Sea Surface Salinity as Measured by SMOS and by

Surface Autonomous Drifters:Impact of Rain

J. Boutin, N. Martin, X. Yin, G. Reverdin, S. Morrisset

LOCEAN, UMR CNRS/UPMC/IRD, Paris, France

+ collaborations with French GLOSCAL SMOS Cal/Val participants (IFREMER, Meteo-France, LEGOS)

and SMOS ESA Expert Support laboratories (ICM/CSIC, LOS/IFREMER, ARGANS-st, CLS, ACRI-st)

Motivations

=> L-band Radiometry (Soil Moisture and Ocean Salinity, SMOS)

• SSSsmos-SSSargo in the tropical Pacific are 0.1 lower than in the subtropical Atlantic. Could it be an effect of rain?

• Skin depth of L-band radiometric signal is 1cm whereas ARGO SSS are measured at typically 5m depth: is rain effect detectable on SMOS-ARGO comparisons?

SMOS SSS - July 2010 L1 & L2 v500 Ascending swaths (center swath, 3m/s<WS<12m/s)

SMOS SSS

ARGO objective analysis (Gaillard et al.)

Comparison with ARGO: SMOS averaged over 10 days-

100km around ARGO

ITCZ trop Pac NSSSsmos-SSSargo=-0.05+/-0.40N=248

Subtropical Atl N (SPURS region)SSSsmos-SSSargo=0.04+/-0.22N=76

SMOS SSS & Rain colocation

• SMOS SSS - 40km resolution retrieved from ~150 Tbs measured at various incidence angles, 2 polarisations=> SSS and ECMWF adjusted wind speed

• Rain Rate deduced from SSMI F16 and F17 (RemSS version 7); 32km spatial resolution

• Colocation of SMOS SSS and SSMI Rain Rate: Maximum of RR (within -5hours and +1hour to SMOS time) and falling in the SMOS ISEA grid point

• Statistics of SSSsmos-SSSargo depending on Maximum RR within -5h/+1h of SMOS SSS

V500 Sun Off - OTT asc - July 2010

SSMI F17 Rain Rate (mm/hr) in July 2010

Rain Rate <1mm/hrN=9419Mean diff: 0.05Std_Diff=0.56Skewness=0.014

Rain Rate >1mm/hrN=3260Mean diff: -0.23Std_Diff=0.75Skewness=-0.708

All Rain RateN=12679Mean diff: -0.02Std_Diff=0.62Skewness=-0.47

SSSsmos-SSSargo Boutin et al. 2011

r=0.47

-0.2 / mm/hr

V500 Sun Off - OTT asc - July 2010

Comparison with ARGO: SMOS averaged over 10 days-100km around ARGO

Subtropical Atl NSSSsmos-SSSargo=0.04+/-0.22Npairs=76Nsmos/argo=64

ITCZ trop Pac NAll Rain:SSSsmos-SSSargo=-0.05+/-0.40N=248

No RainSSSsmos-SSSargo=0.06+/-0.39Npairs=219; Nsmos/argo=28

Ascending orbits (center of orbit; OTT asc) 3<WS<12m/sSMOS SSS averaged around ARGO (+/-50km; +/-10days)

Conclusions & Perspectives

• Precision of 10days-100km SMOS SSS in a warm non rainy

region ~ 0.2 (ascending orbits only in July v500)

• Monthly SMOS SSS (1cm) minus ARGO Salinity (~5m) in the

tropical Pacific (average over 5°N-15°N; 110°W-180°W) 0.1

lower than in the Atlantic subtropics because of rain

freshening (ΔSSS/RRmax=-0.2 / mm/hr)

Perspectives

• Redo/extend this analysis in time with reprocessed SSS

• Look at angular polarised signature of rain

• Rainy drifters - SMOS SSS colocations once SMOS

reprocessed archive available (challenging: need for a lot of

colocated events; punctual versus integrated SSS…)

Small effect of rain in the atmosphere (Rayleigh scattering) at L-band (J Schulz,

2002; Wentz, 2005)

Wentz, 2005

In warm region, RR ~ 10mm/hr =>rayleigh scattering ~ 0.2K => SSS bias ~ 0.14pss at 10mm/hr : we see a much larger effect

Could it be an artefact of L-band radiometric measurements?

• At L-band, weak impact of atmosphere (RR<10mm/hr); • Roughness effect on SMOS rainy measurements?

Differences between ECMWF and SMOS retrieved wind speed are not correlated with RR

(N.B.: 1psu ~0.7K ~3.5m/s)

Motivations (2)

• Surface in situ salinity data between 15cm and 1m depth evidence salinity gradients near the surface in case of rain (TOGA COARE, Soloviev and Lucas, 1997; next slides); however these observations are punctual

• Most in situ measurements (ship; ARGO) are made at a few meters depth (=> they miss near-surface gradients)

• Fresh-lenses => thin surface layer isolated from the water below (stratification; air-sea exchanges)

• Gas properties (e.g. CO2 solubility) depend on salinity

In situ measurements :large salinity stratification in case of rain

Precipitation rate (mm/hr)Precipitation rate (mm/hr) salinity at 1 m depthsalinity at 1 m depthWind speed (m/s)Wind speed (m/s) salinity at 5 m depthsalinity at 5 m depth

salinity at 10 m depthsalinity at 10 m depth

1,1 pss

1,4 pss

1,5 pss

TAO buoy 3.5°N – 95°W, Pacifique Est60

0 31

34

28/01/03 03/02/03

Hénocq et al., JAOT, 2010

Autonomous drifter (~50cm depth)

Example of SSS freshening in Atlantic ITCZ

Float

ARGO at +/-100km, +/-2days

See poster Morrisset, et al; Reverdin et al. JGR 2011, in revision

• Pacific Gyre (SBE 37 SI)

SSS rain-freshening temporal evolution as seen by 60 drifters in the tropical Oceans

Figure 7: Average cycle of salinity (upper panel) among 60 salinity drop events (relative to a common time of beginning of event). Individual records are shifted to a common salinity value at the initial drop time and the magnitude of the drop is adjusted to the mean drop. The average is plotted as well as individual events. The associated average reported rainfall (mm/hour) is plotted in the lower panel by 2-hour average, as well as the individual values at the exact time of reports.

SSS temporal evolution after a freshening event (time 0)

Satellite Rain Rates (SSM/I, TMI, AMSRE (www.ssmi.com)) colocated with floats SSS

0hr 4hr 15hr

35.1

34.7

0hr 4hr 15hr

10

0

See poster Morrisset, et al; Reverdin et al. JGR 2011, in revision

Vertical gradients 15cm – 45 cm17 events SVP-BS / Surplas

Vertical gradients 15cm – 45 cm17 events SVP-BS / Surplas

SURPLAS tied to a SVP-BS drifter (CAROLS2010 cruise, Gulf of Biscay)

See poster Morrisset, et al; Reverdin et al. JGR 2011, in revision