LO calibration frequency impact on ocean salinity

SMOS QWG-5, 30 May – 1 June 2011, ESRIN

LO calibration frequency impact on ocean salinity

V. González, C. Gabarró, J. Martínez,

M. Portabella, J. Font and BEC–TEAMSMOS Barcelona Expert CentrePg. Marítim de la Barceloneta 37-49, Barcelona SPAINE-mail: [email protected]: www.smos-bec.icm.csic.es

SMOS QWG-5, 30 May – 1 June 2011, ESRIN 2

Decimation study using MTS

Assessment of the LO calibration frequency impact on TB.

Data acquired at 2min and decimated at 6min and 10min.

Descending orbit over Pacific: 20100325T040821.

All the comparisons are relative to the 2min LO calibration frequency data.

Only AF-FOV grid points

SMOS QWG-5, 30 May – 1 June 2011, ESRIN 3

TB maps @ 38-40º in the AF-FOV

TBH10min-TBH2min TBV10min-TBV2min

[K]

MIRAS-TS

SMOS QWG-5, 30 May – 1 June 2011, ESRIN

L1pp: TB maps @ 38-40º in the AF-FOV

L1PP_3.5

SMOS QWG-5, 30 May – 1 June 2011, ESRIN 5

TB_10min-TB_2min AF-FOV

TB10min – TB2min

H-pol

V-pol

L1PP_3.5MIRAS-TS

SMOS QWG-5, 30 May – 1 June 2011, ESRIN 6

TBH6min-TBH2min TBV6min-TBV2min

TB maps @ 38-40º in the AF-FOV

MIRAS-TS

SMOS QWG-5, 30 May – 1 June 2011, ESRIN 7

TB_6min-TB_2min AF-FOVTB6min – TB2min

H-pol

V-pol

MIRAS-TS L1PP_3.5

WARNING: different K scale

SMOS QWG-5, 30 May – 1 June 2011, ESRIN 8

SSS 24-25 march 2010 57 orb

SSS 10min – 2min

SMOS QWG-5, 30 May – 1 June 2011, ESRIN 9

SSS 24-25 march 2010 57 orb

SSS 6min – 2min

SMOS QWG-5, 30 May – 1 June 2011, ESRIN 10

Conclusions

The present LO calibration frequency has a clear impact on TB (larger in V-pol) and SSS.

TB differences between LO-10min and LO-2min have a global STD of 0.47 K (1.1K in N. Atlantic, 0.21 K in Pacific)

Differences in SSS (filtered by good retrieval flag): global STD 0.48 psu (0.94 psu in N. Atlantic, 0.20 psu in Pacific)

Accuracy requirement for L2 SSS is 1.2 psu. We don’t know yet the actual retrieval error, but the low calibration freq can represent in average 0.5 psu of additional error!!

SMOS SSS requirements will likely never be met with LO being calibrated at 10 min intervals

SMOS QWG-5, 30 May – 1 June 2011, ESRIN 11

On statistics

During 10 minutes 500 SMOS snapshots are acquired

Calibrating LO every 10 min means loosing up to 8 snapshots, then the error in building an L3 map by averaging L2 products (wrt no calibration) increases by sqrt (500)/sqrt(492) = 1.008

Calibrating every 2 min, 40 snapshots maximum are lost, the L3 error increases by sqrt(500)/sqrt(460) = 1.043

Going back to one orbit, the increase of usable snapshots @10 min compared to @2 min implies an error reduction of sqrt(492)/ sqrt(460) = 1.034

This 3.4% reduction on the required SSS error means 0.04 psu

We showed that calibrating @2 min instead of @10 min implies an error reduction by 0.48 psu in global average

SMOS QWG-5, 30 May – 1 June 2011, ESRIN

SMOS QWG-5, 30 May – 1 June 2011, ESRIN 13

Histogram SSS WORLD

Global L2OS output

When filtering byFg_poor_retrieval=1

STD (SSS10-SSS2)= 0.48 psu

SMOS QWG-5, 30 May – 1 June 2011, ESRIN 14

SSS differences: best & worst cases

% SSS_Nmin -SSS_2min> 0.2 psu

SSS_Nmin -SSS_2min> 0.5 psu

SSS_Nmin -SSS_2min>1 psu

N=10min 30.67%

12.03%

9.59%

2.87%

2.07%

0.55%

N=6min 9.67%

1.85%

1.42%

0.05%

0.22%

0.01%

N=4min 2.85%

0.42%

0.38%

0.03%

0.11%

0.01%

Lat < -50º PACIFICLat=[-50,10]Lon=[-180,-110]

SMOS QWG-5, 30 May – 1 June 2011, ESRIN 15

TB maps @ 23-25º in the AF-FOV

TBH10min-TBH2min TBV10min-TBV2min

[K]

MIRAS-TS

SMOS QWG-5, 30 May – 1 June 2011, ESRIN

L1pp: TB maps @ 23-25º in the AF-FOV

L1PP_3.5

SMOS QWG-5, 30 May – 1 June 2011, ESRIN

L1pp: TB maps @ 38-40º in the EAF-FOV

L1PP_3.5 EAF

SMOS QWG-5, 30 May – 1 June 2011, ESRIN

L1PP: TB differences for 1 orbit EAF-FOV

L1PP_3.5 EAF