Page 1 Page 1 The EUMETSAT Satellite Applications Facility for NWP (NWP SAF) Bill Bell Bryan Conway and John Eyre Met Office, UK JCSDA presentation 17 th April 2007
Page
1Pa
ge 1
The
EU
ME
TSA
TS
atel
lite
App
licat
ions
Fac
ility
for N
WP
(NW
P S
AF)
Bill
Bel
l
Bry
an C
onw
ay a
nd J
ohn
Eyr
e
Met
Offi
ce, U
K
JCSD
A pr
esen
tatio
n 17
thAp
ril 2
007
Page
2Pa
ge 2
Out
line
oB
ackg
roun
d / M
issi
on /
Peo
ple
oP
roje
cts:
•A
AP
P•
RTT
OV
•1D
-Var
•Q
uick
scat
Dat
a P
roce
ssor
/ SD
P•
Mon
itorin
g re
ports
•S
SM
IS_P
repr
oces
sor
•S
uppo
rt to
EA
RS
oC
oncl
usio
ns
Page
3Pa
ge 3
NW
PS
AF
back
grou
nd
Oth
er S
AFs
incl
ude:
oce
an &
sea
ice,
clim
ate,
ozo
ne
and
Gra
sN
WP
SA
F de
liver
s so
ftwar
e m
odul
es fo
r use
in D
A
syst
ems
( oth
er S
AF’
sde
velo
p pr
oduc
ts)
Pha
ses
incl
ude
:5
year
dev
pha
se (1
998
-20
03)
3 ye
ar in
itial
ope
ratio
nal p
hase
(200
3 -2
006)
cont
inuo
us d
ev a
nd o
pera
tions
pha
se (2
007
-)M
anag
emen
t : S
teer
ing
grou
pS
oftw
are
deve
lopm
ent u
nder
qua
lity
syst
ems
(ded
icat
ed Q
A/Q
C a
t loc
al le
vel,
revi
ew p
roce
ss)
Vis
iting
sci
entis
ts (t
o ac
cele
rate
the
deve
lopm
ent o
f de
liver
able
s / t
rain
ing
1 da
y -2
yea
rs)
Page
4Pa
ge 4
The
NW
P S
AF:
bac
kgro
und
•Sat
ellit
e A
pplic
atio
ns F
acili
ty fo
r Num
eric
al
Wea
ther
Pre
dict
ion
(NW
P SA
F)
•one
of 8
SA
Fsth
at fo
rm p
art o
f the
dis
tribu
ted
grou
nd s
egm
ent o
f EU
ME
TSA
T
•Led
by
the
Met
Offi
ce, i
n pa
rtner
ship
with
EC
MW
F, K
NM
I and
Mét
éo-F
ranc
e
•75%
-fund
ed b
y EU
MET
SAT
Page
5Pa
ge 5
The
NW
P S
AF:
mis
sion
MIS
SIO
N
•To
impr
ove
and
supp
ort t
he in
terfa
ce b
etw
een
sate
llite
dat
a/pr
oduc
ts a
nd E
urop
ean
activ
ities
in
glo
bal a
nd re
gion
al N
WP
Page
6Pa
ge 6
The
NW
P S
AF:
peo
ple
•M
anag
er:B
.Con
way
•Pr
ojec
t Tea
m:
•M
et O
ffice
:S.E
nglis
h, R
.Sau
nder
s, D
.Offi
ler,
N.A
tkin
son,
W
.Bel
l, J.
Cam
eron
, B.C
andy
, A.D
oher
ty, M
.For
syth
e,
P.F
ranc
is, R
.Fra
ncis
, F.H
ilton
, S.K
eogh
, U.O
’Kee
ffe, E
.Pav
elin
, P
.Ray
er, S
.Wat
kin
•EC
MW
F:T.
McN
ally
, P.B
auer
, A.C
olla
rd, A
.Gar
cia-
Men
dez,
H
.Her
sbac
h, G
.Kel
ly, J
.-N.T
hépa
ut, G
.Van
derG
rijn
•M
étéo
-Fra
nce:
P.B
rune
l, T.
Labr
ot, L
.Lav
anan
t, P
.Mar
guin
aud,
A
.Mar
soui
n•
KN
MI:
A.S
toffe
len,
A.V
erho
ef, J
.Vog
elez
ang
•St
eerin
g G
roup
:J.E
yre,
L.S
arlo
, S.E
lliott,
J.O
nvle
e, P
.Pyl
kko,
F.
Rab
ier,
P.S
chlu
esse
l, A
.Sim
mon
s•
Visi
ting
scie
ntis
ts: M
any!
Page
7Pa
ge 7
The
NW
P S
AF:
pro
duct
s
At p
rese
nt:
•A
APP
-ATO
VS
and
AV
HR
R P
re-p
roce
ssin
g P
acka
ge•
RTT
OV
-fas
t rad
iativ
e tra
nsfe
r mod
el•
+ m
odel
-bas
ed p
rofil
e da
ta s
ets
•1D
-Var
retri
eval
sch
emes
•Q
DP
-Qui
ksca
tDat
a P
roce
ssor
•M
onito
ring
repo
rts
•SS
MIS
pre
-pro
cess
or
Und
er d
evel
opm
ent:
•U
pdat
esto
the
abov
e•
SDP
-Sca
ttero
met
erD
ata
Pro
cess
or
Page
8Pa
ge 8
The
NW
P S
AF:
AA
PP AA
PP
Page
9Pa
ge 9
The
NW
P S
AF:
AA
PP
Page
10
Page
10
The
NW
P S
AF:
AA
PP
The
ATO
VS
and
AV
HR
R P
re-p
roce
ssin
g P
acka
ge
•P
erfo
rms
ING
EST
and
PRE-
PRO
CES
SIN
Gof
ATO
VS
and
AV
HR
R
data •A
TOVS
= A
MS
U-A
+ H
IRS
+ A
MS
U-B
/MH
S•
ING
EST
•D
ecom
mut
atio
n•
Nav
igat
ion
•C
alib
ratio
n•
PRE-
PRO
CES
SIN
G•
Pre
cipi
tatio
n an
d cl
oud
dete
ctio
n fo
r mic
row
ave
inst
rum
ents
•M
appi
ng to
com
mon
grid
(e.g
. HIR
S fi
elds
of v
iew
)•
Clo
ud a
naly
sis
on A
VH
RR
pix
els
with
in e
ach
HIR
S fo
v
Page
11
Page
11
AA
PP
ver
sion
s 1
to 5
HR
PT
(NO
AA
sat
ellit
es)
ATO
VS
1a/
b
ATO
VS
1c
AV
HR
R 1
a/b
NE
SD
IS
Glo
bal A
TOV
S
EU
ME
TCas
tR
egio
nal A
TOV
S
(EA
RS
)
ATO
VS
1dS-b
and
1a =
cou
nts
1b =
cou
nts
+ ca
libra
tion
+ ge
oloc
atio
n1c
= ra
dian
ces
1d =
radi
ance
s m
appe
d to
com
mon
grid
e.g
. HIR
S
AA
PP
Page
12
Page
12
AA
PP
ver
sion
6 –
incl
udin
g M
ETO
P
ME
TOP
AH
RP
T(C
CS
DS
pac
kets
)
ATO
VS
+ A
VHR
RIA
SI ‘
leve
l 0’
+ A
VHR
R
ATO
VS
1a/
b
ATO
VS
1c
IAS
I 1c
BU
FR d
ecod
e
EU
ME
TCas
tG
loba
l IA
SI
EU
ME
TCas
tG
loba
l ATO
VS
+ E
AR
S
ATO
VS
/IAS
I 1d
ATO
VS
1d
BU
FR d
ecod
e
Exi
stin
g A
AP
P
OP
S
NB
: IA
SI 1
c =
calib
rate
d,
corr
ecte
d &
Gau
ssia
nap
odis
ed
Use
r opt
ions
(fle
xibl
e):
•AM
SU m
appe
d to
IASI
•Spa
tial t
hinn
ing
•Red
uced
cha
nnel
set
and
/or
•Eig
enve
ctor
sco
res
or G
TN
X-b
and
IAS
IP
re-p
roce
ssor
Page
13
Page
13
The
NW
P S
AF:
RTT
OV RTT
OV
Page
14
Page
14
The
NW
P S
AF:
RTT
OV
RTT
OV
-a
fast
radi
ativ
e tra
nsfe
r mod
el
•It
is u
sed
by N
WP
cen
tres
for s
ever
al a
pplic
atio
ns (e
.g. r
adia
nce
assi
mila
tion,
dat
a m
onito
ring,
sim
ulat
ed im
ager
y)
•Th
e N
WP
SA
F m
aint
ains
and
dis
tribu
tes
rece
nt v
ersi
ons
of R
TTO
V
(cur
rent
ly v
ersi
ons
7 &
8)
•Th
e la
test
ver
sion
, RTT
OV
v8,
was
rele
ased
in N
ovem
ber 2
004
•Th
e ne
xt v
ersi
on, R
TTO
V v
9, is
now
und
er d
evel
opm
ent a
nd w
ill b
ere
leas
ed in
Feb
200
7
Page
15
Page
15
RTT
OV
func
tiona
lity
©C
row
n co
pyrig
ht 2
005
05/
0178
M
et O
ffice
and
the
Met
Offi
ce lo
go a
re re
gist
ered
trad
emar
ks
•In
put p
rofil
es: t
empe
ratu
re a
nd w
ater
vap
our;
optio
nally
, ozo
ne a
nd
carb
on d
ioxi
de a
lso
as v
aria
ble
gase
s•
Com
pute
sto
p-of
-atm
osph
ere
radi
ance
s, b
right
ness
tem
pera
ture
s an
d la
yer-t
o-sp
ace
trans
mitt
ance
s fo
r eac
h ch
anne
l•
Com
pris
es:
forw
ard,
tang
ent l
inea
r, ad
join
tand
K (J
acob
ian
mat
rix)
mod
els,
for u
se in
var
iatio
nala
ssim
ilatio
n or
retri
eval
s•
Sea-
surf
ace
emis
sivi
ty: c
ompu
ted
inte
rnal
ly (I
SE
M-6
mod
el fo
r IR
, FA
STE
M fo
r MW
) or v
alue
pro
vide
d by
use
r •
Clo
uds:
•S
ingl
e-la
yer,
spec
trally
-inva
riant
•M
ulti-
laye
r, sp
ectra
lly-v
aryi
ng: u
sing
“wra
pper
”cod
e, R
TTO
V_C
LD•
Mic
row
ave,
with
sca
tterin
g: u
sing
“wra
pper
”cod
e, R
TTO
V_S
CA
TT•
Fort
ran-
90•
Run
und
er u
nix
or li
nux;
test
ed o
n ra
nge
of p
latfo
rms
•R
un-ti
me:
~ 0
.5 m
sfo
r 20
HIR
S c
hann
els
for 1
pro
file
on H
P w
orks
tatio
n
Page
16
Page
16
©C
row
n co
pyrig
ht 2
005
05/
0178
M
et O
ffice
and
the
Met
Offi
ce lo
go a
re re
gist
ered
trad
emar
ks
1-9
TMI
TRM
M1-
10W
indS
atC
orio
lis
1-17
, 1-2
378
1-15
, 1-4
,1-1
4M
OD
IS,A
IRS
AM
SU-A
, HSB
, AM
SRTe
rra
Aqu
a
FY-1
, FY-
2
GM
S-5,
MTS
AT
ERS-
1/2
ENVI
SAT
GO
ES-8
-12
Met
eosa
t-2-8
DM
SP F
-16
DM
SP F
-8-1
5
TIR
OS-
NN
OA
A-6
-18
NO
AA
-2-5Pl
atfo
rms
1-3,
1-2
MVI
SR, V
ISSR
1-3,
1-4
Imag
er
1-3
1-3
ATS
RA
ATS
R
1-4
1-18
Imag
erSo
unde
r
2 4-11
MVI
RI
SEVI
RI
1-24
SSM
I(S)
1-7
SSM
/I
1-19
, 1-4
1-3,
1-1
51-
5, 1
-5,
1-3,
1-8
HIR
S, M
SU,
SSU
, AM
SU-A
AM
SU-B
, MH
S,A
VHR
R, V
TPR
Cha
nnel
ssi
mul
ated
Sens
or
RTT
OV
–se
nsor
s su
ppor
ted
Page
17
Page
17
RTT
OV
-85/
87 s
tatu
s
Num
ber o
f lic
ence
requ
ests
= 1
87N
umbe
r of u
sers
pro
vide
d co
de ~
180
Num
ber o
f bug
s re
port
ed s
ince
rele
ase
= 10
for 8
5 co
rrec
ted
in 8
7 an
d 6
for 8
7 al
l min
or
Effic
ient
vec
toris
atio
nof
cod
e st
ill b
eing
wor
ked
on
for N
EC s
uper
com
pute
r R
ewrit
ten
RTT
OV_
SCA
TT c
ode
for R
TTO
V-87
Part
icip
ated
in A
IRS
RT
com
paris
on (s
ee s
epar
ate
talk
)
Page
18
Page
18
New
coe
ffici
ent f
iles
avai
labl
e
ME
TOP
n.b
. sat
ellit
e id
=2IA
SI a
vaila
ble
for R
TTO
V-7
and
8H
IRS
ava
ilabl
e A
MS
U-A
/MH
S a
vaila
ble
AV
HR
R a
vaila
ble
MS
G-2
SE
VIR
I ava
ilabl
e
GO
ES
-12
soun
der a
vaila
ble
Meg
aTro
piqu
es, S
aphi
rand
Mad
ras
Page
19
Page
19
RTT
OV
-9 d
evel
opm
ents
New
div
erse
pro
file
data
set –
incl
usio
n of
mor
e m
inor
ga
ses,
mor
e le
vels
Incl
usio
n of
mul
tiple
sca
tterin
g fo
r clo
udy
and
aero
sol
radi
ance
cal
cula
tions
Line
ar in
tau
mea
n pa
th v
alue
sZe
nith
ang
le d
epen
denc
e of
pat
hIn
clud
e re
flect
ed s
olar
for S
WIR
Mor
e ac
tive
trac
e ga
ses
CO
, CH
4, N
2O,
Furt
her o
ptim
isat
ion
of p
redi
ctor
sIm
prov
emen
ts to
RTT
OV_
SCA
TT (n
ew M
ie ta
bles
)C
hang
e in
terf
ace
to a
llow
pro
file
inpu
t on
user
leve
lsC
hang
e in
terf
ace
to a
void
nee
d to
spe
cify
pol
aris
atio
n in
dex
Wha
t is i
nclu
ded?
Page
20
Page
20
The
NW
P S
AF:
1D
-Var
sch
emes
1D-V
AR
Page
21
Page
21
The
NW
P S
AF:
1D
-Var
sch
emes
One
-dim
ensi
onal
var
iatio
nalr
etrie
val –
1D-V
ar
Use
d fo
r:
•R
etrie
val o
f atm
osph
eric
/sur
face
var
iabl
es fr
om ra
dian
ce
mea
sure
men
ts
•N
WP
dat
a as
sim
ilatio
n –
pre-
proc
essi
ng a
nd q
ualit
y co
ntro
l
•R
esea
rch
tool
–ra
pid
expl
orat
ion
of n
ew d
ata
Page
22
Page
22
Var
iatio
nald
ata
assi
mila
tion
Ass
imila
ting
obse
rvat
ions
into
a N
WP
m
odel
Min
imiz
e:
J(x)
=
½(x
-xb )
TB
-1 (x
-xb )
+
½(y
o -H
[x])T
(E+F
)-1 (y
o -H
[x])
whe
rex
cont
ains
the
NW
P m
odel
sta
te
xbis
bac
kgro
und
estim
ate
of x
(sho
rt-ra
nge
fore
cast
)B
is it
s er
ror c
ovar
ianc
e,
yois
vec
tor o
f mea
sure
men
tsH
[...]
is “o
bser
vatio
n op
erat
or ”
or “f
orw
ard
mod
el”,
map
ping
sta
te x
into
“mea
sure
men
t spa
ce”
E is
erro
r cov
aria
nce
of m
easu
rem
ents
,F
is e
rror c
ovar
ianc
e of
forw
ard
mod
el.
∇xJ
(x)T
= B
-1 (x
-xb )
-∇
xH[x
]T(E
+F)-1
(yo -
H[x
]) =
0
Page
23
Page
23
1D-V
ar: s
imul
ated
IAS
I ret
rieva
l err
ors
back
grou
nd e
rror
retri
eval
err
ors
bias
SD
973
sim
ulat
ed re
triev
als
1000
IAS
I cha
nnel
s
mid
-latit
ude
win
ter p
rofil
e
Page
24
Page
24
The
NW
P S
AF:
1D
-Var
sch
emes
3 sc
hem
es a
re a
vaila
ble:
•“E
CM
WF”
•gen
eric
har
ness
–m
inim
isat
ion
sche
me
with
“hoo
ks”
•“M
et O
ffice
”•c
ompl
ete
sche
me
-ATO
VS
, AIR
S, I
AS
I
•“S
SM
IS”
•com
plet
e sc
hem
e -S
SM
I, S
SM
IS, A
MS
U
Page
25
Page
25
The
NW
P S
AF:
Sca
ttero
met
erpr
oces
sors
QD
P /
SD
P
Page
26
Page
26
The
NW
P S
AF:
Sca
ttero
met
erpr
oces
sors
•QD
P-Q
uiks
catD
ata
Pro
cess
or –
avai
labl
e N
OW
•In
put –
NO
AA
Qui
ksca
tpro
duct
in B
UFR
•P
re-p
roce
ssin
g –
sorti
ng a
nd s
patia
l ave
ragi
ng•
Win
d re
triev
al•
QC
–ra
in d
etec
tion,
etc
•A
mbi
guity
rem
oval
•M
onito
ring
and
outp
ut
•SD
P–
Sca
ttero
met
erD
ata
Pro
cess
or -
SOO
N•
Gen
eric
sca
ttero
met
erco
de•
ER
S S
CA
T, M
ETO
P A
SC
AT,
Sea
win
ds(Q
uiks
cat,
NS
CA
T)
Page
27
Page
27
Sca
ttero
met
er D
ata
Pro
cess
or
Obs
erva
tions
Inve
rsio
nA
mbi
guity
Rem
oval
Win
d Fi
eld
OU
TPU
TO
cean
Surf
ace
Rad
ar B
acks
catte
r O
bser
vatio
nsIn
vers
ion
Am
bigu
ityR
emov
al
Qua
lity
Con
trol
Pre-
Proc
ess
Win
d Fi
eld
INPU
TO
UT
PUT
Qua
lity
Mon
itor
Page
28
Page
28
NO
AA
S
DP
@ 2
5 km
impr
oved
col
dfr
ont be
tter
arou
ndra
in
50 k
mPl
ots !
Page
29
Page
29
The
NW
P S
AF:
Mon
itorin
g re
ports
Mon
itorin
g re
ports
Page
30
Page
30
The
NW
P S
AF:
mon
itorin
g pr
oduc
ts
•Obs
erva
tion
cove
rage
plo
ts•S
tatis
tics
of o
bser
ved-
fore
cast
diff
eren
ces
•Dat
a ty
pes:
•A
TOV
S, S
SM
I, A
IRS
, geo
-radi
ance
s•A
MV
s•Q
uiks
cat,
ER
S-2
•Ozo
ne: S
BU
V, E
nvis
at
Page
31
Page
31
The
NW
P S
AF:
mon
itorin
g pr
oduc
ts
Exa
mpl
e:
AM
Vs
v. 6
h fo
reca
st
Met
Offi
ce
EC
MW
F
Page
32
Page
32
The
NW
P S
AF:
mon
itorin
g pr
oduc
ts
Exa
mpl
e:
AM
Vs
v. 6
h fo
reca
st
Met
Offi
ce
Page
33
Page
33
The
NW
P S
AF:
mon
itorin
g pr
oduc
ts
Exa
mpl
e:
MS
G 6
.3µm
wat
er v
apou
rch
anne
l
EC
MW
F
Page
34
Page
34
The
NW
P S
AF:
mon
itorin
g pr
oduc
ts
Exa
mpl
e of
link
to
anot
her m
onito
ring
site
:
KN
MI Q
uiks
cat
mon
itorin
g
Page
35
Page
35
SS
MIS
Pre
proc
esso
r
Mai
n fu
nctio
ns:
•Rem
appi
ng
•Ave
ragi
ng•R
efle
ctor
em
issi
on c
orre
ctio
n•S
olar
intru
sion
flag
ging
Cod
e &
pre
proc
esse
d da
ta a
vaila
ble
Page
36
Page
36
SS
MIS
Pre
proc
esso
r: A
vera
ging
•N
E∆
T fo
r LA
S c
hann
els
is ~
0.3K
⇒re
quire
ave
ragi
ng to
ach
ieve
NE∆
T eff
= 0.
1K
•A
lso
bene
fit fr
om im
prov
ed s
cale
mat
chin
g?
•Ope
ratio
nal p
repr
oces
soru
ses
σ=
50km
(FW
HM
= 1
18km
)
•NE∆
T eff
~ 0.
03K
•Pro
cess
ing
time
~ 1
min
ute/
orb
it
Page
37
Page
37
SS
MIS
pre
proc
esso
r: fla
ggin
g an
d co
rrec
tion
resi
dual
bia
ses
Page
38
Page
38
SS
MIS
pre
proc
esso
r: fla
ggin
g an
d co
rrec
tion
re
sidu
al b
iase
s
Page
39
Page
39
The
NW
P S
AF:
sup
port
to E
AR
S
Sup
port
to E
AR
S
Page
40
Page
40
The
NW
P S
AF:
sup
port
to E
AR
S
NW
P SA
F su
ppor
t to
the
EUM
ETSA
T A
TOVS
Ret
rans
mis
sion
Ser
vice
(EA
RS)
:
•Dev
elop
men
t and
mai
nten
ance
of A
AP
P
•Rea
l-tim
e da
ta m
onito
ring:
•f
or e
ach
EA
RS
rece
ptio
n si
te•c
heck
s co
nsis
tenc
y w
ith g
loba
l ATO
VS
dat
a•c
heck
s co
nsis
tenc
y w
ith lo
cally
-rec
eive
d A
TOV
S d
ata
(Lan
nion
)
Page
41
Page
41
The
NW
P S
AF:
sup
port
to E
AR
S
Con
clus
ions
Page
42
Page
42
The
NW
P S
AF:
con
clud
ing
rem
arks
•C
olla
bora
tion
betw
een
4 E
urop
ean
NW
P c
entre
s, w
ith
supp
ort f
rom
EU
ME
TSA
T, h
as p
erm
itted
the
deve
lopm
ent,
deliv
ery
and
supp
ort o
f:•s
oftw
are
mod
ules
for s
atel
lite
data
pro
cess
ing
and
assi
mila
tion,
•dat
a m
onito
ring
serv
ices
,to
a la
rge
and
grow
ing
user
com
mun
ity.
•O
ver t
he n
ext f
ew y
ears
, the
NW
P S
AF
plan
s to
co
ntrib
ute
to th
e ex
ploi
tatio
n of
dat
a fro
m n
ew in
stru
men
ts.
•C
olla
bora
tion
with
the
inte
rnat
iona
l com
mun
ity w
ill b
e ne
eded
to e
nsur
e w
e co
ver a
ll th
e ne
w in
stru
men
ts, i
n a
timel
y m
anne
r, w
ithou
t unn
eces
sary
dup
licat
ion.
Page
43
Page
43
The
NW
P S
AF:
furth
er in
form
atio
n
•Fo
r inf
orm
atio
n, v
isit:
http
://w
ww
.met
offic
e.go
v.uk
/rese
arch
/inte
rpro
j/nw
psaf
/inde
x.ht
ml
•To
obt
ain
softw
are,
vis
it:
http
://w
ww
.met
offic
e.go
v.uk
/rese
arch
/inte
rpro
j/nw
psaf
/re
ques
t_fo
rms/
inde
x.ht
ml
Page
44
Page
44
End
Page
45
Page
45
Sup
plem
enta
ry s
lides
Page
46
Page
46
AA
PP
-R
ecen
t upd
ates
Upd
ates
to A
AP
P v
4.0
Upd
ate
4.4,
31/
8/04
:Li
nux
com
patib
ility
(als
o W
indo
ws
via
MS
Ser
vice
s Fo
r Uni
x)Im
prov
ed ro
bust
ness
in d
ecom
mut
atio
nU
tility
to c
ompa
re o
utpu
t file
s fro
m d
iffer
ent p
latfo
rms
(ato
vsC
ompa
re)
Big
/littl
e-en
dian
conv
ersi
ons
Pro
cess
ing
of N
OA
A-1
7to
leve
l 1d
1d fl
ag fo
r few
er c
ollo
catio
ns th
an e
xpec
ted
in re
-map
ping
A
MS
U-A
to H
IRS
Upd
ate
4.5,
03/
02/0
5:U
pdat
ed A
MS
U-B
cal
ibra
tion
para
met
ers
file
(gro
ss li
mits
)
Page
47
Page
47
AA
PP
ver
sion
5
Incl
udes
the
follo
win
g en
hanc
emen
ts:
•N
OA
A-N
cap
abili
ty (i
nclu
ding
MH
S)
•N
ew H
IRS
cal
ibra
tion
met
hod
(bas
ed o
n N
OA
A v
4)•
Upd
ated
nav
igat
ion
–ab
ility
to u
se 2
-line
ele
men
ts•
Cal
ibra
tion
-allo
w fo
r moo
n co
ntam
inat
ion
in A
MS
U-B
/MH
S•
Pre
cipi
tatio
n te
sts
-add
ed N
WC
-SA
F sc
atte
ring
inde
x (B
enna
rtz) t
o A
MS
U-B
leve
l 1d
•U
se o
f ins
trum
ent-s
peci
fic s
can
char
acte
ristic
s, a
nd re
mov
al o
f m
any
hard
-cod
ed p
aram
eter
s
Rel
ease
d to
use
rs o
n 18
Jul
y 20
05, f
ollo
win
g va
lidat
ion
with
NO
AA
-18
data
(~2
mon
ths
afte
r lau
nch)
Page
48
Page
48
The
NW
P S
AF:
1D
-Var
sch
emes A
ssim
ilatin
gob
serv
atio
ns in
to a
NW
P
mod
el
OR
Ret
rievi
ng a
tmos
pher
ic
varia
bles
usi
ng a
fo
reca
st p
rofil
e as
ba
ckgr
ound
(firs
t gue
ss)
Page
49
Page
49
The
NW
P S
AF:
mon
itorin
g E
AR
S v
glo
bal r
adia
nces
Exa
mpl
e: A
TOV
S -
Trom
so