T
E
L
K
O
M
NIKA
T
elec
o
mm
un
ica
t
io
n Co
m
pu
t
i
ng
E
lect
ro
nics
a
nd
Co
ntr
o
l
Vo
l.
23
,
No
.
6
,
Dec
em
b
er
20
25
,
p
p
.
1
5
6
6
~1
578
I
SS
N:
1693
-
6
9
3
0
,
DOI
: 1
0
.
1
2
9
2
8
/
T
E
L
KOM
NI
K
A
.
v
23
i
6
.
27404
1566
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//jo
u
r
n
a
l.u
a
d
.
a
c.
id
/in
d
ex
.
p
h
p
/TELK
OM
N
I
K
A
M
a
chine learning
-
ba
sed recon
st
ruc
tion o
f
m
iss
ing
ra
infall
ex
tre
m
es:
a
co
m
p
a
ra
tive a
na
ly
sis
w
ith
cla
ss
ica
l
m
o
de
ls
Ya
nu
a
r
H
enry
P
riba
di
1,
2
,
T
a
nia
J
un
e
1
,
I
P
utu
Sa
nti
k
a
y
a
s
a
1
,
4
,
Su
pa
ri
3
,
Ana
T
ury
a
nti
1
1
D
e
p
a
r
t
me
n
t
o
f
G
e
o
p
h
y
si
c
s a
n
d
M
e
t
e
o
r
o
l
o
g
y
,
F
a
c
u
l
t
y
o
f
M
a
t
h
e
mat
i
c
s
a
n
d
N
a
t
u
r
a
l
S
c
i
e
n
c
e
s,
I
P
B
U
n
i
v
e
r
si
t
y
,
B
o
g
o
r
,
I
n
d
o
n
e
si
a
2
C
l
i
m
a
t
o
l
o
g
y
S
t
a
t
i
o
n
o
f
B
a
n
t
e
n
,
M
e
t
e
o
r
o
l
o
g
y
C
l
i
mat
o
l
o
g
y
a
n
d
G
e
o
p
h
y
si
c
s A
g
e
n
c
y
(
B
M
K
G
)
,
S
o
u
t
h
T
a
n
g
e
r
a
n
g
,
I
n
d
o
n
e
si
a
3
D
i
r
e
c
t
o
r
a
t
e
o
f
C
l
i
ma
t
e
C
h
a
n
g
e
,
M
e
t
e
o
r
o
l
o
g
y
C
l
i
mat
o
l
o
g
y
a
n
d
G
e
o
p
h
y
si
c
s A
g
e
n
c
y
(
B
M
K
G
)
,
Jak
a
r
t
a
,
I
n
d
o
n
e
si
a
4
C
e
n
t
r
e
f
o
r
C
l
i
ma
t
e
R
i
sk
a
n
d
O
p
p
o
r
t
u
n
i
t
y
M
a
n
a
g
e
me
n
t
i
n
S
o
u
t
h
e
a
st
A
si
a
P
a
c
i
f
i
c
(
C
C
R
O
M
–
S
EA
P
)
,
I
P
B
U
n
i
v
e
r
si
t
y
,
B
o
g
o
r
,
I
n
d
o
n
e
si
a
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l
15
,
2025
R
ev
i
s
ed
Sep
19
,
2
0
2
5
A
cc
ep
ted
Oct
19
,
2
0
2
5
T
h
e
li
m
it
e
d
a
v
a
il
a
b
il
it
y
o
f
d
a
il
y
ra
in
f
a
ll
d
a
ta
re
m
a
in
s
a
k
e
y
c
h
a
ll
e
n
g
e
in
ra
in
f
a
ll
d
a
ta
a
n
a
l
y
sis.
T
h
is
stu
d
y
a
s
se
ss
e
s
th
e
e
ffe
c
ti
v
e
n
e
s
s
o
f
sp
a
ti
a
l
in
terp
o
latio
n
a
n
d
b
ias
c
o
rre
c
ti
o
n
tec
h
n
iq
u
e
s
u
sin
g
sa
telli
te
-
d
e
riv
e
d
ra
in
f
a
ll
d
a
ta
to
f
il
l
m
issin
g
o
b
se
rv
a
ti
o
n
s
in
th
e
Ba
n
ten
a
n
d
Ja
k
a
rta
re
g
io
n
s.
T
h
re
e
in
terp
o
latio
n
m
e
th
o
d
s
i
n
v
e
rse
d
istan
c
e
we
ig
h
ti
n
g
(IDW
),
k
rig
in
g
,
a
n
d
s
p
li
n
e
w
e
r
e
c
o
m
p
a
re
d
.
Nin
e
sta
ti
stica
l
a
n
d
m
a
c
h
in
e
lea
rn
i
ng
-
b
a
se
d
b
ias
c
o
rre
c
ti
o
n
m
e
th
o
d
s
w
e
re
a
p
p
li
e
d
t
o
c
li
m
a
te
h
a
z
a
rd
s
g
ro
u
p
in
f
ra
re
d
p
re
c
ip
it
a
ti
o
n
w
it
h
sta
ti
o
n
d
a
ta
(
CHIR
P
S
)
,
m
u
lt
i
-
so
u
rc
e
w
e
ig
h
ted
-
e
n
se
m
b
le
p
re
c
ip
it
a
ti
o
n
(
M
S
W
EP
)
,
a
n
d
g
lo
b
a
l
p
re
c
ip
it
a
t
io
n
m
e
a
su
re
m
e
n
t
in
teg
ra
ted
m
u
lt
i
-
sa
telli
te
re
tri
e
v
a
ls
f
o
r
G
PM
(
G
P
M
I
M
ER
G
)
.
P
e
rf
o
r
m
a
n
c
e
wa
s
e
v
a
lu
a
ted
u
sin
g
ro
o
t
m
e
a
n
sq
u
a
re
e
rro
r
(
RM
S
E
)
,
m
e
a
n
a
b
so
lu
te
e
rro
r
(
M
A
E
)
,
b
ias
,
P
e
a
rso
n
c
o
rre
latio
n
(R),
a
n
d
Kli
n
g
-
G
u
p
ta
e
ff
icie
n
c
y
(
K
G
E
)
in
t
h
e
e
x
p
e
r
t
tea
m
o
n
c
li
m
a
te ch
a
n
g
e
d
e
tec
ti
o
n
a
n
d
in
d
i
c
e
s
(
E
T
CCDI
)
e
x
tre
m
e
in
d
e
x
.
T
h
e
re
se
a
rc
h
f
in
d
in
g
s
in
d
ica
te
th
a
t
CHIR
P
S
w
it
h
q
u
a
n
ti
le
m
a
p
p
in
g
(QM)
b
ias
c
o
rre
c
ti
o
n
d
e
li
v
e
rs
th
e
b
e
st
p
e
rf
o
rm
a
n
c
e
,
f
o
ll
o
w
e
d
b
y
ra
n
d
o
m
f
o
re
st
re
g
re
s
sio
n
(RF
R)
a
s
th
e
m
o
st
a
c
c
u
ra
te
m
a
c
h
in
e
lea
rn
in
g
m
e
th
o
d
.
I
n
s
p
a
ti
a
l
i
n
terp
o
lat
io
n
,
IDW
sta
n
d
s
o
u
t
a
s
th
e
lea
d
i
n
g
m
e
th
o
d
.
T
e
stin
g
th
e
e
x
trem
e
in
d
e
x
ET
CCDI
c
o
n
f
irms
th
a
t
CHIRP
S
-
Q
M
c
o
n
s
isten
tl
y
o
u
tp
e
rf
o
rm
s
m
a
c
h
in
e
lea
rn
in
g
a
n
d
in
terp
o
latio
n
m
e
th
o
d
s.
I
n
g
e
n
e
ra
l,
CHIRPS
-
QM
a
n
d
IDW
re
p
re
se
n
t
th
e
m
o
st
e
ffe
c
ti
v
e
c
o
m
b
in
a
ti
o
n
o
f
tec
h
n
iq
u
e
s
f
o
r
re
c
o
n
stru
c
t
in
g
d
a
il
y
ra
in
fa
ll
,
p
a
rti
c
u
larly
e
x
tre
m
e
e
v
e
n
ts.
T
h
is
stu
d
y
u
n
i
q
u
e
ly
in
teg
ra
tes
sp
a
ti
a
l
in
terp
o
latio
n
a
n
d
b
ias
c
o
rre
c
ti
o
n
i
n
a
u
n
if
ied
e
v
a
lu
a
ti
o
n
.
K
ey
w
o
r
d
s
:
B
ias co
r
r
ec
tio
n
E
x
p
er
t te
a
m
o
n
cli
m
ate
c
h
an
g
e
d
etec
tio
n
an
d
in
d
ice
s
E
x
tr
e
m
e
r
ai
n
f
a
ll
Ma
ch
i
n
e
lear
n
i
n
g
R
ain
f
all
d
ata
r
ec
o
n
s
tr
u
cti
n
g
Satellite r
ain
f
al
l d
ata
Sp
atial
in
ter
p
o
latio
n
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
T
an
ia
J
u
n
e
Dep
ar
t
m
en
t o
f
Geo
p
h
y
s
ics a
n
d
Me
teo
r
o
lo
g
y
,
Fac
u
lt
y
o
f
Ma
th
e
m
atics a
n
d
Natu
r
al
Sc
ien
ce
s
I
P
B
Un
iv
er
s
it
y
B
o
g
o
r
,
I
n
d
o
n
esia
E
m
ail: ta
n
iaj
u
n
e
@
ap
p
s
.
ip
b
.
ac
.
id
1.
I
NT
RO
D
UCT
I
O
N
R
ain
f
all
is
cr
u
cial
i
n
v
ar
io
u
s
cli
m
ate,
h
y
d
r
o
lo
g
y
,
a
n
d
n
at
u
r
al
d
is
as
ter
m
a
n
a
g
e
m
e
n
t
s
e
cto
r
s
.
T
h
e
ac
cu
r
ac
y
o
f
r
ai
n
f
all
d
ata
s
i
g
n
i
f
ican
tl
y
d
eter
m
i
n
es
th
e
r
eliab
ili
t
y
o
f
lo
n
g
-
ter
m
cli
m
ate
a
n
al
y
s
is
,
in
cl
u
d
i
n
g
f
lo
o
d
m
o
d
ell
in
g
,
ex
tr
e
m
e
in
d
e
x
ca
lcu
latio
n
s
,
an
d
cli
m
ate
tr
en
d
m
o
n
ito
r
in
g
.
C
li
m
ate
c
h
an
g
e
h
as
in
cr
ea
s
ed
th
e
f
r
eq
u
en
c
y
a
n
d
i
n
te
n
s
it
y
o
f
r
a
in
f
all
a
n
o
m
al
ies
in
tr
o
p
ical
r
eg
io
n
s
,
esp
ec
iall
y
t
h
o
s
e
a
f
f
ec
ted
b
y
t
h
e
E
l
N
iñ
o
So
u
t
h
er
n
Oscil
latio
n
(
E
NSO
)
p
h
en
o
m
e
n
o
n
[
1
]
,
[
2
]
.
Ho
w
e
v
e
r
,
th
er
e
ar
e
r
elate
d
ch
allen
g
e
s
,
as
th
e
d
is
tr
ib
u
tio
n
o
f
r
ain
f
all
m
ea
s
u
r
e
m
e
n
ts
i
n
I
n
d
o
n
esia
lack
s
lo
n
g
s
er
ies,
a
n
d
m
an
y
lo
ca
tio
n
s
ex
p
er
ie
n
ce
tec
h
n
ica
l
d
is
r
u
p
tio
n
s
,
in
f
r
astru
ct
u
r
e
li
m
i
tatio
n
s
,
o
r
m
is
s
i
n
g
ar
ch
i
v
es.
T
h
i
s
p
o
s
es
a
s
i
g
n
i
f
ica
n
t
c
h
alle
n
g
e
i
n
h
y
d
r
o
lo
g
ical
an
d
cl
i
m
a
te
an
al
y
s
is
,
e
s
p
ec
iall
y
i
n
ar
ea
s
with
co
m
p
le
x
to
p
o
g
r
ap
h
y
a
n
d
li
m
ited
o
b
s
er
v
atio
n
n
et
w
o
r
k
s
[
3
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Ma
ch
in
e
lea
r
n
in
g
-
b
a
s
ed
r
ec
o
n
s
tr
u
ctio
n
o
f m
is
s
in
g
r
a
in
fa
ll e
xtremes …
(
Ya
n
u
a
r
Hen
r
y
P
r
ib
a
d
i
)
1567
T
h
e
B
an
ten
an
d
J
ak
ar
ta
r
eg
io
n
s
ar
e
s
tr
ateg
ic
ar
ea
s
in
t
h
e
w
ester
n
p
ar
t
o
f
J
av
a
I
s
lan
d
at
h
ig
h
r
is
k
o
f
ex
tr
e
m
e
h
y
d
r
o
m
eteo
r
o
lo
g
ical
ev
en
t
s
.
J
ak
ar
ta,
a
m
etr
o
p
o
litan
ce
n
ter
w
it
h
a
h
i
g
h
p
o
p
u
latio
n
d
en
s
it
y
,
an
d
B
an
ten
,
th
e
s
o
u
r
ce
o
f
s
e
v
er
al
m
aj
o
r
r
i
v
er
b
asin
s
s
u
ch
as
C
iu
j
u
n
g
,
C
id
u
r
ian
,
an
d
C
id
an
au
(
C
3
)
,
o
f
ten
ex
p
er
ien
ce
f
lo
o
d
in
g
d
u
e
to
ex
tr
e
m
e
r
ai
n
f
all.
P
r
ev
io
u
s
s
tu
d
ie
s
h
a
v
e
i
n
d
icate
d
th
at
t
h
is
ar
ea
h
as
s
ig
n
i
f
ica
n
t
ch
ar
ac
t
er
is
tics
o
f
e
x
tr
e
m
e
d
ail
y
an
d
s
ea
s
o
n
al
r
ain
[
4
]
-
[
6
]
,
an
d
ev
e
n
r
ec
o
r
d
-
b
r
ea
k
in
g
r
ain
f
all
e
v
en
ts
h
av
e
ca
u
s
ed
s
u
b
s
tan
tial
ec
o
n
o
m
ic
lo
s
s
es a
n
d
f
lo
o
d
in
g
d
is
aster
s
i
n
J
ak
ar
ta
[
7
]
.
R
esear
ch
er
s
h
a
v
e
d
ev
e
lo
p
ed
v
ar
io
u
s
m
et
h
o
d
s
to
ad
d
r
ess
t
h
e
is
s
u
e
o
f
m
i
s
s
i
n
g
r
ain
f
all
d
at
a.
Glo
b
al
-
s
ca
le
r
esear
ch
o
n
d
ail
y
r
ai
n
f
a
ll
d
ata
i
m
p
u
tatio
n
h
a
s
u
tili
ze
d
d
iv
er
s
e
d
ataset
s
to
p
r
o
d
u
ce
co
n
tin
u
o
u
s
m
eteo
r
o
lo
g
ical
r
ec
o
r
d
s
b
y
i
n
t
eg
r
atin
g
r
ea
n
al
y
s
i
s
p
r
o
d
u
cts,
i
n
ter
n
at
io
n
al
s
tatio
n
d
ata,
an
d
v
ar
io
u
s
g
ap
-
f
illi
n
g
ap
p
r
o
ac
h
es,
in
clu
d
in
g
i
n
ter
p
o
latio
n
,
q
u
an
ti
le
m
ap
p
in
g
,
an
d
m
ac
h
in
e
lear
n
in
g
[
8
]
.
Mu
ltip
le
m
e
th
o
d
s
f
o
r
f
i
lli
n
g
m
is
s
i
n
g
d
ata
co
n
ti
n
u
e
to
e
v
o
lv
e
f
r
o
m
co
n
v
e
n
tio
n
al
s
tati
s
ti
ca
l
-
b
ased
ap
p
r
o
ac
h
es,
s
u
c
h
a
s
i
n
ter
p
o
latio
n
,
to
m
ac
h
in
e
lear
n
in
g
an
d
d
ee
p
lea
r
n
in
g
-
b
ased
m
et
h
o
d
s
,
w
h
ic
h
ca
n
r
ec
o
g
n
ize
co
m
p
le
x
p
atter
n
s
in
cli
m
ate
d
ata
[
9
]
.
T
h
e
p
e
r
f
o
r
m
a
n
ce
o
f
ea
ch
m
eth
o
d
is
g
r
ea
tl
y
in
f
l
u
e
n
ce
d
b
y
t
h
e
s
p
atial
d
is
tr
ib
u
tio
n
o
f
s
tatio
n
s
an
d
th
e
to
p
o
g
r
ap
h
ical
ch
ar
ac
ter
is
tics
o
f
th
e
ar
ea
[
1
0
]
,
[
1
1
]
.
I
n
v
er
s
e
d
is
tan
ce
w
e
ig
h
ti
n
g
(
I
DW
)
is
k
n
o
w
n
f
o
r
its
co
m
p
u
tatio
n
al
ef
f
icie
n
c
y
a
n
d
i
m
p
le
m
e
n
tat
io
n
w
it
h
i
n
g
eo
g
r
ap
h
ic
in
f
o
r
m
a
tio
n
s
y
s
te
m
(
GI
S
)
s
y
s
te
m
s
[
1
2
]
.
I
t
h
as
p
r
o
v
en
o
p
er
atio
n
all
y
ef
f
ec
ti
v
e
in
f
lat
r
eg
io
n
s
s
u
c
h
as
S
y
d
n
e
y
[
1
3
]
.
Me
an
w
h
ile,
th
e
s
p
li
n
e
in
t
er
p
o
latio
n
m
et
h
o
d
ten
d
s
to
b
e
m
o
r
e
s
en
s
iti
v
e
to
o
u
tlier
s
a
n
d
p
r
o
d
u
ce
s
ir
r
eg
u
lar
i
ties
in
lo
ca
tio
n
s
w
it
h
e
x
tr
e
m
e
d
ata
[
1
4
]
.
On
t
h
e
o
th
er
h
a
n
d
,
b
ias
co
r
r
ec
tio
n
ap
p
r
o
ac
h
es
ar
e
i
m
p
le
m
e
n
ted
to
i
m
p
r
o
v
e
th
e
esti
m
ate
s
o
f
s
ate
llit
e
p
r
o
d
u
cts
s
u
ch
as
cli
m
ate
h
a
za
r
d
s
g
r
o
u
p
in
f
r
ar
ed
p
r
ec
ip
it
atio
n
w
ith
s
tatio
n
d
ata
(
C
HI
R
P
S),
m
u
lti
-
s
o
u
r
c
e
w
ei
g
h
ted
-
e
n
s
e
m
b
le
p
r
ec
ip
itati
o
n
(
MSW
E
P
)
,
an
d
g
lo
b
al
p
r
e
cip
itatio
n
m
ea
s
u
r
e
m
e
n
t
in
teg
r
ated
m
u
lt
i
-
s
atellite
r
etr
iev
als
f
o
r
GP
M
(
G
P
M
I
M
E
R
G
)
to
b
etter
alig
n
w
it
h
lo
ca
l
co
n
d
itio
n
s
.
T
h
ese
p
r
o
d
u
cts
ar
e
b
en
ef
icial
b
ec
au
s
e
th
e
y
co
v
er
lar
g
e
ar
ea
s
an
d
h
av
e
h
ig
h
s
p
atial
r
e
s
o
lu
tio
n
,
b
u
t
t
h
e
y
co
n
ta
in
s
y
s
te
m
atic
b
i
ases
d
u
e
to
s
e
n
s
o
r
esti
m
atio
n
li
m
itat
io
n
s
.
B
ias
co
r
r
ec
tio
n
h
as
b
ee
n
p
er
f
o
r
m
ed
u
s
in
g
v
ar
io
u
s
ap
p
r
o
ac
h
es,
r
an
g
i
n
g
f
r
o
m
s
tati
s
tical
m
et
h
o
d
s
s
u
ch
as
q
u
an
t
ile
m
a
p
p
in
g
(
QM
)
a
n
d
li
n
ea
r
r
eg
r
es
s
io
n
(
LR
)
to
m
ac
h
i
n
e
lear
n
in
g
m
et
h
o
d
s
s
u
ch
a
s
r
an
d
o
m
f
o
r
est
o
r
XGB
o
o
s
t
[
1
5
]
-
[
1
7
]
.
T
h
e
ac
cu
r
ac
y
o
f
d
ail
y
r
ain
f
all
esti
m
ate
s
ca
n
b
e
s
ig
n
i
f
i
ca
n
tl
y
i
m
p
r
o
v
ed
b
y
co
r
r
ec
tin
g
t
h
e
C
HI
R
P
S
d
ata
[
1
6
]
,
[
1
8
]
.
P
r
ev
io
u
s
s
t
u
d
ies
h
a
v
e
d
e
m
o
n
s
tr
ated
t
h
at
m
ac
h
in
e
lear
n
in
g
ap
p
r
o
ac
h
es
ca
n
h
i
g
h
l
y
ca
p
t
u
r
e
co
m
p
le
x
p
atter
n
s
i
n
r
ain
f
all
d
ata.
R
e
s
ea
r
ch
in
I
n
d
o
n
esia,
s
u
c
h
as
[
1
9
]
s
u
cc
es
s
f
u
ll
y
ap
p
lied
m
ac
h
in
e
lear
n
i
n
g
tec
h
n
iq
u
es t
o
m
o
d
el
d
ail
y
r
ai
n
f
all
i
n
tr
o
p
ical
r
eg
io
n
s
w
it
h
h
i
g
h
v
ar
iab
ilit
y
.
Si
m
ilar
l
y
,
s
tu
d
ie
s
co
n
d
u
cted
in
I
n
d
ia
al
s
o
s
h
o
wed
i
m
p
r
o
v
ed
r
ain
f
all
p
r
ed
icti
o
n
ac
cu
r
ac
y
i
n
ar
ea
s
c
h
ar
ac
t
er
ized
b
y
m
o
n
s
o
o
n
cli
m
ate
s
[
2
0
]
.
A
lt
h
o
u
g
h
in
ter
p
o
latio
n
an
d
b
ias
co
r
r
ec
tio
n
h
av
e
b
ee
n
w
id
el
y
u
s
ed
,
th
er
e
ar
e
s
till
v
er
y
f
e
w
s
tu
d
ies
in
I
n
d
o
n
esia
th
a
t
in
te
g
r
ate
b
o
th
ap
p
r
o
ac
h
es
w
it
h
in
a
s
i
n
g
le
s
y
s
t
e
m
atic
ev
a
lu
atio
n
f
r
a
m
e
w
o
r
k
.
I
n
ex
tr
e
m
e
cli
m
at
e
an
al
y
s
is
,
ad
d
itio
n
al
v
alid
atio
n
b
ec
o
m
e
s
cr
u
cial
to
en
s
u
r
e
th
at
th
e
r
ec
o
n
s
tr
u
cted
d
ata
is
s
tatis
t
icall
y
ac
cu
r
ate
an
d
s
u
itab
le
f
o
r
ex
tr
e
m
e
a
n
al
y
s
i
s
.
T
h
er
ef
o
r
e,
th
is
r
esear
ch
ad
d
s
a
f
ea
s
ib
ili
t
y
te
s
t
f
o
r
th
e
r
ec
o
n
s
tr
u
cted
d
ata
b
y
ca
lcu
lati
n
g
ex
tr
e
m
e
i
n
d
ices
b
a
s
ed
o
n
th
e
e
x
p
er
t
tea
m
o
n
cli
m
ate
ch
an
g
e
d
etec
tio
n
a
n
d
i
n
d
ices
(
E
T
C
C
DI
)
,
w
h
ich
ar
e
w
id
el
y
u
s
ed
in
g
lo
b
al
clim
ate
s
t
u
d
ies
an
d
d
is
as
ter
r
is
k
ass
es
s
m
en
t
s
[
2
]
,
[
2
1
]
,
[
2
2
]
.
T
h
e
E
T
C
C
DI
in
d
ex
is
s
u
s
ce
p
tib
le
to
th
e
q
u
alit
y
o
f
i
n
p
u
t
d
ata
,
n
ec
e
s
s
itat
in
g
a
ca
r
e
f
u
l
ap
p
r
o
ac
h
to
en
s
u
r
e
i
ts
r
elia
b
ilit
y
,
i
n
cl
u
d
in
g
f
o
r
ass
es
s
i
n
g
d
is
aster
r
is
k
an
d
clim
ate
ch
a
n
g
e
to
av
o
id
m
i
s
i
n
ter
p
r
etatio
n
o
f
cli
m
ate
ch
an
g
e
s
i
g
n
al
s
[
2
3
]
-
[
2
5
]
.
T
h
e
ef
f
ec
tiv
e
n
e
s
s
o
f
i
m
p
u
tatio
n
m
et
h
o
d
s
is
s
i
g
n
i
f
ica
n
tl
y
in
f
l
u
en
ce
d
b
y
r
ain
f
all
v
ar
iab
ilit
y
a
n
d
th
e
s
p
atia
l
ch
ar
ac
ter
is
tic
s
o
f
t
h
e
ar
ea
,
as
d
em
o
n
s
tr
ated
i
n
[
2
6
]
.
T
h
er
ef
o
r
e,
m
et
h
o
d
s
elec
tio
n
s
h
o
u
ld
b
e
tailo
r
ed
to
th
e
an
al
y
tical
r
eq
u
ir
e
m
e
n
ts
,
p
ar
tic
u
lar
l
y
f
o
r
e
x
tr
e
m
e
ev
e
n
ts
.
I
t
w
a
s
n
o
ted
i
n
[
2
7
]
th
at
e
x
te
n
s
i
v
el
y
f
illed
h
is
to
r
ical
d
ata
m
a
y
a
f
f
ec
t
th
e
d
etec
tio
n
o
f
lo
n
g
-
ter
m
tr
e
n
d
s
in
ex
tr
e
m
e
i
n
d
ices,
esp
ec
iall
y
w
h
en
t
h
e
p
r
o
p
o
r
tio
n
o
f
m
is
s
in
g
d
ata
ex
ce
ed
s
a
ce
r
tain
th
r
e
s
h
o
l
d
.
T
h
is
s
tu
d
y
co
n
tr
ib
u
tes
to
f
i
lli
n
g
th
e
ex
is
ti
n
g
r
esear
ch
g
ap
b
y
s
y
s
te
m
atica
ll
y
i
n
te
g
r
atin
g
i
n
t
er
p
o
latio
n
m
et
h
o
d
s
an
d
b
ias
co
r
r
ec
tio
n
ac
r
o
s
s
m
u
ltip
le
s
atelli
te
r
ain
f
all
p
r
o
d
u
cts,
th
er
eb
y
e
n
h
a
n
ci
n
g
e
x
tr
e
m
e
r
ai
n
f
a
ll
an
al
y
s
is
u
s
i
n
g
E
T
C
C
DI
,
e
s
p
ec
iall
y
i
n
tr
o
p
ical
r
eg
io
n
s
.
T
h
e
u
n
iq
u
e
n
es
s
o
f
t
h
i
s
r
esear
ch
lies
i
n
its
co
m
p
r
eh
e
n
s
i
v
e
ev
alu
a
tio
n
f
r
a
m
e
w
o
r
k
,
w
h
ic
h
ai
m
s
n
o
t
o
n
l
y
to
co
m
p
a
r
e
in
ter
p
o
latio
n
m
e
th
o
d
s
a
n
d
b
ias
co
r
r
ec
tio
n
q
u
an
tita
tiv
e
l
y
b
u
t
also
to
test
th
e
r
esu
l
ts
o
f
t
h
ese
r
ec
o
n
s
tr
u
ctio
n
s
i
n
th
e
co
n
tex
t
o
f
E
T
C
C
DI
ex
tr
e
m
e
in
d
e
x
an
al
y
s
is
.
T
h
i
s
ap
p
r
o
ac
h
is
e
x
p
ec
ted
to
m
ak
e
m
et
h
o
d
o
lo
g
ical
co
n
tr
ib
u
tio
n
s
to
u
s
i
n
g
m
u
lt
i
-
ca
teg
o
r
y
r
ai
n
f
al
l
d
ata
f
o
r
ex
tr
e
m
e
cl
i
m
ate
s
t
u
d
ies in
I
n
d
o
n
esia.
2.
M
E
T
H
O
D
T
h
is
s
t
u
d
y
f
o
cu
s
es
o
n
th
e
B
a
n
t
en
an
d
J
ak
ar
ta
r
e
g
io
n
s
.
Ob
s
er
v
atio
n
al
r
ain
f
al
l
d
ata
w
er
e
o
b
tain
ed
f
r
o
m
th
e
Me
teo
r
o
lo
g
y
,
C
li
m
ato
lo
g
y
,
an
d
Geo
p
h
y
s
ics
Ag
e
n
c
y
(
B
MK
G)
,
p
ar
ticu
lar
l
y
f
r
o
m
th
e
B
an
ten
C
li
m
ato
lo
g
y
Statio
n
.
I
n
i
tiall
y
,
r
ain
f
all
r
ec
o
r
d
s
f
r
o
m
1
8
0
s
tatio
n
s
w
er
e
co
llected
,
an
d
f
o
llo
w
in
g
a
q
u
alit
y
co
n
tr
o
l
(
QC
)
p
r
o
ce
s
s
,
1
4
0
s
tatio
n
s
w
er
e
d
ee
m
ed
v
al
id
.
Am
o
n
g
th
e
s
e,
1
0
5
s
tatio
n
s
w
er
e
s
elec
ted
f
o
r
an
al
y
s
i
s
,
w
it
h
4
4
h
a
v
i
n
g
m
o
r
e
t
h
a
n
8
0
%
d
ail
y
d
ata
av
ailab
ilit
y
b
et
w
ee
n
1
9
9
1
an
d
2
0
2
0
,
w
h
ile
t
h
e
r
est
r
an
g
ed
f
r
o
m
4
1
%
to
8
0
%,
p
r
im
ar
il
y
d
u
e
to
d
ata
g
ap
s
i
n
t
h
e
ea
r
l
y
y
ea
r
s
.
T
h
e
s
p
atial
d
is
tr
ib
u
tio
n
o
f
t
h
ese
s
tatio
n
s
an
d
d
ata
av
ailab
ilit
y
is
illu
s
tr
ated
in
Fi
g
u
r
e
1
.
T
h
r
ee
s
atellite
-
b
ased
r
ain
f
all
p
r
o
d
u
cts
w
er
e
u
s
ed
:
C
HI
R
P
S
,
w
i
t
h
a
0
.
0
5
°
(
~5
k
m
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
23
,
No
.
6
,
Dec
em
b
er
20
25
:
1
5
6
6
-
1578
1568
r
eso
lu
tio
n
an
d
d
ata
a
v
ailab
l
e
f
r
o
m
1
9
8
1
[
2
8
]
;
MSW
E
P
,
w
it
h
a
0
.
1
°
(
~1
0
k
m
)
r
eso
l
u
tio
n
a
n
d
d
ata
s
tar
tin
g
f
r
o
m
1
9
7
9
[
2
9
]
;
an
d
GP
M
I
ME
R
G,
also
w
i
th
a
0
.
1
°
r
eso
lu
tio
n
,
av
ailab
le
f
r
o
m
1
9
9
8
[
3
0
]
.
T
h
ese
d
atasets
w
er
e
ac
ce
s
s
ed
f
r
o
m
t
h
eir
r
esp
ec
ti
v
e
o
f
f
icial
s
o
u
r
ce
s
.
Fig
u
r
e
1
.
Statio
n
l
o
ca
tio
n
s
an
d
r
ain
f
all
d
ata
a
v
ailab
ilit
y
in
B
a
n
ten
a
n
d
J
ak
ar
ta
T
h
e
s
i
m
u
latio
n
o
f
m
is
s
in
g
d
ata
r
ec
o
n
s
tr
u
ctio
n
u
s
ed
s
e
v
en
r
ai
n
s
ta
tio
n
s
w
it
h
co
m
p
lete
an
d
h
o
m
o
g
en
eo
u
s
ti
m
e
s
er
ies
f
r
o
m
1
9
9
1
to
2
0
2
0
:
Ke
m
a
y
o
r
an
,
T
an
j
u
n
g
P
r
io
k
,
So
ek
ar
n
o
Ha
tt
a,
C
u
r
u
g
,
T
an
g
er
a
n
g
Geo
p
h
y
s
ica
l,
B
an
ten
C
li
m
ato
l
o
g
ical,
an
d
Ser
an
g
Ma
r
iti
m
e
Me
teo
r
o
lo
g
ical
Statio
n
s
.
T
h
es
e
s
tatio
n
s
w
er
e
t
h
e
p
r
im
ar
y
r
ef
er
en
ce
f
o
r
test
i
n
g
in
ter
p
o
latio
n
an
d
b
ias
co
r
r
ec
tio
n
m
et
h
o
d
s
b
ef
o
r
e
b
r
o
ad
er
ap
p
licatio
n
.
I
n
th
e
s
i
m
u
lat
io
n
,
r
ain
f
all
d
ata
f
r
o
m
1
9
9
1
–
2
0
0
5
w
er
e
in
te
n
tio
n
all
y
r
e
m
o
v
ed
to
ev
al
u
ate
m
et
h
o
d
p
er
f
o
r
m
a
n
ce
.
T
h
e
m
is
s
i
n
g
d
ata
w
er
e
r
ec
o
n
s
tr
u
ct
ed
u
s
in
g
th
r
ee
i
n
ter
p
o
latio
n
m
eth
o
d
s
,
s
ix
s
tatis
tical
b
ias co
r
r
ec
tio
n
m
et
h
o
d
s
,
an
d
th
r
ee
m
ac
h
i
n
e
lear
n
in
g
-
b
ased
m
et
h
o
d
s
.
I
n
ter
p
o
latio
n
u
tili
ze
d
d
ata
f
r
o
m
1
0
5
o
th
er
s
tatio
n
s
,
w
h
ile
b
ias
co
r
r
ec
tio
n
w
a
s
b
ased
o
n
s
atellite
tr
ai
n
i
n
g
d
ata
f
r
o
m
2
0
0
6
to
2
0
2
0
to
est
i
m
ate
r
ai
n
f
al
l v
al
u
es
f
o
r
1
9
9
1
to
2
0
0
5
.
Satellite
p
r
o
d
u
ct
d
ata
is
u
s
ed
f
o
r
b
ias
co
r
r
ec
tio
n
m
et
h
o
d
s
,
w
h
er
e
th
e
ev
al
u
atio
n
an
d
b
ias
co
r
r
ec
tio
n
p
r
o
ce
s
s
es
ar
e
alig
n
ed
to
th
e
s
a
m
e
p
er
io
d
,
s
p
ec
if
icall
y
d
ata
f
r
o
m
1
9
9
1
to
2
0
2
0
,
ex
ce
p
t
f
o
r
GP
M
I
ME
R
G,
w
h
o
s
e
d
ata
is
av
ailab
le
f
r
o
m
1
9
9
8
.
Su
b
s
eq
u
e
n
tl
y
,
th
e
d
o
w
n
lo
ad
ed
d
ata
w
er
e
ex
tr
ac
ted
at
lo
ca
tio
n
s
m
atc
h
i
n
g
th
e
r
ai
n
g
au
g
e
co
o
r
d
in
ates
f
o
r
e
v
alu
a
ti
o
n
an
d
b
ias
co
r
r
ec
tio
n
p
u
r
p
o
s
es.
I
n
s
u
m
m
ar
y
,
t
h
e
f
lo
w
o
f
t
h
e
r
esear
ch
p
r
o
ce
d
u
r
e
ca
n
b
e
s
ee
n
i
n
Fi
g
u
r
e
2
.
T
h
is
s
t
u
d
y
e
m
p
lo
y
s
v
ar
io
u
s
in
te
r
p
o
latio
n
an
d
b
ias
co
r
r
ec
tio
n
m
et
h
o
d
s
,
w
it
h
t
h
e
in
ter
p
o
latio
n
m
eth
o
d
s
u
s
ed
i
n
c
lu
d
in
g
s
p
li
n
e,
k
r
ig
i
n
g
,
an
d
I
D
W
.
Fo
r
th
e
C
HI
R
P
S,
MSW
E
P
,
an
d
GP
M
I
ME
R
G
s
atellite
p
r
o
d
u
cts,
th
e
s
tat
is
tic
al
b
ias
co
r
r
ec
tio
n
m
et
h
o
d
s
u
s
ed
ar
e
av
er
ag
e
r
atio
(
A
V
G)
,
QM
,
L
R
,
li
n
ea
r
b
ias
co
r
r
ec
tio
n
(
L
B
C
)
,
r
o
b
u
s
t
r
eg
r
ess
io
n
(
R
R
)
,
an
d
w
ei
g
h
ted
r
eg
r
ess
io
n
(
W
R
)
,
as
w
ell
as
t
h
e
m
ac
h
i
n
e
lear
n
i
n
g
m
et
h
o
d
s
r
an
d
o
m
f
o
r
est r
e
g
r
es
s
io
n
(
R
F
R
)
,
XGB
o
o
s
t
r
eg
r
ess
i
o
n
(
XG)
,
an
d
is
o
latio
n
f
o
r
est (
I
F).
Sp
lin
e
in
ter
p
o
latio
n
co
n
s
tr
u
ct
s
a
s
m
o
o
th
s
u
r
f
ac
e
th
a
t m
i
n
i
m
iz
es
cu
r
v
at
u
r
e
b
y
s
ee
k
i
n
g
an
i
n
te
r
p
o
latio
n
f
u
n
ctio
n
(
,
)
th
at
m
in
i
m
izes
th
e
s
q
u
ar
e
d
i
f
f
er
e
n
ce
s
b
et
w
ee
n
th
e
d
ata
an
d
th
e
s
u
r
f
ac
e,
w
h
i
le
co
n
t
r
o
llin
g
t
h
e
s
m
o
o
th
n
e
s
s
o
f
th
e
s
u
r
f
ac
e
th
r
o
u
g
h
t
h
e
s
m
o
o
t
h
n
e
s
s
p
ar
am
eter
λ
w
it
h
(
1
)
.
T
h
e
ef
f
ec
t
iv
en
e
s
s
o
f
s
p
li
n
e
in
ter
p
o
latio
n
co
m
p
ar
ed
to
I
DW
h
as
also
b
ee
n
e
v
alu
a
ted
in
p
r
ev
io
u
s
r
ain
f
all
s
tu
d
ie
s
,
w
h
ic
h
h
i
g
h
li
g
h
t
th
e
s
u
p
er
io
r
p
er
f
o
r
m
a
n
ce
o
f
I
DW
in
v
ar
io
u
s
cli
m
atic
co
n
d
it
io
n
s
[
1
0
]
.
(
)
=
∑
[
−
(
,
)
]
2
=
1
+
λ
∬
[
(
∂
2
∂
2
)
2
+
(
∂
2
∂
2
)
2
]
(
1
)
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Ma
ch
in
e
lea
r
n
in
g
-
b
a
s
ed
r
ec
o
n
s
tr
u
ctio
n
o
f m
is
s
in
g
r
a
in
fa
ll e
xtremes …
(
Ya
n
u
a
r
Hen
r
y
P
r
ib
a
d
i
)
1569
Fig
u
r
e
2
.
R
esear
ch
f
lo
w
ch
ar
t
o
f
r
ain
f
all
d
ata
r
ec
o
n
s
tr
u
c
tio
n
m
et
h
o
d
s
T
h
e
k
r
ig
i
n
g
m
et
h
o
d
tak
e
s
in
t
o
ac
co
u
n
t
t
h
e
s
p
atial
co
r
r
elati
o
n
b
et
w
ee
n
p
o
in
t
s
.
T
h
is
m
eth
o
d
u
s
es
a
s
e
m
i
v
ar
io
g
r
a
m
to
esti
m
ate
s
p
atial
au
to
co
r
r
elatio
n
w
it
h
(
2
)
.
I
n
k
r
i
g
in
g
,
̂
(
0
)
is
th
e
e
s
ti
m
ated
v
alu
e
a
t
th
e
lo
ca
tio
n
0
,
an
d
λ
is
th
e
s
p
atial
w
ei
g
h
t
o
b
tain
ed
f
r
o
m
t
h
e
s
e
m
i
v
ar
io
g
r
a
m
m
o
d
el.
Kr
ig
i
n
g
r
elies
o
n
a
s
e
m
i
v
ar
io
g
r
a
m
to
e
s
ti
m
ate
s
p
atial
au
to
co
r
r
elatio
n
,
a
n
d
h
as
b
ee
n
s
h
o
w
n
to
o
u
tp
er
f
o
r
m
n
o
n
-
g
eo
s
tatis
tical
in
ter
p
o
latio
n
m
et
h
o
d
s
in
tr
o
p
ical
r
eg
io
n
s
[
3
1
]
.
T
h
e
I
DW
m
e
th
o
d
esti
m
ate
s
th
e
v
al
u
e
at
a
s
p
ec
if
ic
p
o
in
t
b
ased
o
n
th
e
d
is
ta
n
ce
to
o
b
s
er
v
atio
n
p
o
in
ts
u
s
in
g
(
3
)
.
T
h
is
m
et
h
o
d
esti
m
ates t
h
e
r
ain
f
all
v
a
lu
e
at
a
p
o
in
t
0
b
ased
o
n
th
e
d
is
ta
n
ce
to
t
h
e
o
b
s
er
v
at
io
n
p
o
in
t
s
.
Her
e,
(
)
is
t
h
e
r
ai
n
f
all
v
alu
e
at
t
h
e
p
o
in
t
,
is
t
h
e
d
is
t
an
ce
f
r
o
m
p
o
in
t
0
to
p
o
i
n
t
1
,
a
n
d
is
th
e
d
i
s
t
an
c
e
w
e
ig
h
ti
n
g
p
a
r
am
e
t
e
r
(
=
2
)
.
T
h
e
v
alu
e
o
f
2
f
o
r
t
h
e
d
is
t
an
c
e
w
ei
g
h
ti
n
g
p
a
r
am
et
e
r
p
r
o
v
i
d
es
an
o
p
t
im
al
b
a
l
a
n
ce
b
e
tw
ee
n
t
h
e
n
e
a
r
e
s
t
d
a
ta
p
o
i
n
ts
’
c
o
n
t
r
i
b
u
t
i
o
n
an
d
th
e
in
t
e
r
p
o
l
a
t
i
o
n
r
e
s
u
lt
s
’
s
p
a
t
i
al
s
t
a
b
il
i
ty
[
3
2
]
-
[
3
4
]
.
T
h
e
AVG
m
eth
o
d
is
a
c
o
r
r
e
c
t
i
o
n
t
h
a
t
in
v
o
lv
es
m
u
l
ti
p
ly
in
g
th
e
s
ate
l
l
i
te
v
a
lu
e
b
y
th
e
r
a
t
i
o
b
e
t
w
e
en
th
e
av
e
r
ag
e
o
b
s
e
r
v
a
ti
o
n
a
n
d
th
e
a
v
e
r
ag
e
s
a
t
el
l
i
t
e
d
u
r
in
g
th
is
c
a
li
b
r
a
t
i
o
n
p
er
i
o
d
in
(
4
)
.
T
h
i
s
m
e
th
o
d
c
an
b
e
u
s
e
d
a
s
a
b
aseli
n
e
to
co
m
p
ar
e
w
it
h
o
th
er
m
et
h
o
d
s
[
3
5
]
.
̂
(
0
)
=
∑
λ
(
)
=
1
(
2
)
̂
(
0
)
=
∑
(
)
=
1
∑
1
=
1
(
3
)
c
o
rr
(
)
=
(
)
⋅
(
4
)
T
h
e
QM
m
eth
o
d
ca
n
ad
d
r
ess
b
ias
in
a
v
er
ag
es
a
n
d
ex
tr
e
m
e
q
u
an
tile
s
i
n
(
5
)
.
QM
ad
j
u
s
ts
th
e
q
u
a
n
til
e
d
is
tr
ib
u
tio
n
o
f
s
ate
llit
e
d
ata
w
ith
o
b
s
er
v
atio
n
a
l
d
ata.
is
th
e
cu
m
u
lat
iv
e
d
is
tr
ib
u
tio
n
f
u
n
cti
o
n
(
C
DF)
o
f
th
e
s
atellite
d
ata,
an
d
−
1
is
th
e
i
n
v
e
r
s
e
C
DF
o
f
th
e
o
b
s
e
r
v
a
ti
o
n
al
d
a
t
a
.
T
h
is
m
et
h
o
d
h
as
b
e
en
d
em
o
n
s
t
r
a
t
e
d
t
o
e
f
f
e
ct
iv
e
ly
im
p
r
o
v
e
C
H
I
R
P/
C
H
I
R
PS
r
a
in
f
al
l
es
t
im
at
es
an
d
s
u
p
p
o
r
t
l
o
n
g
-
t
e
r
m
h
y
d
r
o
l
o
g
i
ca
l
a
p
p
l
i
c
a
t
i
o
n
s
[
1
6
]
.
LR
m
e
th
o
d
s
,
a
l
th
o
u
g
h
s
im
p
l
e
,
a
r
e
s
t
i
l
l
c
o
m
m
o
n
ly
u
s
e
d
i
n
c
o
r
r
ec
t
in
g
c
lim
a
t
e
d
a
t
a
b
i
as
t
o
f
o
r
m
a
t
im
e
s
e
r
i
e
s
o
f
c
o
r
r
e
c
t
i
o
n
r
e
s
u
lt
s
b
a
s
e
d
o
n
t
h
e
h
is
t
o
r
i
c
al
r
e
la
t
i
o
n
s
h
i
p
b
e
tw
ee
n
t
h
e
o
b
s
e
r
v
a
ti
o
n
a
l
d
a
ta
an
d
s
a
te
l
l
i
te
m
o
d
e
l
o
u
t
p
u
ts
[
1
7
]
.
T
h
i
s
m
eth
o
d
is
f
o
u
n
d
in
(
6
)
.
A
s
im
p
l
e
l
in
ea
r
m
o
d
e
l
w
h
e
r
e
i
s
th
e
i
n
t
e
r
c
e
p
t
an
d
i
s
th
e
s
l
o
p
e
d
e
t
e
r
m
in
e
d
b
y
r
e
g
r
es
s
i
o
n
f
i
t
ti
n
g
b
etw
e
en
o
b
s
er
v
a
ti
o
n
al
an
d
s
a
t
e
ll
i
t
e
d
a
t
a
.
T
h
e
R
o
b
u
s
t
l
in
e
a
r
c
o
r
r
e
ct
i
o
n
m
e
th
o
d
i
s
u
s
e
d
e
x
p
l
i
c
it
ly
f
o
r
u
p
p
e
r
q
u
an
t
i
l
es
(
ex
t
r
em
e
s
)
t
o
av
o
id
o
v
er
f
itti
n
g
a
n
d
i
m
p
r
o
v
e
p
r
ed
ictio
n
s
tab
ilit
y
[
3
6
]
.
R
R
is
a
m
o
d
el
r
esis
tan
t
to
o
u
tlier
s
,
u
s
i
n
g
a
n
M
-
est
i
m
a
to
r
ap
p
r
o
ac
h
.
T
h
e
W
R
m
eth
o
d
is
s
i
m
ilar
to
LR
,
b
u
t
w
ei
g
h
t
s
ass
ig
n
ed
b
ased
o
n
th
e
v
ar
ia
n
ce
o
f
s
atel
lite
d
a
ta
at
p
o
in
t
i,
s
o
p
o
in
t
s
w
it
h
lo
w
u
n
ce
r
tain
t
y
ar
e
p
r
io
r
itized
.
Mo
d
el
r
esil
ien
ce
to
o
u
tlier
s
is
al
s
o
an
i
m
p
o
r
ta
n
t c
o
n
s
id
er
atio
n
,
alt
h
o
u
g
h
t
h
e
r
o
b
u
s
t
n
es
s
asp
ec
t is
n
o
t e
x
p
licit
l
y
r
ef
lecte
d
in
(
7
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
23
,
No
.
6
,
Dec
em
b
er
20
25
:
1
5
6
6
-
1578
1570
c
o
rr
(
)
=
−
1
(
(
(
)
)
)
(
5
)
c
or
r
(
)
=
+
⋅
(
)
(
6
)
c
or
r
(
)
=
+
⋅
(
)
,
=
1
Va
r
(
)
(
7
)
T
h
e
R
FR
m
et
h
o
d
is
an
en
s
e
m
b
le
m
o
d
el
o
f
T
d
ec
is
io
n
tr
ee
s
,
w
h
er
e
th
e
f
in
al
r
es
u
lt
is
t
h
e
av
er
ag
e
o
u
tp
u
t
o
f
all
tr
ee
s
ℎ
(
)
,
u
s
i
n
g
(
8
)
.
T
h
is
m
e
th
o
d
h
a
s
b
ee
n
e
x
te
n
s
i
v
el
y
u
s
e
d
f
o
r
GP
M
I
ME
R
G
co
r
r
ec
tio
n
[
3
7
]
.
T
h
e
XG
m
et
h
o
d
h
as
b
ee
n
p
r
o
v
en
to
h
av
e
a
h
ig
h
b
ias
co
r
r
ec
tio
n
ca
p
a
b
ilit
y
o
n
p
r
ec
ip
itatio
n
d
ata
in
m
o
u
n
tai
n
o
u
s
ar
ea
s
,
w
it
h
b
ia
s
v
al
u
es
s
i
g
n
i
f
ica
n
tl
y
s
m
aller
t
h
a
n
co
n
v
e
n
tio
n
al
s
tat
is
tical
ap
p
r
o
ac
h
es
[
3
8
]
.
T
h
e
b
o
o
s
tin
g
m
o
d
el
ad
j
u
s
t
s
th
e
p
r
ev
io
u
s
p
r
ed
ictio
n
(
−
1
)
̂
w
it
h
a
n
e
w
tr
ee
f
u
n
ctio
n
(
)
,
m
u
ltip
lie
d
b
y
t
h
e
lear
n
in
g
r
ate
η
(
9
)
.
R
a
n
d
o
m
p
ar
titi
o
n
in
g
is
u
s
ed
to
d
etec
t
o
u
tlier
s
an
d
r
ec
tify
e
x
tr
e
m
e
v
alu
es.
T
h
e
IF
m
et
h
o
d
h
as
b
ee
n
s
h
o
w
n
to
en
h
a
n
ce
th
e
ac
cu
r
ac
y
o
f
p
r
ec
ip
itatio
n
d
ata
d
is
tr
ib
u
tio
n
an
d
d
e
m
o
n
s
t
r
ate
s
u
p
er
io
r
it
y
in
p
r
eser
v
in
g
p
h
y
s
icall
y
r
ele
v
a
n
t
ex
tr
e
m
e
v
al
u
es
w
it
h
o
u
t c
o
m
p
r
o
m
is
i
n
g
m
o
d
el
s
tab
ilit
y
[
3
9
]
.
̂
=
1
∑
ℎ
(
)
=
1
(
8
)
(
)
̂
=
(
−
1
)
̂
+
η
(
)
(
9
)
T
h
e
r
esu
lts
o
f
e
s
ti
m
ati
n
g
r
ai
n
f
all
ar
e
co
m
p
ar
ed
w
i
th
ac
tu
al
d
ata
f
r
o
m
s
e
v
en
B
MK
G
s
tat
io
n
s
th
a
t
h
a
v
e
co
m
p
lete
d
ata
an
d
ar
e
e
v
al
u
at
ed
u
s
i
n
g
er
r
o
r
m
etr
ic
i
n
d
icato
r
s
,
s
p
ec
if
icall
y
R
MSE
,
M
A
E
,
B
ias,
an
d
co
r
r
elatio
n
co
ef
f
icie
n
t
(
R
)
.
T
h
ese
er
r
o
r
m
etr
ic
ca
lcu
latio
n
s
d
eter
m
i
n
e
th
e
b
est
m
et
h
o
d
f
o
r
f
illi
n
g
in
m
i
s
s
i
n
g
d
ail
y
r
ain
f
al
l
d
ata,
esp
ec
ially
f
o
r
h
ig
h
an
d
ex
tr
e
m
e
r
ai
n
f
al
l
ev
en
ts
.
T
h
e
eq
u
atio
n
s
f
o
r
th
ese
er
r
o
r
m
etr
ic
s
ca
n
b
e
f
o
u
n
d
in
(
10
)
-
(
13
)
.
W
ith
as
th
e
v
alu
e
o
f
th
e
esti
m
ated
r
esu
lt
(
in
ter
p
o
lated
r
esu
lt
o
r
b
ias
co
r
r
ec
tio
n
)
,
as
th
e
o
b
s
er
v
atio
n
v
alu
e
(
ac
t
u
al
r
ain
f
all
d
ata)
,
̅
an
d
̅
as
th
e
esti
m
ated
an
d
o
b
s
er
v
ed
d
ata
av
er
ag
es
,
an
d
n
as
th
e
to
tal
n
u
m
b
er
o
f
d
ail
y
d
ata
p
o
in
ts
co
m
p
ar
ed
.
=
√
1
∑
(
−
)
2
=
1
(
1
0
)
=
1
∑
|
−
|
=
1
(
1
1
)
=
1
∑
(
−
)
=
1
(
1
2
)
=
∑
(
−
̅
)
(
−
̅
)
=
1
√
∑
(
−
̅
)
2
∑
(
−
̅
)
2
=
1
=
1
(
1
3
)
An
e
v
al
u
ati
v
e
ap
p
r
o
ac
h
to
t
w
o
t
y
p
es
o
f
r
ain
f
al
l
d
ata
r
e
co
n
s
tr
u
ct
io
n
m
et
h
o
d
s
,
n
a
m
el
y
s
p
atial
in
ter
p
o
latio
n
an
d
s
atellite
p
r
o
d
u
ct
b
ias
co
r
r
ec
tio
n
,
w
a
s
u
s
ed
in
th
i
s
s
tu
d
y
.
T
h
e
ev
alu
at
io
n
is
co
n
d
u
cted
n
o
t
o
n
l
y
o
n
p
er
f
o
r
m
an
ce
b
ased
o
n
g
en
er
al
s
tatis
tic
s
b
u
t
al
s
o
o
n
its
ab
ilit
y
to
m
ai
n
tai
n
t
h
e
s
i
g
n
al
o
f
ex
tr
e
m
e
ev
e
n
ts
,
h
ig
h
li
g
h
ti
n
g
t
h
e
i
m
p
o
r
tan
ce
o
f
s
e
lecti
n
g
i
m
p
u
tatio
n
m
et
h
o
d
s
b
ased
o
n
s
p
atial
v
ar
iab
ilit
y
a
n
d
s
en
s
i
tiv
it
y
to
ex
tr
e
m
e
s
[
2
6
]
.
T
h
en
,
to
en
s
u
r
e
th
e
r
eliab
ilit
y
o
f
t
h
e
f
illed
d
ata
in
lo
n
g
-
ter
m
an
al
y
s
is
,
E
T
C
C
D
I
in
d
ex
ca
lcu
latio
n
test
s
ar
e
co
n
d
u
cted
.
T
h
is
ap
p
r
o
ac
h
tak
es
i
n
to
ac
co
u
n
t
t
h
e
f
i
n
d
in
g
s
i
n
[
2
7
]
,
w
h
ich
i
n
d
icate
s
th
at
u
s
i
n
g
i
n
-
f
il
led
d
ata
ca
n
s
ig
n
if
ican
tl
y
af
f
ec
t
t
h
e
d
etec
tio
n
o
f
ex
tr
e
m
e
tr
en
d
s
,
esp
ec
iall
y
w
h
e
n
th
e
p
r
o
p
o
r
t
io
n
o
f
f
illed
d
ata
is
s
ig
n
i
f
ica
n
t.
A
cc
u
r
ate
f
o
r
ec
asti
n
g
o
f
th
e
E
T
C
C
DI
i
n
d
ex
i
s
h
ea
v
i
l
y
in
f
l
u
e
n
ce
d
b
y
t
h
e
co
m
p
leten
e
s
s
o
f
d
ail
y
r
ain
f
al
l
d
ata,
as
in
co
m
p
lete
n
es
s
o
f
d
ata
d
u
r
i
n
g
s
p
ec
if
ic
p
er
io
d
s
w
ill
lead
to
b
iases
in
e
x
tr
e
m
e
tr
en
d
ca
lcu
latio
n
s
,
th
u
s
r
eq
u
ir
i
n
g
e
f
f
ec
ti
v
e
d
ata
r
ec
o
n
s
tr
u
ctio
n
tech
n
iq
u
e
s
[
4
0
]
.
Fil
lin
g
i
n
r
ain
d
ata
w
it
h
g
r
id
d
ed
d
ata
o
r
in
ter
p
o
latio
n
r
es
u
lts
ca
n
r
ed
u
c
e
p
ea
k
v
al
u
es
a
n
d
th
e
f
r
eq
u
e
n
c
y
o
f
e
x
tr
e
m
e
v
al
u
es
;
h
e
n
ce
,
s
tr
ict
v
er
i
f
icatio
n
i
s
r
eq
u
ir
ed
b
ef
o
r
e
th
e
d
ata
is
u
s
e
d
in
cli
m
ate
c
h
a
n
g
e
a
n
al
y
s
i
s
[
2
4
]
.
T
h
e
r
esu
lts
o
f
f
i
lli
n
g
i
n
m
i
s
s
i
n
g
d
ata
also
d
e
p
en
d
o
n
th
e
q
u
ality
o
f
th
e
i
n
p
u
t
d
ata,
i
n
cl
u
d
in
g
d
ata
d
er
iv
ed
f
r
o
m
i
n
ter
p
o
latio
n
,
as
it
ca
n
af
f
ec
t
th
e
tr
en
d
s
p
r
o
d
u
ce
d
in
th
e
ex
tr
e
m
e
in
d
e
x
[
4
1
]
,
[
4
2
]
.
W
ith
th
e
p
r
ev
io
u
s
r
esear
ch
,
th
e
ca
lcu
latio
n
o
f
th
e
E
T
C
C
DI
in
d
ex
in
t
h
is
s
t
u
d
y
w
il
l b
e
u
s
ed
f
o
r
clim
a
te
ch
ar
ac
ter
izatio
n
an
d
as a
n
ad
d
itio
n
al
v
alid
atio
n
test
in
g
to
o
l
f
o
r
th
e
i
n
ter
p
o
latio
n
r
es
u
lt
s
an
d
b
ias
co
r
r
ec
tio
n
.
T
h
e
ex
tr
e
m
e
E
T
C
C
DI
i
n
d
ices
to
b
e
test
e
d
in
clu
d
e
1
4
in
d
ices
b
ased
o
n
th
eir
in
te
n
s
i
t
y
,
s
u
c
h
a
s
m
a
x
i
m
u
m
1
-
d
a
y
p
r
ec
ip
itatio
n
a
m
o
u
n
t
(
R
x
1
d
a
y
)
,
m
ax
i
m
u
m
5
-
d
a
y
p
r
ec
ip
itatio
n
am
o
u
n
t
(
R
x
5
d
a
y
)
,
s
i
m
p
le
d
ail
y
in
te
n
s
it
y
i
n
d
ex
(
SDI
I
)
,
an
n
u
al
to
tal
w
et
-
d
a
y
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Ma
ch
in
e
lea
r
n
in
g
-
b
a
s
ed
r
ec
o
n
s
tr
u
ctio
n
o
f m
is
s
in
g
r
a
in
fa
ll e
xtremes …
(
Ya
n
u
a
r
Hen
r
y
P
r
ib
a
d
i
)
1571
p
r
ec
ip
itatio
n
(
P
R
C
PT
O
T
)
,
p
er
ce
n
tag
e
o
f
to
tal
p
r
ec
ip
it
atio
n
f
r
o
m
d
a
y
s
ex
ce
ed
in
g
t
h
e
9
0
th
p
er
ce
n
tile
(
P
9
0
p
T
O
T
)
,
p
er
ce
n
tag
e
o
f
t
o
tal
p
r
ec
ip
itatio
n
f
r
o
m
d
a
y
s
ex
ce
ed
in
g
th
e
9
5
th
p
er
ce
n
ti
le
(
P
9
5
p
T
O
T
)
,
an
d
p
er
ce
n
tag
e
o
f
to
tal
p
r
ec
ip
itati
o
n
f
r
o
m
d
a
y
s
e
x
ce
ed
in
g
th
e
9
9
th
p
er
ce
n
tile
(
P
9
9
p
T
O
T
)
;
b
ased
o
n
f
r
eq
u
e
n
c
y
,
n
u
m
b
e
r
o
f
h
ea
v
y
p
r
ec
ip
itatio
n
d
ay
s
(
≥
1
0
m
m
)
(
R
1
0
mm
)
,
n
u
m
b
er
o
f
v
er
y
h
ea
v
y
p
r
ec
ip
ita
tio
n
d
a
y
s
(
≥
2
0
m
m
)
(
R
2
0
mm
)
,
n
u
m
b
er
o
f
e
x
tr
e
m
el
y
h
ea
v
y
p
r
ec
ip
itatio
n
d
a
y
s
(
≥
5
0
m
m
)
(
R
5
0
mm
)
,
n
u
m
b
er
o
f
ex
tr
e
m
el
y
h
ea
v
y
p
r
ec
ip
itatio
n
d
ay
s
(
≥
1
0
0
m
m
)
(
R
1
0
0
mm
)
,
an
d
n
u
m
b
er
o
f
ex
tr
e
m
el
y
h
ea
v
y
p
r
ec
ip
itatio
n
d
ay
s
(
≥
1
5
0
m
m
)
(
R
1
5
0
mm
)
;
an
d
b
ased
o
n
d
u
r
atio
n
,
co
n
s
ec
u
ti
v
e
w
et
d
a
y
s
(
C
W
D
)
an
d
co
n
s
ec
u
ti
v
e
d
r
y
d
a
y
s
(
C
D
D
)
.
T
h
e
ex
tr
e
m
e
in
d
e
x
r
esu
lts
f
r
o
m
t
h
e
b
est
i
n
ter
p
o
latio
n
,
an
d
all
s
atellite
b
ias
co
r
r
ec
tio
n
s
ar
e
m
ea
s
u
r
ed
u
s
in
g
a
co
m
b
i
n
atio
n
o
f
f
o
u
r
m
etr
ics
:
R
MSE
,
B
ias,
R
,
an
d
au
g
m
e
n
t
ed
w
i
th
k
l
in
g
-
g
u
p
ta
ef
f
icie
n
c
y
(
KGE
)
.
On
e
o
f
th
e
n
e
w
m
etr
ic
s
u
s
ed
is
th
e
KG
E
,
as
p
r
o
p
o
s
ed
in
[
4
3
]
,
w
h
ic
h
co
m
b
in
e
s
th
r
ee
m
ai
n
ev
al
u
atio
n
co
m
p
o
n
en
t
s
:
co
r
r
elatio
n
,
m
ea
n
b
ias
r
atio
,
an
d
v
ar
iatio
n
r
atio
,
as
s
h
o
w
n
in
(
1
4
)
.
W
h
er
e
r
is
th
e
P
ea
r
s
o
n
co
r
r
elatio
n
co
ef
f
ic
ien
t,
is
th
e
m
ea
n
v
a
lu
e,
=
/
is
th
e
co
ef
f
icien
t
o
f
v
ar
iatio
n
,
an
d
s
u
b
s
cr
ip
ts
s
an
d
o
r
ef
er
to
s
i
m
u
la
ted
an
d
o
b
s
er
v
ed
d
ata,
r
esp
ec
tiv
el
y
.
KGE
=
1
−
√
(
−
1
)
2
+
(
−
1
)
2
+
(
−
1
)
2
(
1
4
)
A
d
d
in
g
KGE
co
m
p
le
m
en
t
s
co
n
v
e
n
tio
n
al
e
v
al
u
atio
n
m
etr
ic
s
,
as
it
ca
n
ca
p
tu
r
e
th
e
m
o
d
el
’
s
o
v
er
al
l
ac
cu
r
ac
y
.
Nev
er
t
h
ele
s
s
,
R
MS
E
,
b
ias,
an
d
R
r
em
ai
n
r
etain
e
d
,
as
ea
ch
p
r
o
v
id
es
s
p
ec
if
ic
in
s
i
g
h
ts
n
o
t
al
w
a
y
s
r
ef
lecte
d
in
th
e
a
g
g
r
eg
ate
KG
E
s
co
r
e
[
4
3
]
.
I
n
ass
es
s
i
n
g
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
s
e
f
o
u
r
m
et
r
ics,
th
eir
v
alu
e
s
ar
e
n
o
r
m
alize
d
i
n
to
a
r
an
g
e
o
f
0
–
1
b
ef
o
r
e
th
e
to
tal
s
co
r
e
is
co
m
p
u
ted
.
T
h
e
p
u
r
p
o
s
e
o
f
ap
p
l
y
i
n
g
n
o
r
m
aliza
t
io
n
is
to
en
s
u
r
e
th
at
all
m
etr
ics
co
n
tr
ib
u
te
eq
u
all
y
a
n
d
d
o
n
o
t
d
o
m
in
ate
ea
ch
o
t
h
er
in
t
h
e
ev
al
u
atio
n
p
r
o
ce
s
s
,
th
u
s
en
s
u
r
in
g
tr
an
s
p
ar
en
c
y
i
n
th
e
s
tan
d
ar
d
ized
f
r
a
m
e
w
o
r
k
[
4
4
]
.
T
o
co
n
s
o
lid
ate
th
e
f
o
u
r
m
etr
ics
in
to
a
s
in
g
le
p
er
f
o
r
m
a
n
ce
m
ea
s
u
r
e,
a
n
o
r
m
aliza
tio
n
p
r
o
ce
s
s
i
s
ca
r
r
ied
o
u
t
w
h
er
e
t
h
e
n
o
r
m
a
lizatio
n
f
o
r
m
u
la
f
o
r
R
MSE
a
n
d
b
ias
is
d
e
s
cr
ib
ed
b
y
(
15
)
,
an
d
th
e
n
o
r
m
aliza
t
io
n
f
o
r
m
u
la
f
o
r
KGE
is
d
e
f
i
n
ed
b
y
(
16
)
.
T
h
e
ab
s
o
lu
t
e
s
ig
n
|
|
is
u
s
ed
ex
p
lici
tl
y
f
o
r
m
etr
ic
s
s
en
s
iti
v
e
to
d
ir
ec
tio
n
(
b
ias).
M
e
tr
ik
n
o
rm
=
1
−
|
|
−
|
|
|
|
−
|
|
(
1
5
)
M
e
tr
ik
n
o
rm
=
|
|
−
|
|
|
|
−
|
|
(
1
6
)
T
h
e
to
tal
s
co
r
e
o
f
ea
ch
m
e
th
o
d
is
ca
lcu
lated
u
s
i
n
g
th
e
w
ei
g
h
te
d
av
er
ag
e
o
f
th
e
f
o
u
r
m
etr
ics
n
o
r
m
al
ized
by
(
17
)
.
Fo
r
o
b
j
ec
tiv
it
y
i
n
th
e
class
i
f
icatio
n
o
f
in
ter
v
al
s
c
o
r
es,
th
e
to
tal
s
co
r
e
in
ter
v
als
in
th
i
s
s
t
u
d
y
ar
e
d
eter
m
in
ed
b
ased
o
n
th
e
q
u
ar
ti
les
o
f
th
e
o
v
er
all
to
tal
s
co
r
e
d
is
tr
ib
u
tio
n
.
T
h
is
allo
w
s
p
er
f
o
r
m
an
ce
e
v
al
u
atio
n
to
b
e
ca
r
r
ied
o
u
t a
cc
o
r
d
in
g
to
th
e
p
r
in
cip
le
o
f
d
is
tr
ib
u
tio
n
-
b
ase
d
class
if
icatio
n
,
w
h
ic
h
h
a
s
b
ee
n
w
id
el
y
ap
p
lied
i
n
ass
es
s
i
n
g
cl
i
m
a
te
an
d
h
y
d
r
o
lo
g
y
m
o
d
el
p
er
f
o
r
m
a
n
ce
[
4
5
]
.
E
ac
h
ca
teg
o
r
y
in
ter
v
al
w
i
ll
t
h
en
b
e
ass
ig
n
ed
a
r
atin
g
,
s
u
c
h
as
‘
lo
w
,
’
‘
f
air
,
’
‘
h
i
g
h
,
’
a
n
d
‘
v
er
y
h
ig
h
,
’
to
r
ep
r
esen
t
a
s
ta
tis
t
icall
y
co
n
s
i
s
te
n
t
r
an
g
e
o
f
v
al
u
es
w
h
ile
b
ei
n
g
r
elev
an
t
f
o
r
d
if
f
er
en
tiat
in
g
m
e
th
o
d
s
b
ased
o
n
th
e
q
u
ali
t
y
o
f
r
esu
lt
s
.
=
0
.
25
×
(
1
−
R
M
SE
n
o
rm
)
+
0
.
25
×
(
1
−
B
ia
s
n
o
rm
)
+
0
.
25
×
KGE
n
o
rm
+
0
.
25
×
|
|
(
1
7
)
T
h
e
u
s
e
o
f
th
e
E
T
C
C
DI
in
d
e
x
in
th
i
s
s
t
u
d
y
is
i
n
te
n
d
e
d
as
p
ar
t
o
f
th
e
in
ter
n
al
ev
a
lu
atio
n
to
ass
es
s
th
e
co
n
s
is
ten
c
y
o
f
d
ata
f
illi
n
g
a
g
a
in
s
t
th
e
c
h
ar
ac
ter
is
tic
s
o
f
e
x
tr
e
m
e
p
r
ec
ip
itatio
n
.
I
n
t
h
i
s
s
t
u
d
y
,
th
e
R
MSE
,
B
I
AS,
an
d
KGE
m
etr
ics
ar
e
f
ir
s
t
n
o
r
m
alize
d
b
ef
o
r
e
b
ein
g
u
s
ed
in
t
h
e
co
m
p
o
s
ite
s
co
r
e
w
ei
g
h
ti
n
g
s
y
s
t
e
m
,
ex
ce
p
t
f
o
r
th
e
co
r
r
elatio
n
v
al
u
e,
w
h
ic
h
is
o
n
l
y
ta
k
e
n
in
ab
s
o
lu
te
ter
m
s
as
it
alr
ea
d
y
h
a
s
a
v
al
u
e
r
an
g
e
o
f
0
-
1
.
T
h
is
n
o
r
m
aliza
t
io
n
s
tep
is
ca
r
r
ied
o
u
t
s
o
th
at
ea
ch
m
etr
ic
is
o
n
an
eq
u
al
s
ca
le
an
d
ca
n
b
e
c
o
m
p
a
r
ed
f
air
ly
.
W
ith
o
u
t
n
o
r
m
aliza
t
io
n
,
d
if
f
er
e
n
c
es
i
n
u
n
i
ts
a
n
d
v
al
u
e
r
a
n
g
e
s
a
m
o
n
g
m
etr
ics
ca
n
lead
to
t
h
e
d
o
m
i
n
an
ce
o
f
o
n
e
m
etr
ic
o
v
er
an
o
th
er
,
r
es
u
lti
n
g
i
n
b
ias
ed
ass
ess
m
e
n
t
s
.
T
h
is
ap
p
r
o
ac
h
ali
g
n
s
w
it
h
t
h
e
p
r
ac
tices
p
r
o
p
o
s
ed
in
[
2
6
]
,
[
4
5
]
,
w
h
er
e
a
co
m
p
o
s
ite
ev
al
u
atio
n
s
y
s
te
m
f
o
r
cli
m
ate
m
o
d
els
w
a
s
d
ev
elo
p
ed
b
y
tr
an
s
f
o
r
m
in
g
a
ll e
r
r
o
r
m
etr
ics i
n
to
a
r
an
g
e
o
f
0
–
1
to
m
i
n
i
m
ize
b
ias
ca
u
s
ed
b
y
d
if
f
er
en
ce
s
in
s
ca
le
an
d
s
en
s
iti
v
it
y
,
en
s
u
r
in
g
th
at
ea
c
h
in
d
icato
r
co
n
tr
ib
u
tes
p
r
o
p
o
r
tio
n
all
y
to
th
e
f
i
n
al
ass
e
s
s
m
e
n
t.
An
i
m
p
o
r
tan
t
n
o
te
in
th
i
s
ap
p
r
o
ac
h
i
s
th
at
all
ca
lcu
late
d
p
er
f
o
r
m
a
n
ce
in
d
ice
s
ar
e
b
ased
s
o
lel
y
o
n
th
e
h
i
g
h
e
s
t
an
d
lo
w
e
s
t
r
ef
er
en
ce
v
al
u
e
s
th
at
e
m
er
g
e
f
r
o
m
th
i
s
r
esear
ch
an
d
ar
e
n
o
t
a
g
lo
b
ally
ac
ce
p
ted
s
tan
d
ar
d
,
th
u
s
allo
w
in
g
f
o
r
a
c
o
n
s
is
te
n
t
an
d
r
elev
a
n
t
ev
alu
atio
n
o
f
t
h
e
d
ata
co
n
tex
t
a
n
d
r
esear
ch
ar
ea
.
Nev
er
th
eles
s
,
th
e
f
i
n
al
r
esu
lts
w
il
l
n
o
t
co
n
cl
u
d
e
th
at
t
h
e
m
e
th
o
d
i
s
g
e
n
er
all
y
r
eliab
le.
I
n
s
tead
,
th
e
r
es
u
lt
s
ar
e
th
e
m
o
s
t o
p
ti
m
al
a
m
o
n
g
th
e
te
s
ted
m
eth
o
d
s
an
d
ar
e
li
m
i
ted
to
th
is
s
tu
d
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
23
,
No
.
6
,
Dec
em
b
er
20
25
:
1
5
6
6
-
1578
1572
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
3
.
1
.
Da
t
a
o
bs
er
v
a
t
io
n,
inte
r
po
la
t
io
n,
a
nd
bia
s
co
rr
ec
t
io
n
T
h
is
s
ec
tio
n
p
r
esen
t
s
th
e
p
er
f
o
r
m
a
n
ce
ev
a
lu
at
io
n
o
f
d
ail
y
r
ain
f
all
esti
m
atio
n
u
s
i
n
g
s
p
atia
l
in
ter
p
o
latio
n
a
n
d
b
ias
co
r
r
ec
tio
n
o
f
t
h
r
ee
s
atelli
te
p
r
o
d
u
cts
.
T
a
b
le
1
s
u
m
m
ar
izes
t
h
e
r
es
u
lts
b
a
s
ed
o
n
f
o
u
r
s
tatis
t
ical
m
etr
ics.
Am
o
n
g
in
t
er
p
o
latio
n
m
et
h
o
d
s
,
I
DW
p
e
r
f
o
r
m
ed
b
est
w
i
th
th
e
lo
w
est
R
MSE
(
1
2
.
0
0
m
m
)
,
MA
E
(
5
.
3
7
m
m
)
,
b
ias
(
–
0
.
7
1
m
m
)
,
a
n
d
th
e
h
i
g
h
e
s
t
co
r
r
elatio
n
(
R
=
0
.
4
7
)
.
Kr
ig
in
g
s
h
o
w
ed
s
l
ig
h
tl
y
lo
w
er
co
r
r
elatio
n
th
a
n
I
DW
,
w
h
ile
s
p
lin
e
y
ield
ed
u
n
r
ea
lis
t
icall
y
h
ig
h
er
r
o
r
s
(
R
MSE
>
1
0
0
m
m
)
,
in
d
icati
n
g
p
o
o
r
s
u
it
ab
il
it
y
f
o
r
d
ail
y
d
ata.
Am
o
n
g
s
atelli
te
p
r
o
d
u
cts,
C
HI
R
P
S
w
i
th
LR
a
n
d
GP
M
I
ME
R
G
w
it
h
IF
p
r
o
d
u
ce
d
R
MSE
v
al
u
es
ar
o
u
n
d
1
2
.
9
9
m
m
,
o
u
tp
er
f
o
r
m
i
n
g
MSW
E
P
,
w
h
o
s
e
R
M
SE
co
n
s
is
te
n
tl
y
ex
ce
ed
ed
1
3
m
m
,
r
ea
ch
in
g
2
9
.
9
3
m
m
u
n
d
er
q
u
an
tile
m
ap
p
in
g
.
Ho
w
e
v
er
,
n
o
n
e
o
f
th
e
b
ias
co
r
r
ec
tio
n
m
et
h
o
d
s
ac
h
ie
v
ed
a
h
ig
h
e
r
co
r
r
elatio
n
th
a
n
I
DW
;
all
R
v
alu
e
s
r
e
m
ain
ed
b
elo
w
0
.
2
6
.
T
h
ese
f
i
n
d
in
g
s
a
lig
n
w
i
th
[
1
8
]
,
w
h
ich
h
i
g
h
li
g
h
t
s
C
HI
R
P
S
’
s
r
eliab
ilit
y
i
n
tr
o
p
ical
cli
m
ate
s
,
an
d
w
i
th
[
3
6
]
,
d
em
o
n
s
tr
atin
g
t
h
e
i
m
p
r
o
v
e
m
en
t
o
f
ex
tr
e
m
e
r
ain
f
al
l
ac
cu
r
ac
y
t
h
r
o
u
g
h
h
y
b
r
id
q
u
an
t
ile
m
ap
p
in
g
.
C
HI
R
P
S
co
n
s
is
te
n
tl
y
s
h
o
w
ed
lo
w
b
ias
an
d
m
o
r
e
s
tab
le
co
r
r
elatio
n
th
an
GP
M
I
ME
R
G,
m
a
k
in
g
it
th
e
m
o
s
t
ac
c
u
r
ate
s
atellite
p
r
o
d
u
ct
in
th
i
s
s
t
u
d
y
.
Nev
er
t
h
eles
s
,
o
th
er
s
tu
d
ies
[
3
7
]
,
[
3
8
]
n
o
ted
th
at
GP
M
I
ME
R
G,
w
h
e
n
co
r
r
ec
ted
w
it
h
R
an
d
o
m
Fo
r
est,
p
er
f
o
r
m
s
b
etter
i
n
m
o
u
n
ta
in
o
u
s
r
e
g
io
n
s
.
T
h
e
s
elec
tio
n
o
f
C
HI
R
P
S
h
er
e
also
r
ef
lect
s
its
lo
n
g
er
d
ata
av
ailab
ilit
y
,
w
h
ich
s
u
its
th
e
te
m
p
o
r
al
n
ee
d
s
o
f
th
i
s
r
esear
ch
.
T
ab
le
1
.
Daily
r
ain
f
all
e
s
ti
m
at
io
n
er
r
o
r
s
co
m
p
ar
ed
to
o
b
s
er
v
atio
n
s
(
1
9
9
1
–
2
0
0
5
)
M
e
t
h
o
d
T
y
p
e
P
r
o
d
u
c
t
E
r
r
o
r
me
t
r
i
c
v
a
l
u
e
R
M
S
E
(
mm
)
M
A
E
(
mm
)
B
IAS
(
mm
)
R
I
D
W
I
n
ter
po
lat
i
o
n
Obs
e
r
v
a
t
i
o
n
1
2
.
0
0
0
4
5
.
3
7
2
9
-
0
.
7
1
1
6
0
.
4
6
5
3
K
r
i
g
i
n
g
1
4
.
1
4
9
7
7
.
6
0
7
7
-
0
.
5
6
4
4
0
.
4
1
7
0
S
p
l
i
n
e
1
0
6
.
2
4
1
2
4
4
.
9
5
4
1
3
8
.
7
3
3
4
0
.
1
3
0
3
A
v
e
r
a
g
e
r
a
t
i
o
S
t
a
t
i
st
i
c
a
l
C
h
i
r
p
s
1
3
.
7
4
4
1
6
.
7
3
6
4
-
0
.
0
7
5
1
0
.
2
5
0
4
L
i
n
e
a
r
b
i
a
s
c
o
r
r
1
2
.
9
9
5
8
6
.
8
3
2
0
0
.
0
1
7
1
0
.
2
5
0
4
L
i
n
e
a
r
re
g
1
2
.
9
9
5
8
6
.
8
3
2
0
0
.
0
1
7
1
0
.
2
5
0
4
Q
u
a
n
t
i
l
e
m
a
p
p
i
n
g
1
6
.
7
7
0
5
7
.
2
6
4
0
-
0
.
0
9
7
7
0
.
2
1
0
7
R
o
b
u
st
r
e
g
1
3
.
5
6
3
8
5
.
1
6
8
4
-
3
.
9
9
8
6
0
.
2
5
0
4
W
e
i
g
h
t
e
d
r
eg
1
3
.
4
7
0
2
6
.
4
4
0
9
-
0
.
5
8
1
2
0
.
2
5
0
4
R
a
n
d
o
m
f
o
r
e
st
r
e
g
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
1
4
.
9
7
0
6
7
.
3
5
2
9
-
0
.
0
8
3
1
0
.
1
3
7
2
X
G
B
o
o
st
re
g
1
3
.
2
9
2
5
6
.
9
0
4
1
-
0
.
0
2
7
4
0
.
2
1
8
5
I
so
l
a
t
i
o
n
f
o
r
e
st
1
3
.
0
0
3
5
5
.
9
9
2
0
-
1
.
8
6
6
0
0
.
2
5
0
4
A
v
e
r
a
g
e
r
a
t
i
o
S
t
a
t
i
st
i
c
a
l
M
sw
e
p
1
7
.
3
6
3
6
6
.
9
9
6
3
0
.
4
2
4
5
0
.
2
0
2
2
L
i
n
e
a
r
b
i
a
s
c
o
r
r
1
7
.
0
4
1
4
7
.
1
0
5
2
0
.
4
1
5
8
0
.
2
0
2
2
L
i
n
e
a
r
re
g
1
7
.
0
4
1
4
7
.
1
0
5
2
0
.
4
1
5
8
0
.
2
0
2
2
Q
u
a
n
t
i
l
e
ma
p
p
i
n
g
2
9
.
9
3
3
1
9
.
3
4
9
5
2
.
3
3
8
6
0
.
1
9
6
5
R
o
b
u
st
r
eg
1
3
.
6
4
2
5
5
.
2
8
6
4
-
3
.
5
3
5
4
0
.
2
0
2
2
W
e
i
g
h
t
e
d
r
eg
1
6
.
5
3
0
3
6
.
6
8
7
9
-
0
.
1
0
2
8
0
.
2
0
2
2
R
a
n
d
o
m
f
o
r
e
st
r
eg
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
1
6
.
2
0
3
5
7
.
0
4
9
8
0
.
2
0
2
0
0
.
2
1
5
9
X
G
B
o
o
st
r
eg
1
4
.
8
5
2
4
6
.
6
7
3
5
-
0
.
0
4
7
0
0
.
2
4
7
9
I
so
l
a
t
i
o
n
f
o
r
e
st
1
3
.
9
6
4
7
5
.
9
2
6
2
-
1
.
5
6
6
8
0
.
2
0
2
2
A
v
e
r
a
g
e
ra
tio
S
t
a
t
i
st
i
c
a
l
G
P
M
i
me
r
g
1
5
.
0
3
9
8
6
.
7
3
7
5
-
0
.
2
5
8
5
0
.
2
2
0
7
L
i
n
e
a
r
b
i
a
s
c
o
r
r
1
3
.
4
1
6
3
6
.
7
9
0
1
-
0
.
0
7
1
3
0
.
2
2
0
7
L
i
n
e
a
r
r
eg
1
3
.
4
1
6
3
6
.
7
9
0
1
-
0
.
0
7
1
3
0
.
2
2
0
7
Q
u
a
n
t
i
l
e
ma
p
p
i
n
g
1
7
.
9
0
7
9
7
.
2
1
6
1
-
0
.
2
3
3
8
0
.
2
0
5
7
R
o
b
u
st
r
e
g
1
3
.
3
3
2
0
5
.
1
6
2
0
-
3
.
6
1
9
9
0
.
2
2
0
7
W
e
i
g
h
t
e
d
r
e
g
1
4
.
8
2
7
0
6
.
6
1
2
9
-
0
.
4
6
2
6
0
.
2
2
0
7
R
a
n
d
o
m
f
o
r
e
st
r
eg
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
1
5
.
0
1
2
8
7
.
1
5
5
3
-
0
.
1
7
9
2
0
.
1
7
9
4
X
G
B
o
o
st
re
g
1
3
.
3
9
6
1
6
.
7
1
7
1
-
0
.
1
6
5
2
0
.
2
3
1
5
I
so
l
a
t
i
o
n
f
o
r
e
s
1
2
.
9
9
3
8
5
.
9
4
7
6
-
1
.
7
2
3
9
0
.
2
2
0
7
Fig
u
r
e
3
illu
s
tr
ate
s
s
ca
tter
p
lo
ts
co
m
p
ar
in
g
o
b
s
er
v
ed
an
d
in
te
r
p
o
lated
d
aily
r
ai
n
f
a
ll
v
al
u
es u
s
in
g
th
r
ee
m
et
h
o
d
s
.
Sp
lin
e
in
ter
p
o
latio
n
s
h
o
w
s
t
h
e
w
ea
k
e
s
t
ag
r
ee
m
en
t
w
it
h
o
b
s
er
v
ed
d
ata
(
R
=0
.
1
2
)
,
w
it
h
a
w
id
e
s
p
r
ea
d
b
elo
w
th
e
1
:1
lin
e,
p
ar
ticu
lar
l
y
o
v
er
esti
m
ati
n
g
h
i
g
h
-
i
n
te
n
s
it
y
r
ai
n
f
all.
T
h
is
s
u
g
g
est
s
th
e
s
p
lin
e’
s
s
m
o
o
t
h
i
n
g
n
atu
r
e
i
s
less
s
u
ited
f
o
r
d
ail
y
r
ain
f
all
w
it
h
s
u
b
s
tan
tial
lo
ca
l
v
ar
iab
ilit
y
,
as
n
o
ted
in
[
1
0
]
,
[
1
4
]
.
Kr
ig
in
g
an
d
I
DW
p
r
o
v
id
e
b
etter
alig
n
m
e
n
t
to
o
b
s
er
v
atio
n
s
,
esp
ec
iall
y
I
DW
,
w
it
h
t
h
e
h
i
g
h
e
s
t
co
r
r
elatio
n
(
R
=0
.
6
1
)
,
r
ef
lectin
g
s
tr
o
n
g
er
co
n
s
i
s
te
n
c
y
ac
r
o
s
s
s
t
atio
n
s
an
d
in
te
n
s
i
ties
.
B
o
th
m
eth
o
d
s
s
h
o
w
m
o
r
e
co
n
ce
n
tr
ate
d
clu
s
ter
in
g
ar
o
u
n
d
th
e
id
ea
l
lin
e,
a
f
f
ir
m
i
n
g
th
eir
s
u
itab
il
it
y
u
n
d
er
s
p
ar
s
e
s
tatio
n
co
n
d
itio
n
s
.
T
h
ese
f
i
n
d
in
g
s
alig
n
w
it
h
p
r
ev
io
u
s
r
esear
ch
[
4
6
]
s
u
p
p
o
r
tin
g
I
D
W
an
d
k
r
ig
in
g
’
s
s
tr
e
n
g
th
s
in
p
r
eser
v
in
g
s
p
atial
ac
cu
r
ac
y
w
i
th
o
u
t
ex
tr
e
m
e
d
ev
iatio
n
s
.
Nev
er
th
ele
s
s
,
s
p
lin
e
m
et
h
o
d
s
r
e
m
ain
r
elev
a
n
t
in
s
ea
s
o
n
al
-
s
ca
le
ap
p
licatio
n
s
d
u
e
to
th
eir
s
m
o
o
th
i
n
g
ca
p
ac
it
y
ac
r
o
s
s
lo
n
g
er
te
m
p
o
r
al
ag
g
r
e
g
atio
n
s
[
1
3
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Ma
ch
in
e
lea
r
n
in
g
-
b
a
s
ed
r
ec
o
n
s
tr
u
ctio
n
o
f m
is
s
in
g
r
a
in
fa
ll e
xtremes …
(
Ya
n
u
a
r
Hen
r
y
P
r
ib
a
d
i
)
1573
Fig
u
r
e
3
.
T
h
e
r
elatio
n
s
h
ip
b
etw
ee
n
o
b
s
er
v
atio
n
a
n
d
in
ter
p
o
l
ated
r
ain
f
all
(
s
p
li
n
e,
k
r
i
g
i
n
g
,
a
n
d
I
DW
)
Fig
u
r
e
4
d
is
p
lay
s
t
h
e
p
er
f
o
r
m
an
ce
o
f
n
i
n
e
b
ias
co
r
r
ec
tio
n
m
eth
o
d
s
ap
p
lied
to
C
HI
R
P
S
s
atellite
d
ata,
av
er
ag
ed
o
v
er
s
ev
e
n
r
ai
n
g
a
u
g
e
s
tatio
n
s
.
T
h
e
m
eth
o
d
s
in
cl
u
d
e
th
r
ee
s
tatis
tical
ap
p
r
o
ac
h
es
A
V
G,
L
B
C
,
an
d
L
R
,
an
d
th
r
ee
r
e
g
r
ess
io
n
-
b
ased
tec
h
n
iq
u
es
QM
,
R
R
,
a
n
d
W
R
,
al
o
n
g
s
id
e
th
r
ee
m
ac
h
i
n
e
lear
n
i
n
g
-
b
a
s
ed
m
o
d
els
I
F,
R
F
R
,
an
d
X
G.
T
h
e
R
FR
,
XG,
an
d
QM
m
et
h
o
d
s
v
is
u
all
y
p
r
o
d
u
ce
s
ca
tter
p
lo
ts
w
it
h
p
o
in
t
s
m
o
r
e
co
n
ce
n
tr
ated
alo
n
g
t
h
e
1
:1
r
ef
er
en
ce
lin
e,
s
u
g
g
es
tin
g
s
tr
o
n
g
er
ag
r
ee
m
e
n
t
w
it
h
o
b
s
er
v
ed
r
ain
f
a
ll
v
al
u
es.
T
h
is
in
d
icate
s
th
eir
s
u
p
er
io
r
ab
ilit
y
to
r
ed
u
ce
s
y
s
te
m
a
tic
b
ias
a
n
d
ca
p
tu
r
e
v
a
r
iab
ilit
y
,
p
ar
ticu
lar
l
y
i
n
ex
tr
e
m
e
r
ain
f
all
ca
s
es,
co
n
s
is
ten
t
w
i
th
f
i
n
d
in
g
s
i
n
[
1
5
]
,
[
2
5
]
,
[
3
7
]
,
[
3
8
]
.
Q
M
s
t
an
d
s
o
u
t
in
h
an
d
l
in
g
h
ig
h
-
in
te
n
s
ity
r
a
i
n
f
al
l
ev
en
ts
(
>
1
0
0
m
m
/
d
ay
)
,
w
h
il
e
s
im
p
l
e
r
m
eth
o
d
s
s
u
ch
as
A
VG
an
d
L
B
C
s
h
o
w
m
o
r
e
d
is
p
e
r
s
e
d
a
n
d
b
ia
s
ed
c
o
r
r
e
c
t
i
o
n
s
,
o
f
t
en
u
n
d
e
r
e
s
t
im
a
tin
g
h
e
av
y
r
ain
f
a
l
l
.
I
F
a
n
d
R
R
a
r
e
p
r
o
n
e
t
o
u
n
d
e
r
p
r
e
d
i
c
t
i
o
n
,
e
s
p
e
c
i
al
ly
f
o
r
ex
t
r
em
e
ev
en
ts
,
a
n
d
t
en
d
t
o
p
r
o
d
u
c
e
o
u
t
p
u
ts
c
lu
s
t
e
r
e
d
a
t
l
o
w
e
r
v
al
u
e
s
.
M
e
an
w
h
i
l
e
,
W
R
s
h
o
w
s
a
l
im
it
e
d
d
i
s
t
r
i
b
u
t
i
o
n
r
an
g
e
,
s
u
g
g
e
s
t
in
g
l
o
w
a
d
a
p
t
a
b
i
l
ity
t
o
th
e
o
b
s
e
r
v
e
d
v
a
r
i
a
b
i
l
ity
.
T
h
e
s
e
r
e
s
u
lt
s
h
ig
h
lig
h
t
t
h
at
m
a
ch
in
e
l
e
a
r
n
in
g
m
e
th
o
d
s
a
n
d
q
u
an
t
il
e
-
b
a
s
ed
c
o
r
r
e
c
t
i
o
n
s
o
f
f
e
r
b
e
t
te
r
f
l
ex
i
b
il
ity
an
d
a
c
cu
r
ac
y
th
an
c
o
n
v
en
ti
o
n
a
l
l
in
e
a
r
m
e
th
o
d
s
,
a
s
s
u
p
p
o
r
t
ed
b
y
[
3
5
]
,
[
3
9
]
.
Fig
u
r
e
4
.
C
o
m
p
ar
is
o
n
o
f
r
ai
n
f
all
b
ias co
r
r
ec
tio
n
an
d
o
b
s
er
v
atio
n
s
f
o
r
C
HI
R
P
S
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
23
,
No
.
6
,
Dec
em
b
er
20
25
:
1
5
6
6
-
1578
1574
3
.
2
.
E
v
a
lua
t
i
o
n o
f
perf
o
rm
a
nce
o
f
m
et
ho
ds
ba
s
ed
o
n t
he
E
T
C
CDI
e
x
t
re
m
e
ind
e
x
Fig
u
r
e
5
p
r
esen
ts
s
ca
tter
p
lo
ts
co
m
p
ar
i
n
g
o
b
s
er
v
ed
v
alu
e
s
an
d
p
r
ed
icted
r
esu
lts
f
r
o
m
ea
ch
d
ata
-
f
illi
n
g
m
et
h
o
d
ac
r
o
s
s
f
o
u
r
E
T
C
C
DI
in
d
ices:
i
n
te
n
s
it
y
(
R
x
1
d
a
y
,
P
R
C
P
T
O
T
)
,
f
r
eq
u
en
c
y
(
R
2
0
m
m
)
,
a
n
d
d
u
r
atio
n
(
C
W
D)
.
T
h
e
d
ash
ed
1
:1
lin
e
m
ar
k
s
t
h
e
id
ea
l
ag
r
ee
m
e
n
t;
p
r
ed
ictio
n
s
clo
s
er
to
t
h
is
l
in
e
i
n
d
ica
te
h
i
g
h
er
ac
c
u
r
ac
y
.
T
h
ese
in
d
ices
w
er
e
s
elec
ted
b
ased
o
n
th
e
ap
p
r
o
ac
h
in
[
4
0
]
,
w
i
th
a
to
tal
o
f
1
0
5
d
ata
p
o
in
ts
ev
a
lu
ated
p
er
m
et
h
o
d
.
Fo
r
R
x
1
d
a
y
,
QM
p
r
o
v
id
ed
th
e
h
ig
h
est
n
u
m
b
er
o
f
ac
cu
r
ate
p
r
ed
ictio
n
s
(
2
7
p
o
in
ts
w
i
th
≤
1
0
%
er
r
o
r
)
,
f
o
llo
w
ed
b
y
I
DW
(
1
1
p
o
in
ts
)
an
d
R
FR
(
6
p
o
in
ts
)
.
T
h
is
f
in
d
in
g
s
u
p
p
o
r
ts
[
1
5
]
,
[
2
5
]
,
w
h
ic
h
e
m
p
h
as
izes
QM
’
s
e
f
f
ec
t
iv
e
n
es
s
in
ca
p
tu
r
i
n
g
d
ail
y
r
ai
n
f
a
ll
ex
t
r
e
m
es
b
y
ali
g
n
i
n
g
p
r
ed
icted
a
n
d
o
b
s
er
v
ed
q
u
an
tiles
.
I
DW
also
s
h
o
w
ed
s
tr
o
n
g
r
esu
lt
s
,
r
ein
f
o
r
cin
g
i
ts
r
eliab
ili
t
y
f
o
r
esti
m
at
in
g
ex
tr
e
m
e
d
ail
y
r
ain
f
all
b
ased
o
n
s
p
atial
i
n
te
r
p
o
latio
n
[
1
0
]
.
Fig
u
r
e
5
.
Scatter
p
lo
t o
f
o
b
s
er
v
atio
n
a
n
d
p
r
ed
ictio
n
r
esu
lt
s
f
o
r
R
x
1
d
a
y
,
R
9
5
p
T
O
T
,
R
2
0
m
m
,
an
d
C
W
D
Fo
r
P
R
C
PT
O
T
,
m
o
s
t
m
et
h
o
d
s
p
er
f
o
r
m
ed
b
etter
th
an
f
o
r
th
e
o
th
er
in
d
ices.
T
h
e
A
VG
m
et
h
o
d
r
ec
o
r
d
ed
th
e
h
ig
h
es
t
n
u
m
b
er
o
f
ac
c
u
r
ate
p
r
ed
ictio
n
s
(
4
6
p
o
in
ts
)
,
f
o
llo
w
ed
b
y
W
R
(
3
4
p
o
in
ts
)
,
QM
(
1
9
p
o
in
ts
)
,
an
d
I
DW
(
2
8
p
o
in
ts
)
.
C
u
m
u
lat
iv
e
i
n
d
ice
s
s
u
ch
a
s
P
R
C
P
T
O
T
ten
d
to
b
e
less
s
e
n
s
iti
v
e
to
d
ail
y
f
l
u
ctu
a
tio
n
s
,
m
ak
in
g
t
h
e
m
m
o
r
e
p
r
ed
ictab
le,
as
co
n
f
ir
m
e
d
b
y
[
2
3
]
,
[
3
5
]
.
Fo
r
C
W
D,
th
e
to
p
-
p
er
f
o
r
m
in
g
m
et
h
o
d
s
w
er
e
QM
(
1
9
p
o
in
ts
)
an
d
I
DW
(
1
4
p
o
in
ts
)
.
QM
w
as
t
h
e
m
o
s
t
co
n
s
is
te
n
t m
et
h
o
d
ac
r
o
s
s
in
te
n
s
it
y
,
f
r
eq
u
e
n
c
y
,
an
d
d
u
r
a
tio
n
-
b
ased
i
n
d
ices.
Ho
w
e
v
er
,
I
DW
r
em
a
in
ed
co
m
p
e
titi
v
e,
p
ar
ticu
lar
l
y
f
o
r
c
u
m
u
lati
v
e
an
d
d
u
r
atio
n
in
d
ice
s
lik
e
P
R
C
P
T
OT
an
d
C
W
D.
I
ts
s
tr
en
g
t
h
i
s
p
r
eser
v
i
n
g
d
ail
y
s
eq
u
e
n
ce
p
atter
n
s
,
e
s
s
e
n
tial
f
o
r
d
u
r
atio
n
an
a
l
y
s
is
[
1
3
]
,
an
d
o
f
f
er
in
g
r
eliab
le
p
er
f
o
r
m
a
n
ce
ev
e
n
w
it
h
s
p
ar
s
e
d
ata,
as
h
ig
h
li
g
h
ted
in
[
4
6
]
.
C
o
m
p
ar
ed
to
m
ac
h
in
e
lear
n
in
g
m
et
h
o
d
s
,
I
DW
o
f
f
er
s
p
r
ac
tical
ad
v
a
n
t
ag
es
d
u
e
to
it
s
m
o
r
e
s
tr
ai
g
h
t
f
o
r
w
ar
d
i
m
p
le
m
en
tat
io
n
a
n
d
lo
w
er
r
eso
u
r
ce
r
eq
u
ir
e
m
en
ts
,
as d
is
c
u
s
s
ed
in
[
3
6
]
,
[
4
0
]
.
Fig
u
r
e
6
p
r
esen
ts
th
e
to
tal
n
o
r
m
alize
d
s
co
r
es
(
r
an
g
i
n
g
f
r
o
m
0
to
1
)
b
ased
o
n
f
o
u
r
er
r
o
r
m
et
r
ics
R
MSE
,
B
ias,
P
ea
r
s
o
n
co
r
r
ela
tio
n
co
ef
f
icie
n
t
(
R
)
,
an
d
Klin
g
–
G
u
p
ta
ef
f
icien
c
y
(
K
GE
)
to
ev
alu
ate
th
e
p
er
f
o
r
m
an
ce
o
f
v
ar
io
u
s
b
ias
co
r
r
ec
tio
n
m
et
h
o
d
s
o
n
C
HI
R
P
S
d
ata
f
o
r
1
2
E
T
C
C
DI
ex
tr
e
m
e
cl
i
m
a
te
in
d
ice
s
.
T
h
e
d
ataset
in
cl
u
d
es
n
in
e
b
ias
co
r
r
ec
tio
n
m
et
h
o
d
s
(
s
tatis
tica
l
an
d
m
ac
h
i
n
e
lear
n
i
n
g
)
a
n
d
o
n
e
s
p
atial
i
n
ter
p
o
lati
o
n
m
e
th
o
d
(
I
DW
)
.
R
ed
s
h
ad
es
in
d
icate
h
i
g
h
p
er
f
o
r
m
an
ce
(
s
co
r
es
n
ea
r
1
)
,
w
h
ile
b
lu
e
in
d
icate
s
p
o
o
r
p
er
f
o
r
m
a
n
ce
(
n
ea
r
0
)
.
QM
an
d
I
DW
co
n
s
is
ten
tl
y
s
h
o
w
h
ig
h
p
er
f
o
r
m
an
ce
ac
r
o
s
s
m
o
s
t
in
d
ices,
p
ar
ticu
lar
l
y
in
in
te
n
s
i
t
y
-
b
a
s
ed
ca
teg
o
r
ies
(
R
x
1
d
a
y
,
R
x
5
d
a
y
,
P
R
C
P
T
OT
,
R
9
5
p
T
O
T
)
.
T
h
ese
f
in
d
in
g
s
r
ei
n
f
o
r
ce
p
r
ev
io
u
s
s
t
u
d
ies
th
at
co
n
f
ir
m
QM
’
s
ab
ilit
y
to
r
etain
q
u
a
n
tile
d
is
tr
ib
u
tio
n
s
an
d
ex
tr
e
m
e
v
a
lu
e
s
[
1
5
]
,
[
2
5
]
,
[
3
5
]
,
w
h
ile
I
DW
m
ai
n
tai
n
s
s
tab
le
p
er
f
o
r
m
an
ce
d
esp
ite
its
s
i
m
p
licit
y
an
d
lac
k
o
f
co
m
p
lex
s
tatis
t
ical
m
o
d
eli
n
g
[
1
3
]
,
[
4
6
]
.
C
o
n
v
er
s
el
y
,
R
F
R
an
d
I
DW
ex
h
ib
it
lo
w
p
er
f
o
r
m
an
ce
o
n
v
er
y
h
ig
h
r
ain
f
all
f
r
eq
u
e
n
c
y
i
n
d
ices
(
R
1
0
0
m
m
a
n
d
R
1
5
0
m
m
)
,
lik
e
l
y
d
u
e
to
t
h
e
li
m
ited
tr
a
in
i
n
g
d
ata
i
n
t
h
o
s
e
r
an
g
es
[
3
7
]
.
T
h
is
li
m
ita
tio
n
i
n
R
FR
h
a
s
also
b
ee
n
n
o
ted
in
o
th
er
s
tu
d
ies th
a
t r
ec
o
m
m
en
d
h
y
b
r
id
o
r
ex
tr
e
m
e
-
s
p
ec
i
f
ic
co
r
r
ec
tio
n
m
et
h
o
d
s
[
3
6
]
.
I
DW
,
h
o
w
ev
er
,
p
er
f
o
r
m
s
b
etter
t
h
a
n
R
F
R
i
n
t
h
e
S
DI
I
in
te
n
s
it
y
ca
teg
o
r
y
an
d
d
u
r
atio
n
i
n
d
ices
lik
e
C
D
D
a
n
d
C
W
D,
r
ea
f
f
ir
m
i
n
g
it
s
p
r
ac
tical
ad
v
an
ta
g
e
s
e
v
en
o
v
er
m
ac
h
in
e
lear
n
i
n
g
ap
p
r
o
ac
h
es.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Ma
ch
in
e
lea
r
n
in
g
-
b
a
s
ed
r
ec
o
n
s
tr
u
ctio
n
o
f m
is
s
in
g
r
a
in
fa
ll e
xtremes …
(
Ya
n
u
a
r
Hen
r
y
P
r
ib
a
d
i
)
1575
Fig
u
r
e
6
.
No
r
m
ali
s
ed
to
tal
s
co
r
e
(
R
MSE
,
R
,
B
ias,
KGE
)
p
er
m
et
h
o
d
f
o
r
th
e
E
T
C
C
DI
in
d
ex
T
o
a
i
d
i
n
t
e
r
p
r
e
t
at
i
o
n
,
th
e
n
o
r
m
al
i
z
e
d
s
c
o
r
e
s
w
e
r
e
c
at
eg
o
r
i
z
e
d
q
u
a
li
t
a
tiv
e
ly
u
s
in
g
q
u
a
r
t
i
l
e
th
r
e
s
h
o
l
d
s
f
r
o
m
a
l
l
c
o
m
b
in
a
t
i
o
n
s
o
f
m
eth
o
d
s
an
d
i
n
d
i
c
es
:
q
u
a
r
ti
l
e
1
(
Q
1
)
=
0
.
2
5
,
m
e
d
i
an
=
0
.
5
9
,
a
n
d
q
u
a
r
ti
l
e
3
(
Q
3
)
=
0
.
8
0
.
B
a
s
e
d
o
n
th
es
e
,
th
e
c
a
t
eg
o
r
i
es
a
r
e
d
ef
in
e
d
as
l
o
w
(
<
0
.
3
1
)
,
f
ai
r
(
0
.
3
1
–
0
.
6
0
)
,
h
ig
h
(
0
.
6
1
–
0
.
7
9
)
,
a
n
d
v
er
y
h
ig
h
(
≥
0
.
8
0
)
.
QM
co
n
s
is
ten
tl
y
d
o
m
i
n
ated
th
e
h
i
g
h
a
n
d
v
er
y
h
i
g
h
ca
teg
o
r
ies,
w
it
h
o
n
l
y
R
1
0
0
m
m
a
n
d
R
1
5
0
m
m
f
al
lin
g
in
to
th
e
f
air
ca
teg
o
r
y
.
I
DW
f
o
llo
w
ed
clo
s
el
y
w
it
h
s
tr
o
n
g
p
er
f
o
r
m
a
n
ce
in
m
o
s
t
i
n
te
n
s
it
y
an
d
d
u
r
atio
n
in
d
ices,
o
n
l
y
s
co
r
in
g
lo
w
in
R
1
5
0
m
m
a
n
d
f
air
in
SDI
I
,
R
2
0
m
m
,
R
5
0
m
m
,
an
d
R
1
0
0
m
m
.
Me
an
w
h
ile
,
m
ac
h
i
n
e
lear
n
i
n
g
-
b
ased
R
FR
p
er
f
o
r
m
ed
b
est i
n
P
R
C
P
T
O
T
,
an
an
n
u
al
c
u
m
u
la
t
iv
e
r
ain
f
all
i
n
d
ex
[
4
4
]
.
Oth
er
m
et
h
o
d
s
lik
e
I
F a
n
d
R
R
m
o
s
tl
y
f
ell
i
n
th
e
lo
w
-
to
-
f
air
r
an
g
e,
w
it
h
I
F p
ar
ticu
l
ar
l
y
cr
iticized
f
o
r
o
v
er
-
s
u
p
p
r
ess
i
n
g
o
u
tlier
s
an
d
f
aili
n
g
to
r
ep
r
esen
t
ex
tr
e
m
es
[
3
9
]
.
I
D
W
o
u
tp
er
f
o
r
m
ed
s
e
v
er
al
s
ta
tis
t
ical
an
d
m
ac
h
in
e
lear
n
i
n
g
ap
p
r
o
ac
h
es
i
n
s
p
atiall
y
an
d
te
m
p
o
r
all
y
v
ar
iab
le
e
n
v
ir
o
n
m
e
n
ts
d
esp
ite
n
o
t
b
ei
n
g
e
x
p
licitl
y
d
esi
g
n
ed
f
o
r
b
ias
co
r
r
ec
tio
n
.
T
h
ese
r
esu
lt
s
co
n
f
ir
m
I
DW
’
s
v
iab
ilit
y
i
n
d
a
ta
-
s
ca
r
ce
co
n
te
x
ts
,
al
th
o
u
g
h
a
ll
f
in
d
i
n
g
s
s
h
o
w
n
i
n
Fig
u
r
e
6
r
e
m
ai
n
q
u
a
litati
v
e
an
d
s
h
o
u
ld
b
e
in
ter
p
r
eted
ca
u
t
io
u
s
l
y
,
as
th
e
y
ar
e
n
o
t
d
ef
i
n
iti
v
e
f
o
r
ch
o
o
s
in
g
a
u
n
iv
er
s
a
ll
y
b
est
m
eth
o
d
in
t
h
i
s
s
tu
d
y
.
4.
CO
NCLU
SI
O
N
T
h
is
r
esear
ch
p
r
esen
ts
a
co
m
p
r
eh
en
s
i
v
e
ap
p
r
o
ac
h
to
r
ec
o
n
s
tr
u
cti
n
g
d
ail
y
r
ain
f
all
d
ata
in
tr
o
p
ical
r
eg
io
n
s
w
it
h
li
m
ited
o
b
s
er
v
ati
o
n
al
d
ata
t
h
r
o
u
g
h
a
n
i
n
te
g
r
ati
v
e
e
v
alu
a
tio
n
o
f
s
tatis
t
ical
a
n
d
m
ac
h
in
e
lear
n
i
n
g
-
b
ased
b
ias
co
r
r
ec
tio
n
m
et
h
o
d
s
an
d
s
p
atial
i
n
ter
p
o
latio
n
.
O
f
t
h
e
th
r
ee
i
n
ter
p
o
latio
n
m
et
h
o
d
s
,
I
DW
w
as
th
e
m
o
s
t
ac
cu
r
ate
a
n
d
s
tab
le
f
o
r
esti
m
atin
g
d
ail
y
r
ai
n
f
al
l
v
alu
e
s
,
b
a
s
ed
o
n
s
tatis
tical
m
etr
ic
e
v
al
u
atio
n
s
o
f
s
i
m
u
la
ted
m
is
s
i
n
g
o
b
s
er
v
atio
n
a
l
d
ata.
R
eg
ar
d
in
g
s
atell
ite
d
ata
b
ias
co
r
r
ec
tio
n
,
th
e
C
HI
R
P
S
p
r
o
d
u
ct
ap
p
ea
r
ed
m
o
r
e
co
n
s
is
ten
t
t
h
an
MSW
E
P
an
d
GP
M
I
ME
R
G
.
QM
ac
h
iev
ed
th
e
h
ig
h
est
p
r
ed
ictiv
e
ac
cu
r
ac
y
a
m
o
n
g
th
e
s
i
x
s
tatis
t
ical
b
ias
co
r
r
ec
tio
n
m
et
h
o
d
s
.
R
FR
p
er
f
o
r
m
ed
b
est
am
o
n
g
th
e
th
r
ee
m
ac
h
i
n
e
-
lear
n
in
g
-
b
ased
m
et
h
o
d
s
.
Fu
r
t
h
er
test
i
n
g
u
s
in
g
t
h
e
ex
tr
e
m
e
cli
m
ate
i
n
d
ex
E
T
C
C
DI
r
ein
f
o
r
ce
s
t
h
ese
f
i
n
d
in
g
s
.
C
HI
R
P
S
-
QM
p
r
o
v
id
es t
h
e
m
o
s
t
r
ep
r
esen
tati
v
e
i
n
d
ex
r
esu
lts
ag
ai
n
s
t
o
b
s
er
v
at
io
n
al
d
ata,
s
u
p
er
io
r
to
th
e
m
ac
h
i
n
e
lear
n
i
n
g
-
b
ased
R
F
R
an
d
I
DW
in
ter
p
o
latio
n
.
Nev
er
th
el
ess
,
I
DW
is
s
till
b
etter
th
a
n
m
o
s
t
o
th
er
b
ias
co
r
r
ec
tio
n
m
eth
o
d
s
.
T
h
u
s
,
th
i
s
r
esear
ch
co
n
cl
u
d
es
t
h
at
i
n
teg
r
atin
g
C
HI
R
P
S
-
QM
a
n
d
I
DW
is
a
h
i
g
h
l
y
e
f
f
ec
ti
v
e
s
tr
ate
g
y
f
o
r
r
ec
o
n
s
tr
u
ct
in
g
d
ail
y
r
ain
f
al
l
d
ata,
esp
ec
iall
y
f
o
r
ex
tr
e
m
e
e
v
en
t
a
n
al
y
s
is
n
e
ed
s
,
co
m
p
ar
ed
to
o
th
er
m
eth
o
d
s
p
r
o
d
u
ce
d
in
t
h
i
s
r
esear
ch
.
C
o
n
s
id
er
in
g
all
ev
a
lu
atio
n
r
es
u
lt
s
co
m
p
ar
ed
to
o
th
er
m
et
h
o
d
s
i
n
t
h
is
s
tu
d
y
,
t
h
e
co
m
b
in
a
tio
n
o
f
C
HI
R
P
S
-
QM
an
d
I
DW
p
r
o
d
u
ce
s
th
e
b
est
r
es
u
lts
i
n
r
ec
o
n
s
tr
u
cti
n
g
d
ail
y
r
ain
f
all
d
ata
w
it
h
li
m
ited
d
ata.
T
h
is
ap
p
r
o
ac
h
ex
ce
ls
in
s
tatis
tical
m
etr
ics
an
d
m
ain
ta
in
i
n
g
th
e
p
r
esen
ce
o
f
ex
tr
e
m
e
v
al
u
es
th
a
t
ar
e
im
p
o
r
tan
t
f
o
r
d
is
aster
r
is
k
a
n
al
y
s
is
an
d
cli
m
ate
ch
a
n
g
e
in
th
e
B
an
te
n
an
d
J
ak
ar
ta
r
eg
io
n
s
.
Ho
w
ev
er
,
u
s
i
n
g
i
m
p
u
ted
d
ata
f
o
r
ex
tr
e
m
e
r
ain
f
all
an
a
l
y
s
is
s
h
o
u
ld
b
e
ca
u
tio
u
s
l
y
ap
p
r
o
ac
h
ed
,
a
s
m
o
s
t
m
et
h
o
d
s
s
h
o
w
lo
w
ac
c
u
r
ac
y
in
esti
m
ati
n
g
th
e
f
r
eq
u
e
n
c
y
o
f
r
ain
f
al
l e
v
e
n
t
s
g
r
ea
ter
th
a
n
1
5
0
m
m
.
ACK
NO
WL
E
D
G
M
E
NT
S
T
h
e
au
th
o
r
s
w
o
u
ld
lik
e
to
t
h
an
k
th
e
Me
teo
r
o
lo
g
y
,
C
l
i
m
a
to
lo
g
y
,
a
n
d
Geo
p
h
y
s
ics
Ag
e
n
c
y
(
B
MK
G)
f
o
r
its
f
i
n
a
n
cial
s
u
p
p
o
r
t
an
d
p
r
o
v
is
io
n
o
f
r
ain
f
all
d
ata.
A
p
p
r
ec
iatio
n
is
also
e
x
ten
d
ed
to
th
e
Dep
ar
t
m
e
n
t
o
f
Evaluation Warning : The document was created with Spire.PDF for Python.