SMOS Validation Rehearsal Campaign Workshop, 18-19/11/2008, Noordwijkerhout, The Netherlands SMOS Validation Rehearsal Campaign Mediterranean flights C.

SMOS Validation Rehearsal Campaign Workshop, 18-19/11/2008, Noordwijkerhout, The Netherlands

SMOS Validation Rehearsal Campaign Mediterranean flights

C. Gabarró, M. Talone, J. Font

SMOS Barcelona Expert CentrePg. Marítim de la Barceloneta 37-49, Barcelona (Spain)E-mail: [email protected]: www.smos-bec.icm.csic.es

SMOS Rehearsal Campaign 2

Flight 19 -04-08 from 18:20 to 21h(flight back on 3-05-08 not analysed)

Marseille

Valencia

Barcelona

Buoy 3

Buoy 1-2

Mediterranean flights

SMOS Rehearsal Campaign 3

In situ data: oceanographic buoys

In situ data report and data files available through ESA

Buoys 1-2: ICM specific deploymentclose to Casablanca oil platform

Buoy 3: offshore permanent network (PE)

Buoy 4: coastal permanent network (SMC)

SMOS Rehearsal Campaign 4

Closest data to overflight

19/04/08 20:20

Buoys 1-2

T = 10'

SSS = 38,09 -> - 0.6 m

SST = 14,50 ºC

WS = 4,16 m/s

SWH=0.7 m

Buoy 3

T=60'

SSS = 38,1 -> - 3 m

SST = 14,5 ºC

WS = 5,8 m/s

In situ data: oceanographic buoys

SMOS Rehearsal Campaign 5

Δ Resolution = 1/16 º

Time = 12h

Marseille

Valencia

Barcelona

Marseille

Valencia

Barcelona

Mediterranean Forecasting System, Univ. Bologna (MOON)

In situ conditions: numerical model

SMOS Rehearsal Campaign 6

Scatterometer winds 19-4-2008

provided by KNMI, Netherlands

In situ conditions: Wind speed

ASCAT QuikSCAT

ERS

SMOS Rehearsal Campaign 7

Only QuikSCAT covers the whole area

of interest :

R = 25Km18:18 h

19/04/2008

In situ conditions: Wind speed

ASCAT QuikSCAT

ERS

SMOS Rehearsal Campaign 8

Flight 19 -04-08 from 18:20 to 21h

Analysis by areas:

1 st: Casablanca buoys

2n: Marseille- Buoys

Marseille

Valencia

Barcelona

Buoy 3

Buoy 1-2

Analysis

RFI - not used

SMOS Rehearsal Campaign 9

Simplified emissivity model

Klein & Swift for sea water dielectric constant

Linear fit of Hollinger measurements for wind effect

Linear approximation for the atmospheric and constant external sources contribution

Average of 0.1 s and points with RFI not used in average.

KT

Temphh

hhT

T

eemeTTeemTeTT

eemeTTeemTeTT

sky

surf

up

dn

VskyupVdnVsurfV

HskyupHdnHsurfH

7.3

cos

0377.0861.220001.0

cos

03.04128.0

)cos(

1.2

11

11

2

coscos_

coscos_

coscos

coscos

SMOS Rehearsal Campaign 10

Flight between buoys

MFSTEP MFSTEP

Bet

wee

n b

uoys

SMOS Rehearsal Campaign 11

EMIRAD beam center locationB

etw

een

buo

ys

Hor

n =

0º

Hor

n =

40º

Roll Pitch

SMOS Rehearsal Campaign 12

TB measured and modelled, horn=0ºB

etw

een

buo

ys

SMOS Rehearsal Campaign 13

Bet

wee

n b

uoys

TB measured and modelled, horn=40º

SMOS Rehearsal Campaign 14

Bias of 5 K -> Problem on calibration TY aft? Too simple model?

TB measured - modelledB

etw

een

buo

ys

SMOS Rehearsal Campaign 15

TB measured - modelledB

etw

een

buo

ys

Galatic noise corrected with gal maps + flat sea - error

SMOS Rehearsal Campaign 16

Small differences between models:Hollinger 1971Gabarró 2004

Models comparisonB

etw

een

buo

ys Tb modelled using in situ

SST, SSS, SWH, WS

SMOS Rehearsal Campaign 17

Flight from Marseille to buoys

SMOS Rehearsal Campaign 18

Only cleanest area used for

analysis (3000 measurements)

In situ dataM

arse

ille

to b

uoys

RFI percentages:

H - aft = 12.66%V - aft = 4.33%H - nadir = 3.92%V - nadir = 1.40 %

SMOS Rehearsal Campaign 19

TB measured vs modelled

Again bias on TY aft

Mar

seill

e to

buo

ys

SMOS Rehearsal Campaign 20

TB measured vs modelled M

arse

ille

to b

uoys

Galatic noise corrected with gal maps + flat sea - error ?

SMOS Rehearsal Campaign 21

Theoretically: at surface reference frame, for θ 0 -> TH=TV

TB measured for θ0

Calibration problems?

Only nadir horn used

0.5 K

Mar

seill

e to

buo

ys

SMOS Rehearsal Campaign 22

Conclusions – EMIRAD over sea

An important bias appears on the aft V channel

antenna (~ 5 K) -> galactic noise??? Not accordance between TH & TV on θ0 (~ 0.5 K)

Important RFI and noise are detected (near Valencia)

Measured TB variability fits with modeled variability

WE NEED TO USE A REALISTIC MODELING OF

GALACTIC NOISE –> use a roughness model (not flat

sea) -> review which output from TRAP to use.

SMOS Rehearsal Campaign 23

INTERFEROMETRIC RADIOMETER

DATA – HUT 2D

Flights over the Gulf of Finland

August, 13th and 15th 2007

SMOS Rehearsal Campaign 24

HUT-2D Data Processing

Two series of flights over the Gulf of Finland

August, 13, 2007 (20 flights) August, 15, 2007 (22 flights)

SMOS Rehearsal Campaign 25

Approximations and Models

• Klein & Swift for the sea water dielectric constant• Linear fit of Hollinger Measurements for Wind Effect • Linear approximation for the Atmospheric Contribution

• Assuming Apparent Temperature = Brightness Temperature

KT

Temphh

hhT

T

eemeTTeemTeTT

eemeTTeemTeTT

sky

surf

up

dn

VskyupVdnVsurfV

HskyupHdnHsurfH

7.3

cos

0377.0861.220001.0

cos

03.04128.0

)cos(

1.2

11

11

2

coscos_

coscos_

coscos

coscos

Ulaby F., Moore R., Fung A. - Microwave Remote Sensing Active and Passive - ed. Addison-Wesley Publishing Company

HUT-2D Data Processing

SMOS Rehearsal Campaign 26

In-situ (vessel measured) SSS

Nearest neighbour approximationalong the ground-track for SSS and SST

Nearest neighbour (in time and space) QuikSCAT data for Wind Speed [KNMI]

3.73 m/s 13/08/2007 at 17.583.46 m/s 15/08/2007 at 17.06

Direct measurements

In-situ (vessel measured) SST

Direct measurements

SMOS Rehearsal Campaign 27

Apparent Temperature at X-Pol

Apparent Temperature at Y-Pol

Second flight on August, 13

SMOS Rehearsal Campaign 28

Difference between Apparent Temperature and Modeled Brightness Temperature in the Earth reference frame

The antenna pattern must be included in the processing to transform the apparent temperature in measured brightness temperature!

Kainulainen, J., Rautiainen, K., Hallikainen, M., Takala, M - Radiometric performance of interferometric synthetic aperture radiometer HUT-2D IEEE International Geoscience and Remote Sensing Symposium, 2007. IGARSS 2007.

SMOS Rehearsal Campaign 29

In order to make the SSS retrieval possible and reliable:

• Antenna pattern must be considered

• A previous selection of the measurements must be performed

• Better models for atmospheric and galactic contribution must be used

SMOS Barcelona Expert Centre (SMOS-BEC)Pg. Marítim de la Barceloneta 37-49, E-08003 Barcelona, SPAINTel. (+34) 93 230 95 00; Fax. (+34) 93 230 95 55URL: www.smos-bec.icm.csic.es