou
d-t
o
-g
rou
nd (CG) with ca
rry
a
n
egativ
e
cha
r
ge
to th
e groun
d.Mo
st be
gan
wit
h
the
relea
s
e
of the
CG
clou
d, prelimina
r
y bre
a
kd
own
process. Also o
n
e
of t
he
lower positive cha
r
ges
may be
invov
ed in
the
neg
ative CG
di
scharg
e
. Some
CG
rel
e
a
s
e
wa
s hi
ghe
r in
the
clou
ds a
n
d
carrie
s a p
o
si
tive charge to the gro
und.
Positive CG
discha
rge i
s
more
comm
o
n
at the seve
re
thunde
rsto
rm
s where the v
e
rtical
wi
n
d
shear i
s
p
r
e
s
e
n
t. Usu
a
lly the ce
nter of p
o
sitive charg
e
on
the prima
r
y n
egative ch
arg
e
and di
scha
rge b
e
twe
en
the two e
w
m
a
in in the cl
o
ud. Wind
sh
ar
can
ca
use cl
oud
s to lea
n
i
ng an
d mov
e
so
me of
th
e po
sitive ch
arge
away from the
corre
c
t
posito
n
in
th
e cente
r
of
n
egative
cha
r
g
e
. Reli
ef
can
travel f
r
om
clou
d to
grou
nd. Po
sitive
CG
discha
rge
al
so mo
re
co
mm
on o
n
the
st
orm at hi
gh
l
a
lituteds where t
he
cent
ral
po
sitive cha
r
ge
i
s
clo
s
er to the grou
nd. Then
, this positive CG di
scha
rg
e also comm
on in the win
t
er storm
s
wit
h
the ce
ntral
p
o
sitive charg
e
wa
s
clo
s
e
r
to the
groun
d
.
Also, the po
sitive CG
discha
rge
co
mm
on
at the
end
of
summ
er sto
r
ms
wh
ere
the
clo
u
d
ma
y til
t
and
the
me
an n
u
mb
er of
po
sitive
cha
r
ge
can b
e
found
on the anvil clou
d and a
w
ay from
the main body of
the cloud. T
he se
que
nce
of
pro
c
e
ss th
at occur
duri
ng
the negative
clou
d-to
-g
rou
nd begi
ns
wit
h
som
e
type of early rel
e
a
s
e
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 8, August 2014: 593
8 –
5945
5942
of the clo
ud.
Then, do
wn t
he rel
e
a
s
e of
negativ
ely-ch
ared move started
m
o
ving towards grou
nd.
Neg
a
tive ch
a
r
ge i
s
carrie
d
from the
ce
n
t
er of
ne
agtive ch
arge al
o
ng the
cha
n
n
e
l and
dist
ribi
te
d
leade
r [11].
Return stroke
propa
gate
s
from g
r
oun
d to clou
d. Whe
n
the stepp
e
d
leade
r ha
s lowe
red
the negativel
y charg
ed col
u
mn of high incomi
ng to
n
ear the groun
d, the quality of the resulti
n
g
electri
c
field
at the ground
is sufficient t
o
ca
use up
ward
-moving
di
schar
ged to
b
e
laun
ch
ed from
grou
nd to
wa
rds the
lead
er tip. Whe
n
on
e of the
s
e
co
ntacts
disch
a
r
ge
s the l
ead
er i
s
conn
ect
ed
to groun
d pot
ential effectivly, while the remaind
e
r
of the leade
r is
at negative incomin
g
and it
is
negatively ch
arge
d [11]. The po
sitive electri
c
field ch
ange
s at gro
und level is d
e
fined in term
s of
the di
spla
ce
ment of
neg
ative ch
arge
s d
o
wnwa
rd and po
sitive cha
r
ge
s
bein
g
raised up
ward
simila
r to the electri
c
field chang
e of neg
ative return
strokes.
.
Figure 2. Clo
ud-to
-groun
d Flash
3. Results a
nd Analy
s
is
3.1. The Differenc
es of
Record Time, Polarit
y
and Multiplicit
y
from our Result
w
i
th TNB
Table 2. The
Differen
c
e
s
o
f
Reco
rd Tim
e
, Polarity and Multiplicity from ou
r Re
su
lt with TNB
Rec
o
rd Ti
me
(us
)
Rec
o
rd Ti
me
(TN
B
)
Pol
a
ri
t
y
(us
)
Polarit
y
(TNB)
Multiplic
ity
(us
)
Multiplic
ity
(TNB)
2009.04.11; 18:3
5
:18.406
2009.04.11; 18:3
5
:57.026
N
N
5(-v
e C
G
)
1
2009.04.13; 16:4
0
:46.459
2009.04.13; 16:4
0
:36.348
N
N
3(-v
e C
G
)
1
2009.04.14; 18:0
8
:37.698
2009.04.14; 18:0
8
:36.296
N
N
1(-v
e C
G
)
1
2009.04.14; 18:1
1
:31.682
2009.04.14; 18:1
1
:30.647
N
N
1(-v
e C
G
)
1
2009.04.14; 18:1
4
:18.445
2009.04.14; 18:1
4
:07.619
N
N
3(-v
e C
G
)
1
2009.04.14; 18:1
4
:44.896
2009.04.14;1
8
:1
4:55.546
N
N
2(-v
e C
G
)
1
2009.04.14; 18:1
6
:35.229
2009.04.14; 18:1
6
:49.602
N
N
0
1
2009.04.14; 18:1
8
:46.232
2009.04.14; 18:1
8
:36.646
N
N
2(-v
e C
G
)
1
2009.04.14; 18:1
9
:47.035
2009.04.14; 18:1
9
:53.615
N
N
3(-v
e C
G
)
1
2009.04.26; 11:3
4
:27.015
2009.04.26; 11:3
4
:18.623
N
N
Cloud Flash
1
2009.04.26; 11:3
6
:54.857
2009.04.26; 11:3
6
:46.062
N
N
Cloud Flash
3
2009.04.26; 11:3
6
:54.857
2009.04.26; 11:3
6
:58.858
N
N
Cloud Flash
2
2009.04.26; 11:4
0
:12.777
2009.04.26; 11:4
0
:26.834
N
N
Cloud Flash
1
2009.04.26; 11:4
1
:23.534
2009.04.26; 11:4
1
:37.234
N
N
2(-v
e C
G
)
2
2009.04.26; 11:4
3
:02.452
2009.04.26; 11:4
3
:16.187
N
N
4(-v
e C
G
)
3
2009.04.26; 11:4
7
:44.249
2009.04.26; 11:4
7
:47.259
N
N
Cloud Flash
2
2009.04.26; 11:4
8
:26.260
2009.04.26; 11:4
8
:22.479
N
N
3(-v
e C
G
)
1
2009.04.26; 11:4
9
:26.367
2009.04.26; 11:4
9
:05.279
N
N
Cloud Flash
2
2009.04.26; 11:5
1
:39.295
2009.04.26; 11:5
1
:55.576
N
N
1(-v
e C
G
)
1
2009.04.26; 11:5
3
:15.867
2009.04.26; 11:5
3
:04.414
N
N
2(-v
e C
G
)
2
2009.04.26; 11:5
5
:48.012
2009.04.26; 11:5
5
:45.402
N
N
2(-v
e C
G
)
2
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Miside
ntificati
on of Typ
e
of Lightnin
g
Fla
s
he
s in Malaysia (Puteri Nur Suhaila Ab
Rahm
an)
5943
From
Tabl
e
2 the
differe
n
c
e
s
of
Time
Re
cord, Pola
rity and
Multi
p
licity from
o
u
r
re
sult
with TNB hav
e sho
w
n that nearly 21 dat
a matchin
g
time and date.
This sugg
est that the polari
t
y
of lightning flash bet
wee
n
both data have the sa
m
e
polarity wh
ich are negat
ive polarity. But
there a
r
e
wid
e
ly different d
ue to the mul
t
iplicity
of both data. Multip
licity refers to
the numb
e
r
of
returns
stroke
due to lig
htni
ng groun
d fla
s
h. From
o
u
r
data, there n
o
t only have
clou
d-to
-grou
n
d
flashe
s o
n
ly, but it sh
ows t
hat dat
a we h
a
ve
dete
c
ted the
present
of
clou
d fla
s
h in
the same
da
y.
Our
data
ca
n
be
re
corded
that there
we
re 6
of
21
dat
a cl
oud fla
s
h
and 1
5
of
21
data a
r
e
clo
u
d
-
to-groun
d ligh
t
ning flash
e
s
while recordi
ng TNB data i
s
only from cl
oud-to
-g
rou
n
d
lightning. O
n
ly
5 data that ha
ve same mult
iplicity betwe
en our d
a
ta a
nd TNB.
On 11
th
of April 2009, o
u
t of 136 of our
data only 1 d
a
ta has b
een
matche
d re
cord time
with T
N
B; TNB only
re
co
rded
3 d
a
ta. TNB a
nd d
a
ta we
have
re
corded
th
e sa
me fla
s
h
and
polarity whi
c
h of cloud
-to
-
groun
d flash
with negat
iv
e polarity bu
t with different numbe
rs
of
multiplicity. TNB are only
having a n
u
m
ber of
re
turns stro
ke b
u
t o
u
r data
havin
g five numbe
r of
return stroke
due to gro
u
n
d
flash.
On 13
th
of Ap
ril 200
9, of the 151 d
a
ta we re
corded, o
n
ly 1 data ha
s matched th
e data
recorded by TNB; TNB on
ly reco
rded 4
data.
TNB and our d
a
ta had sh
own the
same flash a
n
d
polarity
of clo
ud-to
-groun
d flash with ne
gative
pol
arit
y but differe
n
t
numbe
r
of
multiplicity. T
h
e
data we h
a
ve re
corded h
a
ve three nu
mber of re
turn stro
ke d
u
e
to lightning grou
nd flash, but
there is o
n
ly one num
be
r the TNB multi
p
licity.
On 14
th
of April 2009, fro
m
59 to our d
a
t
a, only 7 data we
re matched record ti
me with
TNB; TNB on
ly reco
rde
d
18 data. From
7 data ma
tch
ed, only 2 data from TNB a
nd our d
a
ta was
recorded t
he
same
lightni
n
g
flashe
s, pol
arity and
several va
riation
s
of
clou
d-to
-grou
nd li
ghtni
ng
of negative polarity with a
numbe
r of return st
ro
ke
due to gro
u
n
d
flash as a
result. Another 5
data fro
m
th
ese
7 d
a
ta reco
rde
d
mat
h
wa
s
re
co
rded the
sam
e
clo
u
d
-
to-g
round fla
s
h
with
negative p
o
la
rity but different on n
u
mb
er of m
u
ltiplic
ity. From 5 d
a
ta re
co
rde
d
, it indicate
s t
hat
the data
we
have of 3 typ
e
s of
seve
ral
variation
s
p
r
ese
n
t either
zero with
no
numbe
r of
ret
u
rn
stroke
whi
c
h
is initial
return stroke ruins or on
p
r
elimi
nary b
r
ea
kd
o
w
n
state, a n
u
mbe
r
of retu
rn
stro
ke
with 2
data and two numbe
r of return stro
ke with 2 data p
r
esent but th
e data that TNB
poste
d the
r
e
only one
ki
nd
of multipli
city that is o
ne
of seve
ral
ret
u
rn
stroke a
s
a result of th
is
grou
nd lightni
ng.
On 26
th
of Ap
ril 200
9, 159
of our data, o
n
ly
12 have
been m
a
tch
e
d
data re
co
rd
ed with
TNB; TNB o
n
ly record
ed
20 data. Of these 12
mat
c
he
d data, o
n
ly 4 have the sam
e
pola
r
ity
data, the types of lightni
n
g
flash a
nd n
u
mbe
r
of
mul
t
iplicity which
are n
egative
polarity of cl
oud-
to-groun
d flash an
d either the number
of multiplic
ity is one o
r
two.
Out of this 12 data match
ed,
only 6 h
a
ve t
he
same
data
on
negative
polarity b
u
t di
ffere in th
e type of fla
s
h. F
r
om th
e dat
a
we
recorded 6 i
s
found that clo
ud flash i
s
prese
n
t but
TNB reco
rde
d
d
a
ta from clo
u
d
-to-gro
und fl
ash
with differe
nt numbe
r of
multiplicity ei
ther on
e,
two or three of
lightning
ret
u
rn
stro
ke from
grou
nd. Th
e rest of the
dat
a from
26
th
of
April 20
09,
which i
s
2
data
have the
re
corde
d
the
sa
me
types of pola
r
ity but different on num
b
e
r of multip
licity. 1 data out of th
is 2 data re
corded th
at
TNB h
a
s fou
nd th
ree
num
ber of return
stro
ke
but
we
found
the
fo
ur
numb
e
r of
return
stro
ke
due
to grou
nd flash. Another 1
data re
co
rde
d
, TNB
ha
s b
een foun
d th
at there is
onl
y one numb
e
r
of
returns d
ue to lightning g
r
ound fla
s
h b
u
t the data we have fou
nd of three
numbe
r of re
turn
stro
ke du
e to grou
nd flash.
3.2. The Di
fferenc
es be
tw
e
e
n
o
u
r Data
w
i
th
T
N
B Based
on
Total Flas
h
es, Cloud Fl
ash,
Nega
tiv
e
an
d Positiv
e
Cloud-to-g
rou
nd Flashes
Table 3. The
Differen
c
e
s
b
e
twee
n our
Data with TNB
Based o
n
Tot
a
l Flash
e
s, Cl
oud Fla
s
h,
Neg
a
tive and
Positive Clou
d-to-gro
und F
l
ash
e
s
Differences Date
Flashes
(
a
ll
polarit
y
)
Cloud-to-
g
roun
d
Flashes
(negative)
Cloud-to-
g
roun
d Flashes
(positive)
Cloud
Flash
TNB
Report
1
s
t
- 26
th
Ap
ril 2009 (
except 20
th
- 25
th
)
94 84
10
0
Us 11
th
- 2
6
h
Ap
ril 2009
1130
939
0
191
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 8, August 2014: 593
8 –
5945
5944
Based
on
Ta
ble 3
it sho
w
s the
differen
c
e
between
our data
with
TNB
ba
sed
on the
numbe
r of tot
a
l Fla
s
he
s,
Cl
oud Fl
ash, ne
gative and
po
sitive Clo
u
d
-
to-g
rou
nd Fl
a
s
he
s. Th
e T
N
B
Rep
o
rt were recorded
th
e data
from
1
st
of April until
26
th
of April
2
009 ex
cept 2
0
th
of April unti
l
25
th
of April 2009 a
s
sche
d
u
led central p
r
ocesso
r up
grading
work.
Our
data
we
re re
co
rde
d
from 11
th
of A
p
ril until 26
th
of
April 2
009. T
N
B ha
s a
naly
z
ed th
e
total numb
e
r
of flashe
s th
a
t
they re
cord
ed only
from one
type of
flash whi
c
h
is clou
d-to
-grou
nd
flash
with b
o
th pola
r
ity of
positive a
nd
negative
with
a total of 9
4
data fla
s
h. A
l
though th
e d
a
ta
we h
a
ve an
al
yzed that the
flashe
s that
we
re
co
rded
come
from t
w
o types of fla
s
he
s
whi
c
h a
r
e
clou
d-to
-grou
nd flash
e
s a
n
d
clou
d flash
with
total 113
0 data flash
e
s
. TNB data
poste
d there
are
only 84
dat
a ba
se
d o
n
neg
ative cl
oud-to
-g
rou
n
d
flashe
s, b
u
t data
we
re
co
rde
d
9
39
data.Re
gardi
ng to
data
from T
N
B
Rep
o
rts, th
ey al
so note
d
th
at there
wa
s p
r
e
s
ente
d
a
tota
l of
10 po
sitive cl
oud-to
-g
rou
n
d
flash data,
but we
did
not find any positive cl
ou
d-to-gro
und fl
ash
data. Howeve
r, the
data
we
have fo
und
there
a
r
e
a tot
a
l of 1
91
data
we
re
present
ed fro
m
cloud
flash but T
N
B did not record any clou
d flash.
3.3. A Qualitativ
e Comparison bet
w
e
e
n
other
s
Re
s
earch
er on Presen
t of Th
unders
torm,
Nega
tiv
e
Gr
ound Flash,
Positiv
e
ground Flash a
nd Cloud Fla
s
h
From
Tabl
e 4
ba
sed
on
the
a
qualitative
comp
ari
s
o
n
with the
certa
i
n du
ration
of
time o
n
the thun
de
rst
o
rm
exist
an
d types of fla
s
he
s
betwe
e
n
Bah
a
ru
din,
Z. A. 201
2
(a
) [3], Bah
a
ru
din,
Z. A. 2012 (b
) [4], N. Azlin
da 20
09 (a) [
5
], N. Azlinda
2009
(b
) [8], Abidin, H. Z.
[1], Abdullah, N.
[9], TNB Research Repo
rt and Abdull
ah,
N [12].
Table 4. A Qualitative Co
mpari
s
o
n
bet
wee
n
Re
se
arche
r
.
Stud
y Duration
Ti
me
Thunde
rstorm
Negative
Grou
nd
Flash
Positive Ground
Flash
Cloud
Flash
Baharudin, Z.A.
2012
(a)[3]
April - June
2009
21 1299
0
533
Baharudin, Z.A.
2012 (b
) [4]
April - June
2009
- 24
0
-
N.Azlinda, 2009 (
a
) [5]
April - June
2009
- -
0
182
N.Azlinda, 2009 (
b
) [8]
April - June
2009
- 900
0
140
Abidin, H.Z. [1]
1993-200
2
3779
-
0
-
TNB Research
R
eport
April - Ma
y
20
09
-
84
10
0
Abdullah, N. [9]
2004-200
7
-
7 015 030
1 712 368
-
Abdullah, N [12]
2008-201
1
-
3 053 795
611 491
0
From Ba
harudin, Z. A. 2012
(a
) [3], it
shows t
hat the the
r
e have exi
s
ting 21
thunde
rsto
rm
s data
with p
r
esent of 12
99 of neg
ative gro
und fla
s
h and 5
33 of
clou
d flash
of
duratio
n time
April to
Jun
e
of 20
09. T
hen, dat
a
Ba
haru
d
in, Z. A
.
2012
(b
) [4] of duration ti
me
April to June
of 2009, it only mentioned
existing of
24
data of nega
tive ground fl
ash. From Ap
ril
to Jun
e
of 20
09 of data
N.Azlinda
200
9
(a) [5]
p
r
e
s
e
n
ted 182
of cloud flash wh
ile from data
N.
Azlinda
200
9
(b) [8] menti
oned th
e p
r
e
s
ente
d
of 9
0
0
of ne
gative gro
und fla
s
h
and
140
clo
ud
flash.
Data o
f
Abidin, H.
Z. [1] on d
u
ration of time
from yea
r
o
f
1993
until
2002
there o
n
ly
mentione
d of
existing
37
79 thu
nde
rst
o
rm
witho
u
t
notice
d
a
n
y
type flash
e
s present. T
N
B
Re
sea
r
ch Re
port fro
m
April to May 200
9 have
me
ntioned th
at the
r
e have
84 d
a
ta of neg
ative
clou
d flash a
nd 10 data of positive clou
d flash.
From
year 2004 to
2007 of Abdullah, N. [9] data
there
present
ed 7
015
03
0
of ne
gative
clou
d fla
s
h
a
nd 1
712
36
8
of po
sitive cl
oud fla
s
h. F
r
om
year
2008
u
n
t
il 2011
Abd
u
llah, N. [1
2] reco
rde
d
d
a
ta
with
existing
611
49
1
of
positive
gro
u
n
d
flash a
nd 3
053 7
95 of
negative g
r
o
und fla
s
h.A
ll the data
were m
e
a
s
ure
d
in Mal
a
ysi
a
. A
qualitative co
mpari
s
o
n
of our results
with t
hose o
b
tained from Baharudin, Z.
A. 2012 (a
) [4],
Baharudin, Z.
A. 2012 (b
) [3], N. Azlinda
2009 (a) [8], N. Azlind
a
20
09 (b
) [5], an
d Abidin, H.
Z.
[1] suppo
rts t
he hypothe
si
s that t
he positive lightning doe
s not exist in Malaysia. While d
a
ta from
Abdullah,
N. [9], TNB Re
se
arch Repo
rt a
nd Abdul
l
ah,
N [12] repo
rted t
here exist
positive
clou
d-
to-groun
d flash in Malaysia.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Miside
ntificati
on of Typ
e
of Lightnin
g
Fla
s
he
s in Malaysia (Puteri Nur Suhaila Ab
Rahm
an)
5945
4. Conclu
sion
We have a
n
a
l
yzed that the
r
e are no po
sitive li
ghtnin
g
flashe
s in
Malaysia b
u
t our the
data have
fo
und the
r
e
are
existing
clou
d flash
and
n
egative lightn
i
ng falshe
s o
n
ly as
su
ppo
rted
by Baharu
d
in
, Z. A. 2012
(a) [4], Bah
a
rudi
n, Z. A. 2012 (b) [3]
,
N. Azlinda
2009
(a) [8],
N
.
Azlinda 2
009
(b) [5], and A
b
idin, H. Z. [1].
Ackn
o
w
l
e
dg
ements
We
wo
uld li
ke to exp
r
e
s
s
our
mo
st g
r
at
itude a
nd
and
ackn
owl
edge
mnt to the
su
pervisor
and
re
sea
r
ch
ers from
Fa
culty of Electri
c
al En
ginee
ri
ng, UT
eM for their g
u
ida
n
ce and
co
ntinu
ous
effort a
s
well
as motivatio
n
throug
hout
the
p
r
oje
c
t.Futherm
o
re,
we would
like
ackno
w
le
dgm
ent
to Universiti Tekni
k
al Mal
a
ysia Mela
ka
(UTeM) for thei
r financial
sup
port.
Referen
ces
[1]
Abidin HZ, Ibrahim R.
T
hun
d
e
rstorm d
a
y a
nd gro
und flas
he de
nsity in
Malaysi
a
. T
he
confere
n
ce of
Natio
nal Po
w
e
r
and Ener
g
y
. 2
003.
[2]
Pinto O Jr, Pint
o IRCA,
Naccarato KP. Maximum cl
o
ud-to-
g
rou
nd
li
ghtni
n
g
flas
h d
ens
itie
s obs
erved
b
y
light
nin
g
loc
a
ti
on s
y
stems in t
he trop
ical r
egi
on: A revie
w
.
J
ourn
a
l of At
mo
spher
ic Res
ear
ch.
200
6; 84
:
189-
200.
[3]
Bahar
udi
n Z
A
, N Azl
i
nd
a, F
e
rna
ndo
M, C
oora
y
V, Mäk
e
lä J
S
, M R
a
hman. E
l
ectric
field
ch
ang
es
gen
erate
d
b
y
t
he prel
imin
ar
y break
do
w
n
for
the
n
e
g
a
tive
cl
oud-to-
g
rou
n
d
l
i
ghtn
i
ng
flas
he
s in
Mal
a
ysi
a
and S
w
e
d
en.
Journ
a
l of Atmo
spher
ic Solar T
e
rrestrial P
h
ysi
cs
. 2012; 84-
8
5
: 16-24.
[4]
Bahar
udi
n Z
A
, N Azli
nd
a, F
e
rnan
do M, C
o
ora
y
V,
Mäk
e
l
ä
JS. Comp
ar
atative stu
d
y o
n
pre
limi
nar
y
break
do
w
n
p
u
l
s
e trains
obs
e
r
ved i
n
Jo
hor,
Mala
ys
ia
an
d
F
l
orid
a, USA.
Journ
a
l of
Atmos
p
h
e
ric
Research
. 20
1
2
; 117: 11
1-12
1.
[5]
N Azlin
da, M F
e
rnan
do, Bah
a
rud
i
n Z
A
, V
Coor
a
y
, Ahm
a
d H, Abdul Ma
lek Z
.
T
he characteristics o
f
Narro
w
Bip
o
lar
Pulses in
M
a
l
a
y
s
ia.
Jo
urn
a
l
of Atmos
p
h
e
ric
and S
o
l
a
r T
e
r
r
estrial P
h
ysic
s
. 2010;
72:
534-
540.
[6]
Rakov
VA, H
u
ffines GR.
R
e
turn-stroke
m
u
ltipl
i
cit
y
of n
egativ
e cl
ou
d-to-gro
und
l
i
ght
nin
g
F
l
as
hes
.
Journ
a
l of App
l
ied Mete
orol
og
y.
2003; 42: 14
55-1
462.
[7]
Darve
n
iza M, T
u
mma LR, Richter B, Rob
y
DA. Mu
ltipu
l
se
light
nin
g
curre
nt metal-o
x
i
de
Arrester.
IEEE
Tra
n
s
a
c
ti
on
s on
Po
we
r D
e
l
i
v
ery.
1997; 12 (3)
:
1168-1
1
7
5
.
[8]
N Azlin
da, M F
e
rna
ndo, B
aha
rudi
n Z
A
, M Rahma
n
, V Coor
a
y
, Z
i
a
d
Sa
leh,
Josep
h
R D
w
yer, H
a
mid
K
Rasso
ul. T
he first electric fie
l
d p
u
lse
of cl
o
ud a
nd c
l
o
ud-t
o
-gro
und
li
gtni
ng d
i
sch
arges.
Journ
a
l of
Atmos
pher
ic a
nd Sol
a
r-T
erre
strial Physics
. 201
0; 72(2-
3): 143-
150.
[9]
Abdu
lla
h N, Y
a
ha
ya
MP, Hu
di
NS.
Imple
m
en
tation and
us
e of
lig
htni
ng det
ection
netw
o
rk in
Ma
laysi
a,
Po
w
e
r and Energy
Confer
ence. PECon 2008. IEEE 2
nd
Internati
ona
l. 200
8; 383-3
86.
[10] N
Azlin
da.
Brban
d an
d HF
Radi
ation fro
m
Cl
ou
d F
l
as
hes an
d Narr
ow
Bipolar P
u
lses
. [
T
hesis].
Upps
ala: Acta
Univers
i
tatis U
p
sali
ens
is. Digi
t
al
Compr
ehe
n
s
ive Summar
i
e
s
of Uppsal
a
Dissertati
ons
from the F
a
culty of Scie
nc
e a
nd T
e
chnol
og
y. 2011; 82
2.
[11]
Martin A Uman
,
Lightni
ng
, Min
eol
a Ne
w
York;
Dover Pub
licat
ions, Inc. 198
2; 1-12.
[12]
Abdu
lla
h N,
H
a
tta NM.
Cl
ou
d-to-gro
und
li
g
h
tnin
g occ
u
rre
nces i
n
P
eni
ns
ular M
a
l
a
ysia
and
its us
e i
n
improve
m
ent of
distrib
u
tion
line
lig
htni
n
g
perfor
m
anc
es
. Po
w
e
r
a
nd En
erg
y
(
PECon).
IEEE
Internatio
na
l C
onfere
n
ce
o
n
. 201
2; 819-
822.
Evaluation Warning : The document was created with Spire.PDF for Python.