L-band sea su ace emissi i y: P elimina y esul s o he
WISE-2000 campaign and i s applica ion o salini y e ie al
in he SMOS mission
A. Camps,1I. Co bella,1M. Vall-llosse a,1N. Du o,1F. To es,1R. Villa ino,1
L. En ique,1J. Mi anda,1F. Julbe´,1J. Fon ,2A. Julia`,2C. Gaba o´,2J.
E che o,3J. Bou in,3A. Weill,4V. Caselles,5E. Rubio,5P. Wu s eisen,6M.
Be ge ,6and M. Ma ı´n-Nei a6
Recei ed 1 Ma ch 2002; e ised 30 May 2002; accep ed 4 June 2002; published 19 June 2003.
[1] Soil mois u e and ocean salini y a su ace le el can be measu ed by passi e
mic owa e emo e sensing a L-band. To p o ide global co e age da a o soil mois u e
and ocean salini y wi h h ee-day e isi ime, he Ea h Explo e Oppo uni y Mission
SMOS (Soil Mois u e and Ocean Salini y) was selec ed by ESA (Eu opean Space Agency)
in May 1999. SMOSsingle payload is a Y-shaped 2-D ape u e syn hesis in e e ome ic
adiome e called MIRAS (Mic owa e Imaging Radiome e by Ape u e Syn hesis).
SMOS p esen s some pa icula imaging peculia i ies: a ia ion o incidence and azimu h
angles, di e en adiome ic sensi i i y and accu acy a each di ec ion (pixels), and
geome ic pola iza ion mixing. The e o e, he accu acy o he geophysical pa ame e
e ie al depends on he knowledge o he angula dependence o he emissi i y o e a
wide ange o incidence and azimu h angles. The accu acy o he sea su ace salini y
e ie als depends on ou capabili y o co ec he wind-induced a ia ion o he b igh ness
empe a u es. To be e unde s and wind e ec s, ESA sponso ed he WInd and Salini y
Expe imen 2000 (WISE-2000) om No embe 15, 2000, o Janua y 16, 2001, in he
Casablanca oil ig, a 40 km o he coas o Ta agona (Spain). This pape is di ided in o
wo pa s. Fi s , i p esen s he de i ed sensi i i ies o he b igh ness empe a u es a
e ical and ho izon al pola iza ions wi h wind speed, and compa es o Hollinge s
measu emen s and nume ical simula ions. Second, hese esul s a e applied o he SMOS
sea su ace salini y (SSS) e ie al p oblem o di e en acks wi hin he swa h. I is
shown ha , excep o low SSS and sea su ace empe a u e (SST), he e ie ed SSS has
a RMS e o o app oxima ely 1 psu in one sa elli e pass. INDEX TERMS:6994 Radio
Science: Ins umen s and echniques; 6969 Radio Science: Remo e sensing; 6974 Radio Science: Signal
p ocessing; KEYWORDS:sea salini y, mic owa e adiome y, mul iangula , in e e ome ic adiome e
Ci a ion: Camps, A., e al., L-band sea su ace emissi i y: P elimina y esul s o he WISE-2000 campaign and i s applica ion
o salini y e ie al in he SMOS mission, Radio Sci., 38(4), 8071, doi:10.1029/2002RS002629, 2003.
1. In oduc ion
[2] In May 1999, he Eu opean Space Agency (ESA)
selec ed he Ea h Explo e Oppo uni y Mission SMOS
[Ma ı´n-Nei a and Gou oule, 1997; Sil es in e al., 2001]
o p o ide soil mois u e and ocean su ace salini y global
co e age measu emen s wi h h ee-day e isi ime.
SMOSpayload is MIRAS ha will be he i s wo-
dimensional syn he ic ape u e adiome e o Ea h
obse a ion. I s scanning con igu a ion p esen s new
challenges, mainly wo-dimensional imaging o he
scene, wi h a ying pixel esolu ion, azimu h and inci-
1Depa men o Signal Theo y and Communica ions, Uni e si-
a Poli e`cnica de Ca alunya, Ba celona, Spain.
2Ins i u de Cie`ncies del Ma , CMIMA-CSIC, Ba celona, Spain.
3LODYC, Uni e si e´ Pie e e Ma ie Cu ie, Case 100, Pa is,
F ance.
4Cen e de´ ude des En i onnemen s Te es e e Plane´ ai es,
Ve´lizy, F ance.
5Depa men o The modynamics, Facul y o Physics, Valencia,
Spain.
6Eu opean Space Agency, Eu opean Space Resea ch and Tech-
nology Cen e, Noo dwijk, Ne he lands.
Copy igh 2003 by he Ame ican Geophysical Union.
0048-6604/03/2002RS002629$11.00
RADIO SCIENCE, VOL. 38, NO. 4, 8071, doi:10.1029/2002RS002629, 2003
MAR 36–1
dence angles as he pixel a els h ough he alias- ee
ield o iew.
[3] Sea su ace salini y can be measu ed by passi e
mic owa e emo e sensing a L-band, whe e he e is a
ese ed equency band (1.400-1.427 MHz). This e-
quency band is a comp omise be ween he sensi i i y o
he b igh ness empe a u e o he salini y, small a mo-
sphe ic pe u ba ion and easonable pixel esolu ion
[Swi and McIn osh, 1983]. The WISE 2000 campaign
was sponso ed by ESA o collec expe imen al da a
unde he wides possible ange o wind condi ions o
be e unde s and he pola ime ic emission a L-band
o he sea su ace and i s dependence wi h wind and
salini y. The WISE-2000 campaign ook place a he
Casablanca oil ig, loca ed a 40⬚43' 4⬙N1⬚21' 34⬙E,
40 km away om he Eb o i e mou h, a he coas o
Ta agona (Spain), whe e sea condi ions a e ep esen-
a i e o he open Medi e anean wi h pe iodic in luence
o he Eb o i e esh wa e plume. Sys ema ic mea-
su emen s we e acqui ed om No embe 16 h o De-
cembe 18 h, 2000 and con inued du ing Janua y 9 h o
15 h, 2001.
[4] The WISE 2000 pa icipan s we e he Poly ech-
nic Uni e si y o Ca alonia (UPC, Ba celona, Spain,
p ime con ac o ), he Ins i u o Ma ine Sciences
(ICM-CSIC, Ba celona, Spain), he Labo a oi e
dOce´anog aphie Dynamique e Clima ologie (LODYC,
Pa is, F ance), he Uni e si y o Vale`ncia (UV, Vale`n-
cia, Spain), he Cen e dE
´ udes Te es es e Plan-
e´ ai es (CETP, Pa is, F ance) and he Uni e si y o
Massachuse s–Amhe s (UMass, USA) as a gues ins i-
u ion.
[5] The ollowing ins umen s we e deployed: (1) an
L-band pola ime ic adiome e (UPC), (2) a Ka-band
pola ime ic adiome e (UMass), (3) a s e eo-came a
o de e mine opog aphy and RMS slopes o he sea
su ace (CETP), (4) h ee oceanog aphic and clima o-
logical buoys wi h salini y, empe a u e, wind speed and
di ec ion, wa e heigh and pe iod senso s e c. (ICM-
CSIC and LODYC), (5) a po able me eo ological s a-
ion (UPC) wi h a mosphe ic p essu e, ai empe a u e,
ela i e humidi y and ain a e senso s, (6) a ideo
came a (UPC) moun ed on he L-band adiome e
pedes al o de e mine sea su ace oam co e age, and
(7) an IR adiome e (UV) o de e mine sea su ace
empe a u e es ima es. In addi ion, simul aneous sa el-
li e image y was acqui ed: AVHRR-de i ed sea su ace
empe a u e, SeaWi s ocean colo , and QuickSca wind
speed. (Quicksca da a we e acqui ed in neighbo pixels
o he sou h and eas , ee om coas al e ec s.)
2. Resul s o he WISE-2000 Campaign
[6] The analysis o he da a acqui ed by he di e en
senso s is desc ibed in de ail by Camps e al. [2002].
Figu e 1 shows he de i ed sensi i i y o he ho izon al
and e ical b igh ness empe a u es o wind speed
(ci cles), de i ed by i ing he adiome ic da a ob ained
o e he whole ange o wind speeds (WS ⫽0–18 m/s)
wi h leas squa es eg ession lines. When compa ed wi h
he ones de i ed om Hollinge s [1971] measu emen s
(diamonds) and he nume ical simula ions pe o med
wi h a wo-scale model [Yueh, 1997] using he Du den
and Vesecky [1985] sea spec um mul iplied by wo and
Figu e 1. (a) WISE 2000-de i ed b igh ness empe a u e
sensi i i ies o wind speed e sus incidence angle a e ical
and ho izon al pola iza ions: 1-
␴
e o ba s compu ed om
adiome ic e o and wind speed measu emen e o .
La ge e o ba s a ho izon al pola iza ion and la ge
incidence angles due o lowe numbe o a ailable da a
poin s (elimina ed because co up ed by RFI). (b) Simu-
la ed b igh ness empe a u e sensi i i ies unde WISE con-
di ions using he Yueh [1997] wo-scale model and he
El ouhaily e al. [1997] and Du den and Vesecky [1985] sea
su ace spec um, mul iplied by 1 and 2.
CAMPS ET AL.: L-BAND SEA SURFACE EMISSIVITYMAR 36–2
un in he obse ed WISE SSS, SST and WS condi ions
(solid line) (E. Dinna , pe sonal communica ion), expe -
imen al esul s show a nadi sensi i i y o 0.22 K/(m/s),
inc easing wi h incidence angle a ho izon al pola iza-
ion, and dec easing a e ical pola iza ion, being null
be ween 55⬚and 60⬚.
[7] Figu e 2 shows he measu ed azimu hal a ia ion
(ci cles) o he i s h ee S okes pa ame e s a cons an
incidence angles and model esul s ob ained using a
modi ica ion o he Camps and Reising [2001] model a
low equency. The la ges azimu hal a ia ion o he
i s wo S okes pa ame e s is on he o de o 0.1–0.2 K
app oxima ely, much smalle han he noise le el in any
SMOS pixel (⬇5 K), and he e o e will be igno ed in he
salini y e ie al p ocess (sec ion 3). (These esul s ha e
o be con i med by he esul s o he ongoing ESA-
sponso ed LOSAC campaign.)
[8] The SSS e ie al equi es he knowledge o WS
and SST ancilla y da a. Mos o en, wind s ess and sea
s a e was co ela ed and WS and di ec ion can be used
o desc ibe i s s a e. E en hough i was ound a good
ag eemen be ween buoy da a, me eo ological s a ion
da a and Quicksca image y, he in e compa ison o
di e en WS measu emen s showed an associa ed RMS
e o ⬃1.7–2.5 m/s [Camps e al., 2002], and aking hese
WS inpu s as ixed pa ame e s, he induced e o in he
e ie ed SSS may be e y impo an . In o he si ua ions,
he co ela ion be ween wind s ess and sea s a e was
qui e low. A cha ac e iza ion and modeling o he swell
would be equi ed, which will be he objec o u u e
s udies.
[9] SST es ima es om he IR adiome e and
AVHRR image y ha e p o en o be accu a e enough o
he SSS e ie al p ocess, exhibi ing a small bias (⬃-0.2
K) ha inc eases a high wind speeds, p obably because
o a lack o accu a e models o he sea oam emissi i y a
he IR.
3. Sea Su ace Salini y Re ie al in he
SMOS Con igu a ion
[10] SSS has al eady been e ie ed om L-band
adiome ic measu emen s a cons an incidence angles
[Mille e al., 1998]. Howe e , he SMOS con igu a ion
p esen s a pa icula i y: he same pixel is obse ed unde
di e en iews a di e en incidence angles, wi h a ying
spa ial esolu ion, adiome ic accu acy and sensi i i ies,
depending on he posi ion o each pixel wi h espec o
he c oss- ack. Figu e 3 ep esen s his pa icula imag-
ing cha ac e is ic o he selec ed SMOS con igu a ion.
The a ows indica e he acks ollowed by a pixel, since
i en e s in he ield o iew ( op), un il i lea es i
(bo om). Depending on he dis ance wi h espec o he
sa elli e g ound ack, he numbe o snapsho s in which
he pixel will be imaged will a y. As i goes a away
Figu e 2. (a) Th, (b) T and (c) U azimu h scans a
␪
⫽
45⬚,WS⫽2.8 m/s, SST ⫽14.1⬚C, SSS ⫽37.9 psu (da a:
ci cles, model i : solid line). See colo e sion o his igu e
a he back o his issue.
CAMPS ET AL.: L-BAND SEA SURFACE EMISSIVITY MAR 36–3
om he g ound ack, no only i is imaged less and less
imes, bu also he angula a ia ion is educed and
measu emen s become noisie , which ansla es in a
deg aded pe o mance in e ms o he e ie ed pa am-
e e s.
[11] The SSS e ie al p oblem p esen s some di i-
cul ies ha a e lis ed below:
1. The i s is he e ec o wind. A nadi , he
sensi i i y o he b igh ness empe a u es o SSS is
⌬T ,h(
␪
)/⌬SSS ⬇0.5 K/psu a SST⫽25⬚C (Figu e 4a)),
and ha o WS is ⌬T ,h(
␪
)/⌬WS ⬇0.2 K/(m/s) (Figu e
1a). Taking in o accoun he ange o a iabili y o he
SSS (30-38 psu) and he WS (0 o ⬃25m/s), WS masks
he SSS signa u e, unless i is p ope ly co ec ed.
Since he wind p oduces a small a ia ion on he
b igh ness empe a u es a bo h pola iza ions (Th, ), a
i s o de app oxima ion has been used o model i s
e ec :
Th, (
␪
,SST,SSS,WS) ⬇eh,
F esnel(
␪
,SST,SSS) 䡠SST
⫹⌬Th, (
␪
)
⌬WS 䡠WS (1)
whe e ⌬T ,h(
␪
)/⌬WS is he sensi i i y o he b igh ness
empe a u es o wind speed (Figu e 1) and eh,
F esnel is he
emissi i y o a pe ec ly la sea su ace (no wind), which
is ela ed o he elec ic ield F esnel e lec ion coe i-
cien s ⌫ ,h by
e ,h
F esnel(
␪
,STT,SSS) Ⳏ1⫺兩⌫ ,h(⌬,SST,SSS)兩2. (2)
2. The second di icul y p esen ed by he SSS e-
ie al p oblem is pola iza ion mixing. SMOS images a
e y wide wo-dimensional ield o iew, ailo ed o
p o ide h ee-day e isi ime a he equa o . Howe e ,
as we mo e away om he E- and H- planes o he
an ennas, he e is a pola iza ion mixing be ween e ical
and ho izon al pola iza ions [Claasen and Fung, 1974].
Tha is, he b igh ness empe a u es in he an enna
e e ence ame a e ela ed o he b igh ness empe a-
u es in he local e e ence ame o e he Ea h by
冋
TV
TH
册
(an )
⫽
冋
cos2
␺
sin2
␺
sin2
␺
cos2
␺
册冋
TV
TH
册
(pixel)
, (3)
whe e he angle
␺
depends on he obse a ion geome y.
The angle
␺
also depends on he Fa aday o a ion ha
canno be neglec ed a L-band. (Es ima es o be used in
Figu e 3. The SMOS imaging cha ac e is ics: a pixel is
imaged many imes unde di e en incidence angles (0⬚-
60⬚), wi h di e en spa ial esolu ions and adiome ic
sensi i i ies (⬎3K). Pa ame e s: 21 an ennas pe a m, d⫽
0.875 wa eleng hs an enna spacing,
␤
⫽32⬚ il angle, and
h⫽755 km pla o m heigh .
Figu e 4. (a) Dependence o he b igh ness empe a u e
a nadi wi h SSS and SST. (b) Incidence angle dependence
o he di e ence be ween b igh ness empe a u es com-
pu ed using Ellison e al. [1998] o Klein and Swi [1977]
seawa e pe mi i i y model (SST ⫽18⬚C, SSS ⫽36 psu, F
⫽1.41 GHz). See colo e sion o his igu e a he back o
his issue.
CAMPS ET AL.: L-BAND SEA SURFACE EMISSIVITYMAR 36–4
equa ion (3) may be used, bu hei accu acy is no good
enough o he equi ed SSS accu acy. Addi ionally,
local ionosphe ic inhomogenei ies c ea e di e en Fa a-
day o a ions wi hin a pixel.)
[12] The pola iza ion mixing p oblems (geome y
plus Fa aday o a ion) can be a oided i he e ie al
p oblem is o mula ed in e ms o he i s S okes
pa ame e I de ined as
I⫽T
an ⫹TH
an ⫽TV
pixel ⫹TH
pixel, (4)
which is in a ian o o a ions, as i can be eadily
deduced om equa ion (3). A second ad an age o his
app oach is ha i minimizes he e ec o he unce -
ain y in he dielec ic pe mi i i y models (Klein and
Swi [1977] o Ellison e al. [1998]). Since he sum o he
wo plo s in Figu e 4b is app oxima ely cons an , his
unce ain y appea s as an o se . Addi ionally, he sen-
si i i y o I o SST is ⬃0.6 K/(m/s), smalle educed om
ha i TVwe e used.
[13] The p oposed SSS e ie al algo i hm is based
on he minimiza ion o he e o de ined as
␧⫽ 1
Nobse a ions
冘
n
冋
(Imodel(
␪
n,SS
ˆS,SS
ˆT,W
ˆS)
⫺Ida a(
␪
n,SSS,SST,WS)) (
␪
,
␾
)
cos
␪
册
2
, (5)
Figu e 5. Simula ed s anda d de ia ion o he e ie ed SSS o di e en wind speed 0 m/s (le ) and
10 m/s ( igh ), di e en SSS (a and b) 30 psu, (c and d) 35 psu and (e and ) 40 psu, and di e en SST:
5⬚C (s a s), 15⬚C (c osses) and 25⬚C (ci cles). See colo e sion o his igu e a he back o his issue.
CAMPS ET AL.: L-BAND SEA SURFACE EMISSIVITY MAR 36–5

whe e Nobse a ions is he o al numbe o measu emen s
in a gi en pixel ack (Figu e 3) only limi ed by he
alias- ee FOV, Imodel is he i s S okes pa ame e
(Imodel共SS
ˆS,SS
ˆT,W
ˆS兲⫽TV
an ⫹TH
an ) compu ed om he
model wi h he es ima ed SS
ˆS,SS
ˆT, and W
ˆS pa ame e s
(azimu hal dependence is negligible and he e o e is no
included in he modeling), Ida a co esponds o he i s
S okes pa ame e e alua ed om he sum o consecu i e
Thand T measu emen s in a pixel by pixel basis (each
one obse ed unde a di e en incidence angle
␪
n) and
(⌬,
␾
)/cos
␪
is he no malized an enna adia ion pa e n
di ided by he obliqui y ac o used o gi e mo e weigh
o he samples wi h he be e adiome ic sensi i i y.
[14] In his s udy, Ida a con ains simula ed da a com-
pu ed om equa ions (1) and (2) and a pixel-dependen
andom noise e o ha depends on he angula dis ance
o he bo esigh . The e o e, simula ions show he bes
esul s ha could be achie ed in one sa elli e passage
wi h an ideal e o - ee ins umen . The minimiza ion
p oblem uses he Le enbe g-Ma qua d me hod a ail-
able in he Ma lab䉷op imiza ion oolbox, wi h es ic-
ions o SSS (30-40 psu), SST (0-40⬚C), and WS (0-30
m/s). Since he sensi i i y o SSS o he unc ion being
minimized is low (equa ion (5)), he speed and accu acy
o he e ie al is imp o ed by educing he sea ch limi s
o an in e al cen e ed a ound he alues ha could be
p o ided by o he senso s and an in e al leng h gi en by
hei associa ed e o bounds (SS
ˆT僆[SST0⫺1⬚C,
SST0⫹1⬚C] and W
ˆS僆[WS0⫺2.5 m/s, WS0⫹2.5
m/s]). I is ound ha his app oach p oduces be e
esul s han o cing he measu ed SST0and WS0 alues,
which may ha e an e o wi h espec o he eal alues
SST and WS used o gene a e he “simula ed”da a
(equa ion (5)).
[15] Figu e 5 shows he s anda d de ia ion o he
e ie ed SS
ˆS(
␴
SS
ˆS⫺SSS) o di e en wind speeds 0 m/s
(le ) and 10 m/s ( igh ), di e en SSS (Figu es 5a and
5b) 30 psu (Figu es 5c and 5d), 35 psu and (Figu es 5e
and 5 ) 40 psu, and di e en SST: 5⬚C (s a s), 15⬚C
(c osses) and 25⬚C (ci cles).
[16] The ollowing is obse ed:
1. In gene al, SSS e ie al imp o es wi h inc easing
SST, excep o high SSS and WS. The eason o ha is
no know, bu i is suspec ed ha i s due o he change in
he slope o TB(SST) om 35 psu o 40 psu.
2. As expec ed, because o he lowe numbe o
samples and he highe noise, SSS es ima es wo sen a
swa h edges because o highe noise le el.
3. Fo cing he alues o SS
ˆT⫽SS
ˆT0and W
ˆS⫽WS0
o he “measu ed” alues (including he measu emen
e o s: ⫾1⬚C, ⫾2.5 m/s) induces a la ge e o in SS
ˆS,
han le ing hem be adjus ed by he minimiza ion
algo i hm.
4. As shown in Figu e 6, he SS
ˆS e o includes a bias
(posi i e o nega i e), and a andom e o . In gene al, a
low WS, he es ima ed W
ˆS ends o be highe (posi i e
bias), which he algo i hm compensa es inc easing he
es ima ed SS
ˆS (posi i e bias). A highe WS, he si ua-
ion is e e sed. In all cases, he adjus ed SS
ˆT has a
negligible bias.
5. The e ie ed SS
ˆS has an e o o app oxima ely
⬇1 psu, in one sa elli e passage, in a swa h o ⫾300 km
om he sa elli e g ound ack.
4. Conclusions
[17] B igh ness empe a u es sensi i i ies o wind
speed de i ed om WISE-2000 da a con i m exis ing
expe imen al da a [Hollinge , 1971] and a e in ag ee-
Fig 5. Con inued.
CAMPS ET AL.: L-BAND SEA SURFACE EMISSIVITYMAR 36–6
men wi h nume ical esul s p edic ed by a wo-scale
model using Du den and Vesecky [1985] sea spec um
mul iplied by wo. A 25⬚incidence angle, he expe i-
men al esul s show a sensi i i y o 0.24 K/(m/s) and 0.20
K/(m/s) a e ical and ho izon al pola iza ion. A highe
incidence angles, he sensi i i y o wind speed inc eases
a ho izon al pola iza ion, and dec eases a e ical
pola iza ion, being null be ween 55⬚and 60⬚incidence
angle. The maximum azimu hal a ia ion is e y small
(⌬Th, peak- o-peak ⬃0.1–0.2 K). Mo e da a needs o be
acqui ed o imp o e he de e mina ion o he sensi i i-
ies o wind speed, mainly a la ge incidence angles, and
oam and swell e ec s need o be accoun ed o o
imp o e he emissi i y models.
[18] These esul s ha e been applied o a pe o -
mance s udy o an SMOS-o ien ed SSS e ie al algo-
i hm. To a oid geome ical and Fa aday- o a ion pola -
iza ion mixing, he algo i hm is based in he
minimiza ion o he e o be ween he sum o he
measu ed Thand T and he sum o he modeled Thand
T , o di e en pixel acks in he alias- ee FOV.
Excep o low SSS and low SST, he e ie ed SSS ha e
an e o o app oxima ely 1 psu, wi hin a ⫾300 km
swa h, in one sa elli e passage. Since pa o his e o is
a bias, u he esea ch will be ocused on i s educ ion
in se e al sa elli e passages in a 10 o 30 day in e al,
wind condi ions will be di e en , and he s udy o
ins umen al sys ema ic e o s.
[19]Acknowledgmen s. The WISE 2000 campaign has
been sponso ed by he Eu opean Space Agency unde
ESTEC con ac 14188/00/NL/DC. The sea emission nu-
me ical simula ions ha e been pe o med unde ESTEC
Salini y S udy 3618. We hank he Cen e Na ional dE
´-
udes Spa iales o he F ench moo ings and Me´ eo F ance
o help in he buoy da a handling. The implemen a ion o
he L-band AU oma ic RAdiome e (LAURA) o he
Poly echnic Uni e si y o Ca alonia and he SSS e ie al
s udy ha e been suppo ed by he Spanish go e nmen
unde g an CICYT TIC 99-1050-C03-01. The iendly
coope a ion and help p o ided by he pe sonnel o Repsol
In es igaciones Pe olı´ e as–Base Ta agona–Pla a o ma
Casablanca o he o ganiza ion o he campaign is g ea ly
app ecia ed.
Re e ences
Camps, A., and S. C. Reising, A model o he wind
di ec ion signa u e in he S okes emission ec o o he
ocean su ace a mic owa e equencies, Mic owa e Op .
Technol. Le .,29(6), 426–432, 2001.
Camps, A., e al., Sea su ace emissi i y obse a ions a
L-band: Fi s esul s o he Wind and Salini y Expe imen
WISE-2000, IEEE T ans. Geosci. Remo e Sens.,40(10),
2117–2130, 2002.
Claasen, J. P., and A. K. Fung, The eco e y o pola ized
appa en empe a u e dis ibu ions o la scenes om
an enna empe a u e measu emen s, IEEE T ans. An en-
nas P opag.,22(3), 433–442, 1974.
Du den, S. L., and J. F. Vesecky, A physical ada c oss-
sec ion model o a wind-d i en sea wi h swell, IEEE J.
Oceanic Eng.,10(4), 445–451, 1985.
El ouhaily, T., B. Chap on, K. Ka sa os, and D. Vande-
ma k, A uni ied di ec ional spec um o long and sho
wind-d i en wa es, J. Geophys. Res.,102, 15,781–15,796,
1997.
Ellison, W., A. Balana, G. Delbos, K. Lamkaouchi, L.
Eyma d, C. Guillou, and C. P igen , New pe mi i i y
measu emen s o sea wa e , Radio Sci.,33(3), 639–648,
1998.
Hollinge , J. P., Passi e mic owa e measu emen s o sea
su ace oughness, IEEE T ans. Geosci. Elec on.,9(3),
165–169, 1971.
Klein, L. A., and C. T. Swi , An imp o ed model o he
dielec ic cons an o sea wa e a mic owa e equencies,
IEEE J. Oceanic Eng.,2(1), 104–111, 1977.
Ma ı´n-Nei a, M., and J. M. Gou oule, A wo-dimensional
ape u e-syn hesis adiome e o soil mois u e and
ocean salini y obse a ions, ESA Bull.,92,95–104, 1997.
Mille , J. L., M. A. Goodbe le , and J. B. Zai ze , Ai bo ne
salini y mappe makes debu in coas al zone, Eos T ans.
AGU,79(14), 173–177, 1998.
Sil es in, P., M. Be ge , Y. Ke , and J. Fon , ESAs second
Ea h Explo e Oppo uni y Mission: The Soil Mois u e
and Ocean Salini y mission–SMOS, IEEE Geosci. Remo e
Sens. Newsl., Cumul. Issue 118,11–14, 2001.
Swi , C. T., and R. E. McIn osh, Conside a ions o
mic owa e emo e sensing o ocean su ace salini y, IEEE
T ans. Geosci. Remo e Sens.,21(4), 480–491, 1983.
Yueh, S. H., Modeling o wind di ec ion signals in pola i-
me ic sea su ace b igh ness empe a u es, IEEE T ans.
Geosci. Remo e Sens.,35(6), 1400–1418, 1997.
Figu e 6. Typical his og ams o e ie ed SS
ˆS, SS
ˆT and
W
ˆS o (a) SSS ⫽30 psu, SST ⫽15⬚C and WS ⫽0 m/s
(posi i e bias in SSS and WS); (b) SSS ⫽35 psu, SST ⫽
15⬚C and WS ⫽10 m/s (nega i e bias in SSS and WS),
compu ed om 16 simula ion wi h di e en andom noise.
CAMPS ET AL.: L-BAND SEA SURFACE EMISSIVITY MAR 36–7
M. Be ge , M. Ma ı´n-Nei a, and P. Wu s eisen, Eu opean
Space Agency, Eu opean Space Resea ch and Technology
Cen e, Keple laan 1, 2200-AG Noo dwijk, Ne he lands.
J. Bou in and J. E che o, LODYC, UPMC, Case 100, Tou
15 2e e´ age, 4 Pl Jussieu, 75252 Pa is Cedex 05, F ance.
A. Camps, I. Co bella, N. Du o, L. En ique, F. Julbe´,J.
Mi anda, F. To es, M. Vall-llosse a, and R. Villa ino, Depa -
men o Signal Theo y and Communica ions, Uni e si a Po-
li e`cnica de Ca alunya, Campus No d-D3, Jo di Gi ona 1-3,
08034 Ba celona, Spain. ([email p o ec ed])
V. Caselles and E. Rubio, Depa men o The modynamics,
Facul y o Physics, D . Moline 50, 46100 Bu jasso , Valencia,
Spain.
J. Fon , C. Gaba o´, and A. Julia`, Ins i u de Cie`ncies del
Ma , CMIMA-CSIC, Passeig Ma ı´ im 37-49, 08003 Ba celona,
Spain.
A. Weill, CETP, 10-12 a de lEu ope, 78140 Ve´lizy,
F ance.
CAMPS ET AL.: L-BAND SEA SURFACE EMISSIVITYMAR 36–8