Tous droits de propriété intellectuclle ré' ,\, [
,
sen cs. ~cproductlOn, rcpréscntation ct ditllision interdites. Loi du 01/07/92.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSlNG, VOL. 4:1, NO. 5, MAY 2()05
025
T'Ile Emissivity of Foanl-Covered Water Surface at
L-Band: Theoretical ModeliI1g and Experilnentai
Results FrOln t11e Frog 2003 Field Experiment
Adriano Camps, Senior Membel; IEEE, Mercè Vall-Ilossera, Membel; IEEE, Ramon Villarino, Student Mel/dm; IEEE,
Nicolas Reul, Bertrand Çhapron, Ignasi Corbeil a, Membel; IEEE, Nuria Duffo, Membel; IEEE,
Franccsc Torres, Mem.bel; IEEE, Jorge José Miranda, Student MembeJ; IEEE, Roberto Sabia, Studenl Membel; IEEE,
Alessandra Monerris, Studem Membel; IEEE, and Rubén Rodrîguez
f\bstract-Sea SUl-t'ace salinity Cllll be measured by microwave
radiometry at L-band (1400-1427 MHz), This frequency is a
compromise between sensitivity to the salillity, small atmospheI"ic
perturbation, and reasonable pixel resolution, The description
of the ocean emission depends on two main factors: 1) the sea
water permittivity, which is a function of sali nity, temperature,
and frequency, and 2) the sea surface state, which depends on the
wind-induced wave spectrum, swell, and raill-induced roughness
spectrum, and by the foam coverage and its emissivity, This study
presents a simplified two-layel' emission model for foam-covered
watel' and the results of a controlled experiment to measure the
t'oam emissivity as a function of salinity, foam thickness, incidence
angle, and polarization, Experimental results are presented, and
then compared to the two-layel' t'oam emission model with the
measured toam parameters used as input model parameters, At
37 psu sait water the foam-induced emissivity inCl'ease is ",0,007
per milIimetel' of l'oam thic!mess (extrapolated to nadir), increasing with increasing incidence angles at vertical polarization,
and oecreasing with increasing incidence angles at hol'Ïzontal
po!arÏzation,
Index 1liJ'/lls-BI"ightness temperature, emission, t'oam, microwave radiometry, salinity, sea,
I. INTRODUCTION
T
HE IMPROVEMENT of weather prediction, e.g., for natural catastrophes, recluires the knowledge of soil 1110isture (SM) and sea surface salinity (SSS) on a global scale. The
knowledge of the SSS distribution at a global scale with a 1110derate revisit time is important to climate predictions, since SSS
is a tracer of sea surface currents and an indicator of the difference between evaporation and precipitation (E-P). Water density is determined by temperature and sali nit y, and hence the
Manuscript reccived July 8, 2004; revised October 9, 2004. The ('oam
emission modeling study was sponsored by the European Space Agency under
Contracl5165/0 I/NLISF. The FROG 2003 campaign was sponsorcd in part by
MCyT AE. ESP2001-4524-PE, in part by the Spanish Comisi6n Interministerial de Ciencia y Tccnologîa and EU FEDER (CICYT TIC20D2-04451-C02-0 1),
and in part by Grant "Distinciô dc Recerca de la Generalitat de Catalunya,"
resolution: UNI/2120/20D2 on July 19, DOGC 3684, 24/07/20D2.
A Camps, M. Vall-llossera, R. Villarino, 1. Corbeil a, N. Du1To, F. Torres,
J. J. Miranda, R. Sabia, A. Moncrris, and R. Rodriguez arc \Vith the Department
0(' Signal Theory and Communications, Universitat Politècnica de Catalunya,
E-08034 Barcclona, Spain (e-mail: [email protected]).
N. Reul and B. Chapron are with the IrREMER, Depareillent d'Océanographie Physique et Spatiale, Laboratoire d'Océanographie Spatiale, 29280
Plouzané, France.
Digital abject Identifier 1D.ll 09!rGRS.2004.839651
thermo-haline circulation can also be monitored by SSS measurements. Salinity monitoring is useful to improve the quality
of the El Nino Southern Oscillation (ENSO) numerical model
predictions, as weil [1], [2].
At present, there are two planned space missions to measure
global sea surface salinity maps with 10-30-day revisit time:
the Soil Moisture and Ocean Salinity (SMOS) Earth Explorer
Opportunity mission l'rom the European Space Agency (ES A),
and the Aquarius Earth System Science Pathfinder (ESSP) mission from the National Aeronautics anc! Space Administration
(NASA), with launch dates in February 2007 and September
2008, respectively.
The SSS retrieval l'rom microwave radiometric measurements
is based on the fact that the dielectric constant of seawater is
a function of sali nity and temperature [3]. The sensitivity of
brightness temperature (TB) to SSS is maximum at low l11icrowave üequencies, and the optimum conditions for salinity
retrieval are found at L-bancl, where there is a protectec! banc! for
passive observations (1400-1427 MHz). However, even at this
frequency the sensitivity of TB to SSS is low: 0.5 K pel' psu for
a sea surface temperature (SST) of 20 oC, decreasing to 0.25 K
pel' psu for an SST of 0 oc. Since other variables influence the
TB signal (polarization, incidence angle, sea surface temperature, roughness and foam), unless they are properly accounted
for, the SSS determination will be erroneOllS.
During SMOS Phase A two field experill1ents named the
Wind and Salinity Experill1ent (WISE) were sponsored by the
ESA in the fall of 2000 and 200 l to better understand the wind
and sea state effects on the L-band brightness temperatures.
They consisted of acquiring long time series of TB l'rom an
oil rig in the northern Mediterranean Sea to relate it to the sea
surface rollghness (wind speed, significant wave height) and
the instantaneous foam coverage [4].
Several effects were not complete!y understood during the
WISE field experiments: the emissivity of foam and the impact
of rain and oil slicks on the 'lB variations. Under Spanish National funds, the Foam, Fain, Oil Slicks and OPS Reflectometry (FROO) field experiment was carried out at the Institut cie
Recerca i Tecnologia Agroalill1entàries CIRTA) facilities at the
Ebro River delta.
This work is organized into two well-defined parts. ln the
first one, a two-Iayer sea foam emission mode! at L-bancl is
o196-2892/$20.DO © 2005 IEEE
Tous droits de propriété intcllectucllc réscrvés. Reproduction, rcprésentation ct dimlsion interdites.
L;Xiii,o;/07/92.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 4J, NO. 5, MAY 2005
926
Radlotrleler
T, is the Foam layer physical temperature, the coefficient
Il
RJ) is the rel1ection coeftlcient of the foam layer medium with
p
the effective dielectric constant C N" and is given by
te
where
Region 0 (air)
li
y
(2)
p
where I[i is an attenuation factor that depends on the foam layer
thickness d, the electromagneticwavelength /\0, and the effective permittivity CNe,
o
Region 2 (waler)
Ew
\\
(3)
Fig. 1. Oeomclrical conljguration l'or thermal emission l'rom foam-covered
ocean. The foam layer is region 1 and is absorplive and scatlering. Region 2
is air bubbles cmbedded in sea \Valer and is absorplive (l'rom rS]; sec Fig. 3).
presented. In the second one, the FROG 2003 field experiment
is described, and the foam emissivity measurements at vertical
and horizontal polarizations are presented in a wide range of
water salinities from 0-37 psu. These measurements are then
compared to the model presented in the first part using the
measured foam parameters: bubble radii histograms, bubbles'
water coating thickness, derived bubbles' packing coefficient,
foam layer thickness, and the void fraction beneath the foam
layer.
In (2), R~l are the Fresnel rellection coefficients between the air
(region 0) ancl the l'oam (region 1)
COS(Bi) -
+ JENn -
cos(O;)
cNn
JENn - Si11
COS(Oi) -
ENa COS(Oi)
2
Bi
(4a)
2
+ VENrx
2
sin Oi
2
- Sill (1.;
tl
tl
ti
1
c
VCNet - sin
2
(Oi) - Jc,," -
Sill
2 fJ·
i
(Sa)
VEND: -
Rt,2(B i )
=
sin 2 (0'i)
EwVCNet - sin
EwVENe, -
2
2
+ Jc,,"
- sin
2
Bi
(0.;) - cNaVc w - sin 2 Bi
sin (0;)
+ ENa JE",
-
.
(Sb)
2
sin fJ·i
Region 2 consists of air bubbles embedded in the ocean bacleground and is assumed to be absorptive. To solve (1 )-(5), one
needs to define an effective permittivity for region 1, namely
CNe" and for region 2, namely Ew.
A. Brightness Temperature Modeling of the
Foam-Waler System
B. Effective Pennittivity ofSea Foam Formations at L-Band
Following Guo et al. [5], it is assumed that foam on the
ocean surface is composed of nearly spherical coatecl bubbles
describecl by an outer radius '/', made of an air core with permittivity C(J, sllrrollnded by a shell of sea water with thickiless
band permittivity C\V' The foam-covered ocean is moclelecl
by the succession of three media: the air (region 0), a fDiuH
layer defined as a region of effective permittivity cNo: with a
layer thickness cl (region 1), and the lInderlying seawater with
some air bubbles (region 2) with permittivity Ew (see Fig. 1).
Boundaries between each region are assumed l1at.
The brightness tempe rature of the foam-water system at incidence angle Oi and polarization li = h (horizontal) or v (vertical)
is then equal ta
The main parameter of the previous multilayer emissivity
model for fmun is tbe effective permittivity EN" of the foam
layer considerec1. To c1etlne tbis parameter, the well-known
Lorenz-Lorentz and van cle Hulst equations can be usec1 and
modifiecl for the polyclispersed system of bubbles. The tlrst
formula takes into account dipole interaction of bubbles in a
close-packed dispersecl system (the quasi-static approximation). The van clé Hulst equations describe the contribution of
the multipole moment of bubbles into the effective permittivity
of the system. Spectral calculations by Chemy ancl Raizer [7]
show that the fi.rst resonant electromagnetic elTects by van de
Hulst's mechanism occur for bubbles' radius (], ::::: /\0/'1. At
L-band (/\0 = 21 cm); this corresponds ta bubble diameters
on the orCIer of 10 cm. Such very large bubbles are extremely
rare at the sea surface and therefore, the multipole mechanism
may be neglected at L-band for which only the clipole term
(1)
"a
P
(4b)
and RJ,2 are the Fresnel rellection coeftlcients between fa<lll1
(region 1) and water (region 2)
II. SEA FOAhl ;]MISSIVITY MODELINO AT L-BAND
Foam and whitecaps belong ta the class of colloidal systems
that include two phases: atmospheric gases and sea water. In
the present worle, the foam layers at the ocean surface are modeled as a medium of clensely paCked sticky spherical air bubbles, coated with thin seawater coating, following the approach
of Guo et ar. [5]. The dipole approximation model developed by
Dombrovskiy ancl Raizer [6] is then usecl to clescribe the effective permittivity of the system.
c
tl
sin Oi
VENa: -
a
Tous droits de propriété intellectuelle réservés. Reproduction, représentation ct difTusion intcrditcs. Loi du 01/07/92.
CAMPS /'/ (II.: EMISSIVITY OF FOAM-COVERED WATER SURFACE AT L-BAND
(l05
927
may be considered. In the present work, we use the dipole approximation mode! developecl by Dombrovskiy and Raizer 161
to describe the cffective permillivity of the system. Il involves
the use of a modification of the Lorenz-Lorentz equation and
yields the following simple formula for the complex effective
pennitlivity ENnof a foam layer [6], ITI:
"nt
ith
1 -1- ~7r N . o{r)
er
,,y
l -17rN. n{r)
(6)
!
where
(7)
and N is the volumetric concentration of the bubbles, n:('r) is the
complex polarizability of a single bubble with external radius 'l',
N. is the so-called packing coefficient or stickiness parallleter,
and Jl J CI') is the normalized probability distribution function of
the bubbles' size. In natural media such as foam, the densely
paclœd particles can have adhesive forces that make them adhere
to form aggregates. This e[fect is accounted for in the model by
the stickiness parallleter h;, which is inversely proportional to
the strength of the attractive forces between bubbles [8].
According to Dombrovskiy and Raizer [6], the complex polarizability depends on the external radius of the bubbles 'l', the
complex permittivity of the shell medium (salt water) Ew, and
the bubbles' filling factor q = 1- 6/1' following
Fig. 2.
Location
or the WISE ancl FROO lielcl expcriments.
(aged [oam). Numerous observations of oecanic bubble size distributions are reported in the literature based on acoustic, photographic, optical, anc! holographic methods [1 1]. Currently, it is
not clear how to parameterize the ocean surface bubble size distribution. Following Bordonskiy et al. [12] and Dombrovskskiy
and Raizer [6], we used a'T distribution for the size distribution
function of the bubbles
(9)
where A and B are parameters of the distribution with '/'1' =
being the !11ost probable radius.
Finally, to ca\culate Ew, a simple physical model based on induced c!ipoles is used. Let Es", denote the permittivity of the seawater (see [13] for Es", at L-band), anc! fn the fractional volume
occupied by the air bubbles. Then, the effective permiltivity Ew
is given by the Maxwell-Garnett mixing formula [5]
.11/ B
\
')
Ill,
il
1\
!
:,1
•
!
p!'.j
LJJ
Since the bubble size distribution depends on the vertical position within the foam layer, the effective pennittivity depends
on the vertical position as weil: ENu = ENo:(;;). In the simplest case, the foam-water system may be modeled as a succession of elementary foam layers, each of them having a homogeneous effective dielectric constant. However, the exact dependence of such fUl1ction with vertical position, which de pends
on the vertical distribution of the bubbles' size, is very poorly
known. It is very like!y that the vertical distribution of the bubbles' size p.rCr, ;;) is a function of the intensity and sc ale of the
underlying breaking event. Moreover it will certainly strongly
evolve during a transient breaking evenl. Nevertheless, in order
to keep a tractable number of parameters in the presel1t model,
we choose to consider a uniform vertical distribution of bubbles'
sizes ]lfC",;;) = JJfCr) within the foam layer.
The foam void fraction (i.e., the ratio of the volume of air to
the total volume of the foam) depends on the distribution of the
bubbles' filling factor q. Therefore, the distribution of bubbles
radii p./'Cr) together with the distribution of coating thicknesses
f (6) determine the foam layer void fraction. In the present simplified model, we fixed the value of the shell thickness b, but
the outer bubble radius T is a ranclom variable. According to
Dombrovskiy [9], this approximation reflects an experimentally
establishec! fact for an emulsion layer of foam (young foam),
but it requires verification for a foam with honeycomb structure
Ew
= Esw
1 -1- 2f"y
l - j'nY
'1
(l0)
where
Y=
1. - Esw
1. -1- 2Esw
(lI)
Note that the effective permittivity Ew here c!oes not include
scatlering extinction, which is small due to the fact that the seawater is heavily absorptive. According to our simplillec! emissivity !11odel for [mlln, the brightness temperature induced by a
sea foam layer is a function of
TB
= function(O.,, ./',p,T,,, 'l'Ill b, li;, d, fn, SSS, SST)
(12)
where (Ji is the rac\iometer incidence angle, f the electromagnetic frequency, p is the polarization, T" is the foam physical
temperature, TT' is the most probable radius, 6 is the bubbles'
water coating thickness, li; is the bubbles' packing coefficient
(same as stickiness parameter), cl is the l'oam layer thickness,
fa is the void fraction beneath the l'oam layer, and finally, SSS
Tous droits de propriété intellectuelle réservés. Reproduction, représentation ct dintlsion interdites. Loi du 01(07/92.
HiEE TRANSACTIONS ON GEOSCIENCE AND IŒMOTE SENSING, VOL. 43, NO. 5, MAY 200S
928
c)
(a)
e)
fl
(b)
Fig. 3. (a) Pool filled with \Vater during a foam emissivity measurement
incidence angle scan. (b) Detail of the air diffusers used to gencrate foam.
1)
and SST are the sea surface salinity and temperature, respective!y, Ali these parameters are either known (Oi, j', li) or measurecl (TH, '/'J!' li, cl, fa., SST, SSS), except the packing coeff1cient
(h:) that will be derived by minimizing the rms error between
the FROO 2003 measurements and lhc lheoretical mode! (see
Scction III-B).
TIl, SEA FOAM EMISSIVITY MEASUREMENTS AT L-BAND
A, FROC 2003 Field Experilllellf Description
Thc FROO 2003 cxpcrimenl was sponsorccl by lhe Spanish
Oovernmcnt 10 detcrmine thc emission of these factors. The
controlled cxpcrimcnt was performcd at the facililics of thc 1nstilllt de Rccerca cn Tècniqucs Agropecllitries CIRTA) at Poble
Nou dcl Dclla, in lhc Ebro Rivcr mouth (Fig. 2).
The L-banc1 Automatic Rac1iometer (LAURA) radiomeler [4]
was mountcd on onc sidc of a 3 m x 7 m pool fillcd with a mixture of sea watcr and frcsh water l'rom the Ebro River. The river
watcr was mixcd with the sea waler 10 adjusl lhc salinily ovcr
thc range l'rom 0-37 psu [Fig. 3(a)], A melallic net was mounted
around the pool, and at the edges the net was inclinecl45° to reHect lhe sky radiation, avoicling contamination of the measurements l'rom ground racliation collecled through thc scconclary
h)
Fig. 4. Suite of instruments dcployed to acquire ancillary data. (a) Six subsurface temperature sensors. (b) Metal bars of the instantaneous surface level
sensor. (c) and (d) Array of 16 gold electrodes of the air fraction measurement
systcm. (e) Periscope \Vith vidco camera to acquire vertical l'oam profiles.
(1') lnfrarecl radio!1leter to measure surface cmission with l'OHm. (g) Video
camera to clcrive surface foam coveragc [as in (h)[.
lobes of the antenna, and cnlarging the raclio-eleclric horizon.
The measurcmcnl slralcgy consistee! of performing incidence
angle scans l'rom 25° to 55 0 , in 5° sleps, and measuring lhe surfacc's emissivity wilh the foam gencrators off, on, ancl finally
off again at each incidence angle, Ali measllremenls were pcrformccl ancr Sllnset to avoid sllnconlamination, andlhe pool was
orientecl to the north which minimizecl the Oalaclic backgrouncl
noise, Foam was createcl by a network of 104 air c1iffusers placecl
at the bottolTl of the pool, and conncclecl to an acljustable air
pump with a maximum air flow of 500 m 3 /h [Fig. 3(b)].
A series of instrumenls was cleployecllo acquire ancillary clala
to be usecl as inputs to theoretical moclcls as follows:
• six temperature scnsors locatecl below the water surface
[Fig, 4(a)];
two metallic bars usecl in a condllctivity-basecl inslanlaneous surhlce level scnsor [Fig. 4(b)J;
c
Tous droits dc propriété intellectuellc réservés. I{eproduction, représcntation ct difllision interdites. Loi du 01/07/92.
CAMPS 1'1 (//.: EMISSIVITY OF FOAM-COVEIŒD WATER SURFACE 1\1' L-BAND
)05
929
c)
f)
i)
Fig. 5. Foam vertical profiles at di l'l'ercnt salinities (seale in ccntimetcrs). (a) () psu, (b) 5 psu, (c) 10 psu, (d) 15 psu, (e) 20 psu, (f) 25 psu,(g) 30 psu, (h) 37 psu,
and (i) horizontal view of natural sea surface bubbles (photograpl) l'rom Gran Canaria island, SSS33.8 psu, rel'erence size = 0.50 Euro com).
• array of 16 golc1 eleclrodes [Fig. 4(c)] of a conductivilybasec1 subsurface voie! fraction profîler [Fig. 4(d)];
• uttrale video camera wilh a large angular lens mounlec1 on
a periscope to acquire foam vertical profiles [Fig. 4(e)];
• CIMEL CE 312 four-bane! lhermal-inl'rarecl rae!iomeler:
8-13, 11.5-/2.5, 10.5-11.5, ane! 8.2-9.2 l/,Ill [Fig. 4(0],
l'rom the Universitat de València;
• Sony SSC-DC393 video camera and zoom [Fig. 4(g)]
10 match the field of view to the radiometer' s anlenna
beam [Fig. 4(h)], usecl to derive the water surface's l'oam
coverage;
• porlable meteorological station usee! to measure air temperalure, pressure and relative humiclity, and wincl speee!
ane! direction.
))
Il
L-J. ___________~
j
B. Radiometrie Measure/llents
Foam-Free Measurements: The fîrsl measuremenls consisted of measuring lhe emissivily of the flal waler surface at
clifferent salinily levels
=1 -
(0 , E.(
, ,/' , SSS , SST)).
.
(13)
Down-welling radiation ane! finite beamwic\lh effecls were
correclec\. As expected, the brighlness temperalures (or alternatively emissivilies) decrease with increasing salinily levels, in
agreement with the lheory using lhe Klein-Swift dieleclric constant model [9].1
,Wat.el'( .
/' ,0' SSS ,~
SST)
cp
rF'l',,-,ncl
p
IThe goodness of the agrcement bctween the moclelcd Tu ancl the measured
~().12 K. Salinity retrievals using measurcd Tu cxhibit a 0.26 psu bias
for the SSS l'rom 0-37 psu, and SST l'rom 14 oC to 20 oc.
1'/1 is
Tous droits de propriété intellectuelle réservés. Reproduction, représentation cl dil1i1sion interdites. Loi du 01107/92 .
.......... _.. ___ ._. _ _ _
(no
~
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _..........._
...
""·I..i",W",tfW
Olli'_ _
W_hN1 iit
tt
L.b·'~'""i;;ÎiliO_"
IEEE TI(ANSACTlONS ON GEOSCIENCE AND REMOTE SENSINCi, VOl. .•U, NO. 5, MAY 2(H15
a)
SSS: a psu. mean f p: a0211m, medlan f p: 78911m,
c)
Fig. 6.
Delail of foam and bubbles verlical profiles. (a) Fresh \Vatel' and (b) 15 psu. Corresponding bubble radius histograms. (c) Fresh waler: mean radius
=
80211.111, median radius = 789/1.111 and (d) 15 psu (mean radius = 437 Ilm, median radius = 423 J1m).
Foam Mcasurclllcnfs: As explained in Section II, in addition
to the foam surface coverage needed to estimate the foam emissivity from the measurements, modeling foam emission requires
the knowledge of: 1) the fmuu layer thickness; 2) the bllbbles'
radii distribution; 3) the bubbles' water coating thickness; 4) the
stickiness parameter; and 5) the air fraction beneath the [oam
layer, as follows.
• The foam layer thickness was measurecl by the vicleo
camera mountecl on the periscope. For the same air f10w
rate the foam thickness is very small for Fresh water
(Iower surface tension) and increases with increasing
salinity (higher surface tension). Fig. 5 presents ni ne
vertical profîles corresponcling to nine clifferent salinities
l'rom 0-37 psu. Alltomatecl image processing was llsecl
to cletermine the mean thickness of each photogram in a
systemalic manner.
• The estimation of the bubbles' radii clistribution requirecl
the recognition of isolated bubbles, ancl fîncling the equivaIent spherical bubble. Fresh water bubbles are lar'ger,
ancl their shape is less spherical than salt y water bllbbles.
The measurecl mean raclii are: ~0.80 mm for fresh water
bubbles [Fig. 6(a) ancl 6(c)], ~0.63 mm at 5 psu, ancl
nearly constant ancl approximately equal to ~0.44 mm
above 10 psu [Fig. 6(b) ancl6(d)J.
In orcier to check the gooclness of the artifîcially generatecl fmul1 as comparecl to natural foam encountered over
the sea aner a wave splash, 80 photographs of the sea
surface (SSS ~ 33.8 psu) were acquirecl in the coasts
of the Gran Canaria islancl al'ter wave breaking. A top
view is shown Fig. 7(a), with its corresponcling histogram
[Fig. 7(b)]. This is representative of the whole volume,
since the foam containecl only one layer of bubbles. 2 The
most noticeable fealures of the histogram are that the average radius is ~0.60 mm, as compared to ~0.40 mm
for the artificially generated bubbles (Fig. 6), and that the
histogram is less symmetric around the mean value, with
a much longer tail, which corresponds to Im'ger bubbles
formed l'rom combinations of smaller ones. The best fît
by a gamma function bubble radii pelf' is also shown in
Fig. 7(c), showing that the mcasurccl histogram is more
peaky and narrower them the gamma pdf.
• Bubbles' coating thickness measurecl from zooms of the
vertical profiles (e.g., Figs. 6 and 7) are in the 10-20-/l,m
range. However, this is not a critical parameter, since the
foeun emissivity, as preclictecl by the theoreticalmodel explain in Section Il, shows little variation with il.
20lher experimenls [1 (rI have shown lhe agreemenl belween lhe bubble sizc
hislogram of arlificially gencralecl sca foam anclnalural sea l'Dam.
Tous droits dc propriété intelleetucllc réservés. Reproduction, représentation ct ditllIsion interdites. Loi du 01/07/92.
CAMPS cl 1//.: EMISSIVITY OF FOAM·COVEIŒD \V1\ITil< SURFACE 1\1' L·BAND
l):J1
781 pm, ffi"dlun r
600
b)
1000
Radius ulm]
1500
2000
Gamma functloo fil a'" 2.9 b=211.9
:lA 10"
î
1.6
695 Ilm,
r
pd/Ir).
'I.G
--!--.r'·I.e ï ;
b··J'lJ;l)
1.4
1,2
0.8
0.6
f \,
\1
il
,-\
,~
"
:1'
,1
0.;\
0.2
a)
c)
°0
1
1
i
\\'\ \
,\\
\. '
\.'.
.",~~~
1000
~-~'\:.~>..;:,,-,,,,:?,,,,.,,.~~~,.~__,...~J.,;;o...
.1iOOO
2000
6000
'" d
ifflil11ml
Fig. 7. (a) Top view of nalural sea f'oam (Gran Canari a Island, scale: 50 euro cenl coin). (b) Bubble radius hislogram (mcan radius
radius = G9G Il.m. (c) Corresponding bubble radius hislogram (higher pcak f'unclion), and bcst fil by a gamma f'unclion (Jower peak).
TABLE 1
ESTIMAI'ED PACKINO COEFFICIENT l" MINIMIZINO THE RMS ERROR BETIVEEN
THE FROG 2003 MEASUREMENTS AND THE THEORETICAL MODEl
SSS
Optillluill K (H-pol)
Oplillluill K (V-pol)
Average K
of H· and V-pol
5
10
15
20
25
30
34
0.14
0.'10
0.'13
0.13
0.17
0.17
0.21
0.03
0.04
0.11
0.12
0.15
0.15
0.17
0.08
0.07
0.'12
0.12
0.16
0.16
0.19
• The bubbles' packing coefflcient or stickiness parame ter
plays a very important role in the foam emission. However, since no direct means has been found to measure it,
the approach followecl has been its determination by minimizing the l'ms error between the moclel ancl the measurements. Table 1 shows the clifferent values of h~ founcl
for clifferent salinities ancl polarizations. Above 10 psu the
values at both polarizations are very close, as expectecl, but
for low salinities they are significantly clifferent, which is
not unclerstoocl, although it may be attributecl to the more
flattenecl shape that the large bubbles have at low salinities, as comparecl to high salinities. ln the thircl co[umn the
L
iSOOiJ
= 781 Ilm, mosl probable
optimum Il. values are chosen to minimize the overall l'ms
error at both polarizations.
• The void fraction beneath the faLun layer is also a critical
parameter. Ils cletermination From video imagery is subject
to large uncertainties clue to the clepth of the fielcl of view
(~I cm). To avoicl this problem, the relative conductivity
between foam ancl the air-water mixture below ancl water
was measurecl. Curtayne's equatian [14]
(14)
relates the ratio of the foam ancl water conductivities to the
liquicl fraction (fit, The void fraction beneath the foam layer
fu can then be c\eterminecl as
(lS)
Fig. 8 shows two lime series of the evolution of fa. for each
of the 16 golcl electrocles. Fig. 9 shows the corresponc\ing average ancl stanclarcl cleviations of fu. of each electrocle. From
these plots, c\erivecl l'rom measurements acC]uired with the same
air f1ow, it can be appreciatecl that far fresh water .la is ~ [S% to
Tous droits de propriété intellectuelle réservés. Reproduction, représentation ct dillllsion interdites. Loi du 01/07/92.
t '
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 4:1, NO. 5, MAY 2(JO)
932
Temporal sequence (ail the electrodas)
E
o
SSS: 0 psu
100
00
SSS: 5 psu
16
r--I.....-J-"
r
':c
11
1'1'
8,
6
~
c
23c:
0
10
0
50
E
....'"
w
êiî
0
00
$
.!.
ë(
a)
20
40
60
80
100
1(
50
SSS: 10 psu
16
'là 1
1
100
50
SSS: 15 psu
°111/10
.
10
o
Samples [sec]
~
100
fi 10
fl
16·
11 '
6·
50 ---1.J00
'/0
E
o
Temporal sequence (ail the electrodes)
°nllO
SSS: 37 psu
16
J
50
SSS: 25 psu
. -- --" .. _.
'100
°fal1O
--~.
50
SSS: 37 psu
E
o
00
~I
16
r
1'11
100
0fal10
~1
1
'
6
·1l."----~_----'.
1
1
,1
o
Air-waler content (C) lime sequence lor every electrodc. (a) SSS
(15 electrodes, the top elcctrocle \Vas air) and b) SSS = .37 psu (16
electrodes).
Pig.8.
o psu
20% beneath the [oam layer, increasing sharply just below the
water surface, while for salt y water, fit is l11uch lower, ~S%,
and the foam layer is thicker,
Fig. 10 shows a typical fmlll1 emissivity l11easurement corresponding to a salinity of 33.21 psu. Radiometers' raw clata (m V)
are shown at H- (lower plot) and V- (upper plot) polariza'tions,
ancl inciclence angle (l'rom 25° to 55°, lower panel) as a funclion of the time [Fig, lO(a)], The L-band surface emissivity is
plotted at H- (lower plots) ancl V - (upper plots) polarizations of
the foam-free surface (clottecl line) ancl with 1'0 a 111 (solicl line).
The foam coverage as a funelion of the incidence angle is shown
in green in Fig, IO(b). The average foam coverage for aIl incidence angles is 86.4%.
The foam emissivity can be clerivecl Ü'OI11 (16)
Total(o) =
eh}v
F,
eFOalll(O)
. . li ,v
-1- [1 - FJ . e"'h,v
Wat.el·(O)
(16)
0
1
50
Alr·Waler ratio (%]
100
50
Air·Water ratio 1%)
100
Fig. 9. Air-water l'l'action U,,) mean value ancl standard deviation 01'
instantaneous values (uf" divided by 10) l'or every clcctrode at SSS
0 tn
37 psu. Each parallel gold clectrode (3 mm x 4 mm, total 16 e1eclrodes) is
separatecl by 2 mm l'rom the adjacent c1cctroclc (vertical scale equals 8 cm).
Element number 16 is locatccl ab ove the surface, ancl hence the air-water ratio
is 100%.
=
where F is the foam fraction coverage (in the sea it is cletermined by the wind, the air-sea instability, the feteh, the salinity
etc., [15]) From which the variation with respect to the nat surj~lce can be derivecl
A.,
Deh,\'
(0)= Cil,v
.,Foalll(f1)_ ,Wl1ter(fl)=
Lh,v
{7
{7
{7
e To 1.11 1(0)
'Ii,v
_ ('Water(o)
li' 'il,v
.
( 17)
Fig, 1 1(a)-(h) shows the intercomparison of the FROG ll1easurell1ents of e WaLer (dotted line) and eFOillIl (solid lines with
circIes) scaled to 100% coverage (16), and the theoretical
values computed using the model describecl in Section II with
the ll1easurecl foam parameters as inputs (solicl lines with triangles), The values of the stickiness parameter usecl are the
Tous droits de propriété intellectuelle réservés. Reproduction, représentation ct dil1lision interdites. Loi du 01/07/92.
CAMPS c/ (/1.: EMISSIVITY OF FOAM·COYERED \VATER SURFACE AT L·BAND
933
= 0.397, 100'''v
30-Apr-2003,100'''H
1 W~::_ ':::"':":':'---·:_"-~~?~'~~~:_"':'-'-
30-Apr-2003, SSS: 33.21 psu, SST: '18.7 oC
-200
--~.~.~~._':'",:~--
-
'
0.258, SSS: 33.21 psu, SST: 18.7 oc
, -
H10am
Vroom
H foarn
V
foam
••••• Hfoam.tree
0.9 -
0.8
Hfonm.free
V,oam.rree
c
!o'!.',n.•
•••••
0.7
j'
A
V
foam·free
Hspccular
.. !
~O.6 :,_~~~~:~
__
.
•
____
:;cn:O.5 ___~_:.L..~:
... ~~.
.!!l 0.4
E
LU O.
;
..;. .............. ·0
..;.. ;;;..~.• Q.." ............. ,
.
••••••••••••••• (i)-........... .
••• " .........
.
~
..
,
;~; .............~ .....:.....~~••l...............l ...............~
, "
~2
0.1
a)
b) ~5
35
30
40
45
50
Incidence Angle [0]
Fio 10
Foam cmissivity mcasurcment at 33.21 psu. (Lower curves) H-polarization and (upper curves) Y-polarization, with (solid line) foam and (dotted line)
(a) Radiometers' raw data (millivolls) versus (lower plot) incidence angle and (h) surface's emissivity and (Iight gray plot) foam coverage as a
function of incidence angle. The variation of the surface's foal11 coverage versus incidence angle is due ta the variallon of the 100U11 patterns gcncrated by the array
of diffusers in the 3 m x 7 m pool [see Fig. 4(h)J.
wi~ilOU; foam.
optimum ones found at each sali nit y, which in general increases with SSS as the bubbles are more densely packed, as
shown in Table I. The l'ms error between the measured data
and the theoretical foam emissivity model is calculated and
shown in each figure ((J' "'h,,,) and varies l'rom 0.008 to 0.017
at H-polarization, and l'rom 0.011 to 0.033 at Y'polarization.
In general the agreement is much better at H-polarization th an
at Y'polarization. At Y -polarization, the measured values show
a larger variation with incidence angle than the model predictions, which requires further analysis and refinement of the
mode!. At H-polarization the agreement is excellent, except at
low salinities, where there is a bias between the measured and
preclictecl emissivites at ail incidence angles.
As shown by Guo et al. [5, Figs. 6~8], the presence of foam increases the brightness temperature (emissivity) almost linearly
with foam thickness until it saturates. 3 Finally, Fig. 12 shows the
measured 6eh,,,(f)) values normalizecl to the foam thickness as
a function of the salinity at H- and Y -polarizations. The oscillations in Figs. 12(a) and 12(b) at low salinity leve1s are due
to the small values of the foam thickness, which are clifficult to
measure [e.g., Fig. s(a)]. Fig. 12(c) and (d) shows the smoothecl
values.
As founcl during the WISE measurements [16], ·at L-band
the foam emissivity increase is Im'ger at Y -pohu'ization than at
H-polarization, and it increases with inciclence angle at Y -polarization, while it clecreases at H-pohu·ization. In order to get the
same foum brightness temperature increase as in WISE [161, the
effective foam thickness (assumed to be the same for ail foam
patches regardless of the type of foam) should be approximately
8 mm.
3Brightness tcmperature saturation depends 011 bubble sizc and frequcncy. At
37 GHz ittakes place l'or l'oamthickness around 4-8 cm, and at 19 GHz around
16-22 cm [5, Fig. 6]. No numerical simulations have been reported at L-band,
but thc saturation thickness must be much larger. lndeed, in the FROG 2003
Illcasurements, the cmissivity increasc is very modcst, which suggcsts that the
thickness or the generated t'oam is far away l'rom the saturation thickncss.
IV.
CONCLUSION
The presence of foam increases the emitted brightness
temperature, since it acts as a transition layer that aclapts the
wave impedance of the two media: water ancl air. The increase
depends on the fraction of the sea surface coverecl by foam
F(U10 ), which is usually parameterizedin tenllS of the wincl
speed, but it clepends on other factors, such as lile air~sea
temperature difference, the sea surface temperature, the fetch,
etc. [15]
,Total(o) -el'
F(U10 ). "
oFoitm ,1. [1 - F(U )]. eSea
p'
10
'1'
-- el'Sca
--
+ F(U10 ). [e'1'Foam _ ePSea]
"e'1'Foa7n .
"L F(UHl ). w.
e Sca.
p
(18)
The FROG 2003 experiment has provicled the first reportecl
L-band emissivity measurements of artificially generated foam
on a salt sea water surface over a wide range of incidence angles and salinities at both polarizations. The ancillary clata acquired (foam thickness, air~water fraction, bubble size distribution and water coat thickness) have been usecl as inputs of
a simple two-Iayer theoretical mocle!. The intercomparison of
model outputs with the measured emissivities has allowed the
retrieval of the packing coeHicient, which is approximately h; =
0.112 for ail salinities, and,,; = 0.19 at888 = :34 psu. At Y-pol
the model and the measurements tend to disagree at higher salinities, and at H-pol at lower salinities. From 6~34 psu, the agreement is very good. In addition, the contribution of errors in the
estimation of the foam emissivity to the error of the sea surface emission are altenuated by a factor equal to the sea surface
foam coverage (~ 1% at 15 mis). The results obtained in FROG
2003, and the t'oam emissivity and coverage clerived from WISE
2001 measurements, indicate that an effective foam thickness of
~8 mm may be appropriate to moclel in a simple way the foam
contribution to sea surface emission.
Tous droits de propriété intellectuelle réservés. Reproduction, représentation ct ditllIsiofl interdites. Loi du Ol/07/92.
IEEE TRANSACTIONS ON GEOSCIENCE AND IŒMOTE SENSING, VOL. 43, NO, 5, MAY 2(0)
934
06·Apr·2003, "'u" 0,016, ",w" 0,013, SSS: 1.14 psu, SST: 14·C
2G·Apr.2003, n.,fI ' 0,011, ",w" 0.011, SSS: 6,18 psu, S5T: 19 oC
1
-.... HfonIlHrlc~:surcd
0.9 -'i'- VIo"n",'... o".
0.8 -... H,o.am.mod(!ol
-1â- Vtoam . model
0.7 ..... Hfo.1)m,'r~e
~. V'O.:JAl.ffOO
:;0,6
:>
l;'0,5
i:!:
.,
~ H'o~m.nlO.a,ur~d
Fonm purametero:
thlokllo• ., 6,11 mm
0.9 -.I/f- Vto.a-m.nlt-aJUtlKl
t~: 17.05%
:i: 10 Ilnl
'fi: 802.35
...
:~O
1
t.: :
0.175
1-1 ln
........... .
•••• •••
~.,
••
b: 10 Ilnl
0,7
r'l: 833,131101
e:
~ ..... ,
0.2
45
35
40
Incidence Angle [')
30
50
~
b)
21-Apr·2003, ""II" 0.009, ",w" 0.011, SSS: 10.49 psu, SST: 20.6 oc
-w- Vro.sm.fM:lsuted
'Jo: 22.27%
0.8 ..... H,o.am-model
...... Hto.nm."Clo
'p: 440.35 1Irn
...... V'o.itm.fre-o
... : 0,oa5
-.. Vtoam,model
0.115
,..
~O,6
................ .
~::J3~. :··~. ~. ~. =
..~..:..~..:..:..:...~..~..-..-..-..-..~..-..-..-..-..-..~..~..-..-..-...-.~...
0.2
0.1
~~5------~30~----~3~5~----~4~O------~45~----~50'
c)
Incidence Angle [")
........ H'oam-m~a;unld
0.9 -19- V'o>!TI.mmore<l
0.8 ..... H,o...lm.mOdel
-8- Vf04m.modcl
•••• HfCl,3m.rtec
..
1
'.11:
0.9
~2.21%
rp: A43.4S flm
.t: :
'
0.125
- , ...........
S
~
.;;
;;0,5 _ _ _~'P"""-'
=-;;;..---
o. . ......';;:::!t.
24-Apr·2003, n,IH. 0.008, n,IV· 0.032, 5SS: 25,5 psu, S5T: 18.8 'C
Foam paramoters:
lhlcknc ..: 16,31 mm
t
.•.. '
0.7
.': 10 'lm
rp: 536,75 ~m
...... V~?-~~~
l;:0.160.~
O6
•
·~O
0.3 ..................................... : ......................~
0.2
0.1
50
35
40
Incidence Angle [0)
30
23~Apr ..2003,
50
45
",'H= 0.011, ",WC 0.033, SSS: 37.33 psu, 55T; 15.6
<le
1
1
-v- H,o.llm.fM.uufcd
Foam paramotors:
'hICkno .. : 10,65mm
V'O.1m.model
..... Hr~m-freo
'.1: ~,59%
1-: 10 'lm
'p: GM.1111m
u"V'o.lm.hvo
... :0.160
iO.6 _.. _....... -.~_._---
............ ..
••••• _....... ~ ........ - ...... ~ ....... .
f)
30·Apr·2003, 0,11" 0,017, ".IV" 0,012,555: 33,21 psu, 55T: 18,7 oC
-~
~
0.3 .............................................................. ..
0,1
~ Hfoam.mll';l$I.rrfi'd
0,9 -'l!'- V'o.m-m... orcd
0.8 ......- Hfoam.mod~1
Hfoam-model
.ll ..... ..
45
thfeknc!l'; 1J,76 mm
',1: 9.21%
-e-
!0.5t:===~~
$............ ~...... ..
35
40
Incidence Angle l')
Foam paramotors:
-$- VrO/lm.mOdcl
..... Hf~m.trec
0,2
30
...... H'Oolm-fMtlSUrnd
-'1p- Vtoam.moluurnd
0.8
f,: 10 !Im
V'o.1m.frco
50
~
35
40
Incldenco Angle [')
30
d)
28-Apr·2003, n"u" 0.009. n,IV' 0,024, S5S: 23,24 psu, 55r: 20 oC
>'
~:
~ ................................ .
0.1
0.7
rp : .437,46 Jtm
•••• Vfrutm,ftee
~~-"'"
0,2
e)
IhlCknou: 13,0" mm
':.: 10.oit'/Î,
li: 10pm
0.7 .. u. Hfo.tm,fleil
.ll0:3 .............................................................. ..
i
~
Foam parumeters:
-... H,oam-mfJUUrOd
0,9 ......- V'""", __•• or<l<l
.-; ; 10 Iim
.ou
~
•
ft
Incidence Angle [0)
thlcknoslS; 10.99 mm
-ID- V ro.;trn-modtl
:fO,5
·2!O
W
25·Apr·2003, ""H U 0.010, n,,y" 0.016, S5S: 16,29 psu, 55T: 16.4 oC
Foam parametars:
........................
0.7
>
",0,6
........... ....... .
1
-'f'- H'~m,me')5ured
0.8 ..... H,o,am-model
1
~
.
~1
a)
0.7
: 0,060
0.2
0.1
0.9
~
" ... Vfo..ml.free
0 ,6
.. .. ..
.ll0:3 .............................................................. ..
1
H,oittn.rtu
f:r=::=·~::::~.:::;c~'~'
_~~:: ........ ..
~twr{*-e;j(..~{~~tt
thlèkne-u: 10.41 tnm
',,: n.63 1A-.
0.8 -ft- H,oam.model
"""@- Vtoam.model
:;;
~~
Foam paramotors:
~!... Vfoam-rnC;tSUrCd
0.8 .... H'oam-modol
-1)}- Vfoam-modo-I
0.7 H.' H'Mm."e-e
0.9
~0.6 ~ ...
.""tiW
~,~="'>
f!'~=,,~""""'-~;:O.5 _----=-:~- . ,.."."
v,oam:"...,
Foam parameters:
thlcJmes5: 12.73 mm
f,.: 4,14%
,'1 : 10 Jlm
'p: 457.96 Iim
,: 0,190-------i~
~0,5.!..._ _-~~
j::3::::::::::::::::::::::::..........~
..• '
..........................,
:~O,
~
w
0.2
0,3 ............................................................. '"
0.2
0,1
0.1
1
g)
30
35
40
Inclde"c. Angle r)
45
50
h)
30
35
40
Incidence Angle [')
45
50
l,
i~
Fig. II. Emissivities at (lower three plots in each panel) H·poJarization and (upper three plots in each panel) V-polarizfltion for dil'ferent salinities in the 1-37-psu
range, Dashee! lines ine!icate fOflm-free conditions. (Sol id Jine with circles) FROO measureinents. (Solid line marked \Vith triangles) Theoretical model \Vith
measurcd paramcters: [oam thickness, bubbles radii, air-water fraction, /j == 10 1r.n1, ane! optimum packing coefficicnt derived by filling the cmission modelto the
d
ri
i
l
,~
l';
measurements,
\'t
l,
Tous droits de propriété intellectuelle réservés. Reproduction, représentation ct dimlsiol1 interdites.
200S
CAMPS
1'1
al.: EMISSIVITY OF FOAM-COVrmEIl \VATER SUI(I'i\CE AT L-BAND
935
0.012
0.012
'"
0,008
0,004
o
40
50
50
10
sss
a)
[psu)
Filtered
b)
Incidence Angle [0)
lleH
=
OH (!OIIm) - eH (1",,".lr..)
sss
[l/mm]
Filtered
0,012
o
[psu]
llev
25
= ev (Io.m) -
30
35
Incidence Angle [']
ev (IOIIm./r..)
[l/mm]
0,012
0,004
0
40
30
50
50
30
20
20
10
c)
sss [psu)
10
Incidence Angle [0)
d)
sss [psu]
0 25
30
35
40
Incidence Angle [']
Fig. 12. Foam emissivity increase per foam thickness unit inmi\limeters, at 100% coverage, at H- and Y-polarizations as a function of SSS and Il;. (a) Measured
data points (H-polarization). (b) Measured data points (Y-polarization). (c) Panel (a) low-pass filtered. (d) Panel (b) low-pass filtered. Osci\lations in (a) and (b)
are due ta errors in the foam thickness measurement, which is very sma\l l'or low salinities.
ACKNOWLEDGMENT
Special th an les are given to the Institut de Recerca de Tècniques AgropeCjuàries (IRTA)-Poble Nou ciel Delta personnel
for their help cluring the experiment at their facilities ancl to
J. Fant, C. Gabarro, ancl A. Julià (Institut cie Ciències ciel Mar
CMIMA/CSIC-Barcelana) for analyzing the water samples.
The authars wish ta aclenawleclge twa ananymaus reviewers
for their comments ta improve the clarity af the text.
REFERENCES
Il] G. S. E. Lagerloer. (2000, Jan.) Report of the Third Workshop,
Salinity Sea Iee Working Group, Field Programs and Algorithms:
Sate\lites and Science, San Antonio, Texas, USA. [Online]. Availablc:
hllp:llwww.esr.org/ssiwg3/SSIWG_3.html
[21 .1. Johanncsscn el al., "Scientilic reqllirements and impact or space observations orocean salinity l'or modeling and climate stlldies," ESA, NOOl'dwiik, The Netherlands, Final Report, ESA Contract 14273/00/NLlDC,
2002.
13] C. T. Swift and R. E. Mclntosh, "Considerations l'or microwave remote
sensing of ocean-surrace salinity," IEEE nal/s. Geosci. Eleclrol/., vol.
GE-21, no. 4, pp. 480-491, Oct. 1983.
[4] A. Camps, .1. Font, M. Yall-llossera, C. Gabarrô, 1. Corbella, N. Duffo,
F. Torres, S. Blanch, A. Aguasca, R. Yillarino, L. Enrique, .1. Miranda,
J. Arenas, A. Julià, .1. Etcheto, Y. Caselles, A. Weill, J. Boutin, S. Contardo, R. Niclôs, R. Rivas, S. C. Reising, P. Wursteisen, M, Berger, and
M. Martin-Niera, "The WISE 2000 and 2001 campaigns in support or
the SMOS mission: Sea surface L-band brightness temperature observations and their application ta multiangular salinity retrieval," IEEE
Tral/s. Geosci. Remole Sel/s., vol. 42, no. 4, pp. 804-824, Apr. 2004.
[5] J. Guo, L. 'l'sang, W. Asher, K.-I-!. Ding, and C.-T. Chen, "Applications
of dense media radiative transfer theory for passive microwave remote
sensing of foam covered ocean," IEEE nal/s. Geosci. Remole Sel/s" vol.
39, no. 5, pp. 1019-1027, May 2001.
[61 L. A. Dombrovskiy and Y. Y. Raizer, "Microwave model of a two-phase
medium at the ocean surface," Izvesliya, AI/l/os. Oceanic Phys., vol. 28,
no. 8, pp. 650-656, 1992.
l7] y. Cheilly and Y. Y. Raizer, Passive MicrOlvave Re/l/ole Sel/sil/g (~r
Oceal/s. New York: Wiley, 1998.
[8] L. M, Zurk, L. Tsang, K. H. Ding, and D. P. Winebrenner, "Monte
Carlo simulations of the extinction rate or densely packed spheres with
clustered and nonclustered geometries," .1. 01'1. Soc. A/l/el:, vol. 12, pp.
1772-1781, Aug. 1995.
19] L. A. Dombrovskiy, "Ca\clllation of the thermal radiation emission of
roam on the sea surfaccw," !zvesliya, AlI/lOS. Oceal/ic Phys., vol. 15, no.
3, pp. 193-198, 1979.
[10] R. D. Peltzer and O. M. Grimn, "Stability and decay propertics or l'oam
in seawater,", NRL Memo. Rep. 5949, Apl'. 24,1987.
r Il.1 J, Wu, "Variation of whitecap covcrage witll wind stress and water temperature," J. Pllys. Oceal/ogl:, vol. 18, pp. 1448-1453, 1988.
Tous droits de propriété intellectuelle réservés. Reproduction, représentation ct dil1l1s1ol1 intcrditcs. Loi du 01/07/92.
r
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOT!'. SENSING. VOL. 43. NO. 5. MAY 20D5
112] G. S. Bordonskiy, 1. B. Vasil'kova, N. N. V. V. M. Vesclov. y. A. Militskiy. V. G. Mirovskiy, V. N. Nkitin, V. Y. Raizer, Y. B. Khapin. Y. A.
Sharkov. and V. S. Etkin, "Speclral characleristics of the emissivity of
foam formations," Izve.Hiya, AllI/os. Oeeal/ie PI/ys., vol. 14, no. 6, pp.
'164-'169, 1978.
113 J L. A. Klein and C. T. Swift, "An improved modcl for the diclectric constant of sea water at microwave frequcncics," IEEE J. Occl/I/ie EI/g., vol.
OE-2, no. l, pp. 104-111, 1977.
1141 D. Wcaire and S. Hlltzlcr, ?'lie l'hysics o/,F(){//II.1'. Oxford: Clarendon.
1999.
r15] E. C. Monahan and IVI. Lu, "ACollstically relevant bllbble assemblages
and their dependcnce on metcreological paramctcrs," IEEE J. Oceal/ie
EI/g., vol. 15, pp. 340-349, Oct. 1990.
r16] R. villarino, A. Camps, M. Vali-liossera, 1. Miranda, and .1. Arenas,
"Sea foam and sea state cfTects on the instantaneous brightncss temperatures at L-band," in Proc. 1.1'1 Resulls Worksl/Ofl 01/ WISEILOSACIEIIroS7i\RSS Ca/llflaigl/s, Nov 4-6, 2002, pp. 95-103.
Adriano Camps (S'91-A'97-M'00-SM'02) was
born in Barcelona, Spain, in 1969. He received the
Telecommunications Engineering degree and the
Ph.D. clegree in telecommunications engineering in
1992 and 1996, respectively, both l'rom the Polytechnic University of Catalonia (UPC), Barcelona,
Spain.
From 1991 to 1992, he was with the ENS cles
Télécommunications de Bretagne, Bretagne, France,
\Vith an Erasmus Fellowship. In 1993, he joined
the Electromagnetics and Photonics Engineering
group, at the Department of Signal TheOI"y and Communications, UPC, as
an Assistant Professor, and since 1997 as an Associate Professor. In 1999,
he was on sabbatical leave at the Microwave Remote Sensing Laboratory,
University of Massachusetts, Amherst. His research interests are mierowave
remote sensing, with special emphasis in microwave racliometry by aperture
synthesis techniques. He has performed numerous studies within the frame of
European Space Agency SMOS Earth Explorer Mission. He is an Associate
Editor of Radio Sciel/ce.
Dr. Camps reccived the second national award of university studies in
1993, the INDRA award of the Spanish Association of Telecommunication
Engineering to the best Ph.D. in 1997, the extraordinary Ph.D. award at the
Universitat Politècnica de Catalunya in 1999, the First Duran Farell Award
and the Ciudad de Barcelona Award, in 2000 and 2001, respectively, both for
Technology Transfer, in 2002. the Research Distinction of the Generalitat de
Catalunya for contributions to microwave passive remote sensing, in 2003,
the Premi Nacional de Telecomunicacions (Gencralitat de Catalunya) with the
members of the Electromagnetics and Photonics Engineering group, ancl in
2004 a EURYI (European Young Investigator) Award. Also, as a member of the
Microwave Radiometry Group at UPC, he received in 2000, 200 l, ancl 2004 the
Ist Duran Farell and the Ciudad dc Barcclona Awards for Technology Transfeer, and the Salvà i Campillo Award of the Telecommunications (Engineering
College of Catalonia) to the mast innovative research project. He \Vas Chair of
Cal '01. He is editor of the IEEE Geosciel/ce al/d Relllole Sel/sil/g News/el/el'
and President-Founder of the IEEE Gcosciencc and Remote Sensing Society
Spain Chapter.
Mcrcè vall-llossera (M'99) rcccived the Senior
Telecommunication Engineer and the Doctor
Telecommunication Engineering dcgrecs in 1990
and 1994, respectively, both l'rom the Polytechnic
University of Catalonia (UPC), Barcclona, Spain.
She has been lecturing and doing rescarch at the
Departlllcnt of Signal Theory and COllllllunications,
UPC from 1990 until 1997 as an Assistant Prof'cssor
and l'roll! 1997 until present as an i\ssoeiatc ProJ'essor. She spent a sabbatical year in Montreal with
the seholarship of the "Programme Québécois de
Bourses cl' cxcellence" (1996-1997): "Stages de Formation postdoctorale au
Québec pour jcunes diplômés étrangers." Hel' researeh interests include numerical methods in clectromagnctism, microwavc radiomelry, antcnna analysis.
ancl design. Current!y, her rcseareh is mainly rclated to the study of numcrical
methods applied ta the sea surface emissivity and their characterization at
L-band and the MIRAS/SMOS project.
Dr. Vali-liossera, along with the ather members of the racliometry group at
UPC, was awardecl the "9th Edition of the Salvà i Campillo Award" in 2004, the
"Primel' Premio Duran Farell de Investigaciôn Tecnol6gica" in 2002, and the
"Primer Premio Ciutat cie Barcelona d'lnvestigaciô Tecnolàgica" in 2001.
Ramon Villarino (S'04) recciveclthe Ingeniero and
Ph.D. degrees in telecommunications from the Polytechnic University of Catalonia (UPC), Barcelon<l
Spain, in 2000 and 2004, respectivcly.
I-Ie has participated in the development and construction of an L-bancl radiometer and in the tlVa
WISE field experiments sponsorecl by ESA cluring
2000-2002, in tbe FROG 2003 fielcl experiment to
cletermine the emissivity offaam at L-band, and in the
MOUSE 2004 experiment carried out at JRC-Ispra
to determine the emissivity of clif'ferent types of soils.
Nicolas Reul reccived the B.S. dcgree in marine sciences engineering l'rom Toulon University, La Garde
Cedex, France, and the Ph.D. clegree in physics (fluid
mechanies) l'rom the University Aix-Marseille Il,
Marseille, France, in 1993 and 1998, respcctively.
From 1999 to 200 l, he was a Postdoctoral Resc archer at the Applied Marine Physics Department,
team of Pral'. M. Donelan, Rasensticl Schaol of Ma- .
rine and Atmospheric Science, University of Miami.
Since 200 l, he has been a Permanent Rsearcher at
the Département Océanographie Spatiale, Institut
Francais de Recherche et cI'Exploitation cie la Mer, responsible for the activities
conccllling the future SMOS satellite mission. The focus of his research program is to improve understanding of the physical processes at air-sea interülCC
and passive/active remote sensing of the ocean surface. I-Ie has cxperiencc in
applied mathematics, physical oceanography, clectromagnetic wave theOl'Y,
and its application to ocean remote sensing. He is presenlly a member of the
ESA/SMOS Science Advisory Group.
Tous droits de propriété intellectuelle réservés. Reproduction, représentation cl diflllsion interdites. Loi du 0 l/07/92.
CAMPS
fi
{/I.: EMISSIVITY OF FOMvl-COVERED WATER SURI'ACE AT I.-\JAND
Bertrand Chapron received the B.Eng. degree
l'rom Institut National Polytechnique de Grenoble,
Grenoble, France, and the Ph.D. clegrec in physics
(fluid mcchanics) l'rom the University or Aix-Marseille II, Marseille, France, in 1984 ancl 1988,
respeetively.
He spent three l'cars as a Pastcloctoral Research
Associate at the NASA Goddard Space Flight
Centcr, Wall,,!,s Flight Facility. He has cxpcricncc
in applicd mathcmatics, physical occanography,
elcctromagnetic wave theO/'y, and its application
to ocean remote sensing. He is currcntly a Rcsearch Scientist and is Heacl
or the Spatial Oceanography group at the Institut Francais de Recherche ct
d'Exploitation de la Mer, Responsible or the Centre ERS Archivage ct Traitement. He is a Collaborator or Principal [nvestigator in seve rai ESA (ENYISAT,
Global Navigation Satellite System), NASA, and CNES (TOpEX/pOSElDON,
JASON) prajeets. He is Coresponsible (with H. Johnsen, NORUT) or the
ENYISAT ASAR-wave mocle algorithms and scientific preparation l'or the
ENYISAT wind and wave proclucts.
937
Jorge José Miranda (S'04) was !Jorn in Las Palmas
de Gran Canaria, Canary Islands, Spain, in 1973.
Hc received the Telecollllllunication Technical
Engineering degree l'rom Las Palmas de Gl1\n
Canari a University (ULpGC), Spain, and the Senior Telecomlllunication Engineering degree rrom
the rolytechnic Univcrsity or Catalonia (UPC),
Barcclona, Spain, in 1996 and 200 l, respectively. He
is currcntly pursuing the Ph.!). dcgrcc in te1CCOIlllI1unication engineering l'rolll Ul'c. His ph.D. thcsis
is rocused on the sea surrace emissivity at L-band.
It includes the devclopment of new Illodcls for clctermining the sea surrace
l'Oughness and the comparison and improvelllent or existing emissivity modcls.
From 1996 to 1999, he was \Vith thc Remote Scnsing and RaclaI' Laboratory
(EUITT-ULPGC), where collaborated in satellite image proccssing. [n 1999, he
joined the Signal Theory and ComIl1unieations Department, UPC. Since 2002,
he has been an Assistant Proressor at the UPC.
Ignasi COl'bella (M'99) received the Teleeommunic~llions Engineering and Dr.Eng. clegrees, both
l'rom Universitat Politècnica de Catalunya (UPC),
Barcelona, Spain, in 1977 and 1983, respectively.
[n 1976, he joined the School of Telecommunication Engineering in UPC as a Research Assistant in
the Microwave Laboratory, where he worked on passive microwave integrated circuit design ancl characterization. During 1979, he worked at Thomson-
Roberto Sabia (S'l)3) \Vas born in Naples, ltaly,
on Novcmber 13. 1975. He gracluated (cum laude)
in marine environ mental sciences (five-year course
study), curriculum in oceanography, rrom the Università degli Studi di Napoli "Parthenope," Naplcs,
in 2002. He is is cUITently pursuing a ph.D. degrec in
signal theory and communications at the Universitat
Politécnica de Catalunya (UPC), Barcelona, Spain.
ln 2002, he was recipient of a grant l'rom the
CSF, Paris, France, on I11icrowave oscillators design.
In 1982, he became an Assistant Professor at urc,
an Assaciate ProfessaI' in 1986, and a Full ProfessaI' in 1993. He is currently
teaching microwaves at the undergraduate level in urc and has designed ancl
taught graduate courses on nonlinear microwave circuits. During the school
year 1998-1999, he worked at NOAA/Environmental Technology Laboratory,
Boulder, CO, as a Guest Researcher, developing methods for radiometer calibration and data analysis. His research work in the Department of Signal Them'y
and Communications, UPC includes microwave airborne and satellite radiometry and micrawave system design.
Nuria Duffo (S'91-!V['99) received the Telecommunication Engineer degree from the Polytechnic
University of Catalonia (UPC), Barcelona, Spain,
and the Doctor in Telecommunication Engineering
l'rom UPC, in 1990 and 1996, respectively.
Since 1997, she has been an Associate Professor
at Upc. Hel' current research interests are numerical
methods in electromagnetics, microwave radiometry,
antcnna anal l'sis, and design.
Francesc TOlTCS (M'96) received the Ingeniero
and Doctor Ingeniera degrees in telccommunication
engineering l'rom the Polytechnic University of
Catalonia (UPC), Barcelona, Spain, in. 1988 and
1992, respectively.
In 1988-1989, he was a Research Assistant in the
RF System Division, European Space Agency, Noordwijk, The Netherlands, devoted ta microwave device testing and characterization. ln 1989, he joined
the Antenna-lvIicrowave-Raciar group, ure, where
he is currently an Associate Professor. His main research interests are focused on the design andtesting of microwave systems and
subsystems. He is currently engaged in research on interferametric radiometers
devoted to earth observation.
Microwave Remote Sensing Laboratory . Univcrsità
dcgli Studi di Napoli "Parthenope," on sea surface
scaltering models. In 2003, he joined the Departmcnt of Signal Them'y and
Communications, UPC. In 2003, he participated in FROG 2003 and REFLEX
2003 measurements campaigns. He is actua]]y worldng on a research contract within the ESA's SlvIOS project. His main research interests deal \Vith
mÎcrowave radiometry and sea surface salinity retrieval.
Alcssandra Moncrris (S'03) was born in Valencia, Spain, in 1977. She received the degree in
telecommunication engineering rrom the Unil'ersitat
politècnica de Yalència (UPY), Yalencia, Spain, in
2001. She she is cUITently pursuing the ph.D. degree
at the Universitat politècnica de Catalunya (UPC),
Barcelona, Spain.
She was with the Group or Microwave Applications l'rom UPY developing numerical methods for
the electromagnctic anal l'sis or waveguide-based
devices. In 2002, she joined the Department or
Signal Theory and Communications, upc. Hel' main research interests deal
with surface soil moisture retrieval algorithms l'rom brightness temperaturc
Ineasurements.
Rubén RodrîgllCZ, photograph and biography not available at the time or
publication.