www.argans.co.uk

SMOS L2 Ocean Salinity

Roughness model deltas

v550 OTTs made fromREPR v504 L1c

30 January 2012

ARGANS

www.argans.co.uk

SMOS L2 Ocean Salinity

v550 OTTs made from REPR v504 L1c

OTT1-OTT2: -0.56 K OTT1-OTT3 +0.33 K

www.argans.co.uk

SMOS L2 Ocean Salinity

v550 OTTs made from REPR v504 L1c

OTT1-OTT2: -0.57 K OTT1-OTT3 +0.30 K

www.argans.co.uk

SMOS L2 Ocean Salinity

v550 OTTs made from REPR v504 L1c

OTT1-OTT2: -0.22 K OTT1-OTT3 +0.18 K

www.argans.co.uk

SMOS L2 Ocean Salinity

v550 OTTs made from REPR v504 L1c

OTT1-OTT2: -0.22 K OTT1-OTT3 +0.19 K

www.argans.co.uk

SMOS L2 Ocean Salinity

v550 OTTs made from REPR v504 L1c

OTT1-OTT2: -0.45 K OTT1-OTT3 +0.23 K

www.argans.co.uk

SMOS L2 Ocean Salinity

v550 OTTs made from REPR v504 L1c

OTT1-OTT2: -0.45 K OTT1-OTT3 +0.22 K

www.argans.co.uk

SMOS L2 Ocean Salinity

v550 OTTs made from REPR v504 L1c

OTT1-OTT2: -0.44 K OTT1-OTT3 +0.23 K

www.argans.co.uk

SMOS L2 Ocean Salinity

v550 OTTs made from REPR v504 L1c

OTT1-OTT2: -0.43 K OTT1-OTT3 +0.20 K

www.argans.co.uk

SMOS L2 Ocean Salinity

Conclusions:

• Very small differences between ascending & descending OTT1/2/3 deltas

• Models are biased, especially in HH & in AFFOV:

OTT1-OTT2: -0.56 K OTT1-OTT3 +0.33 K OTT2-OTT3 +0.89 K

• Maybe deltas in Stokes 3 & 4 too – not tested

• Biases are corrected by OTTs

• Which model is nearer to ‘ground-truth’?

www.argans.co.uk

SMOS L2 Ocean Salinity

delta OTT1 – OTT2, descending 20111109 (event 54)

Why strong delta at top edge of VV?

www.argans.co.uk

SMOS L2 Ocean Salinity

delta OTT1 – OTT2, ascending 20111109 (event 54)

Why strong delta at top edge of VV?

www.argans.co.uk

SMOS L2 Ocean Salinity

delta OTT1 – OTT2, ascending 20100704 (event 21)

Why strong delta at top edge of VV, independent of sun, ascending/descending, year, etc?

www.argans.co.uk

SMOS L2 Ocean Salinity

delta OTT3 – OTT2, ascending 20100704 (event 21)

Delta at top edge in VV seems to come from model 2?

www.argans.co.uk

SMOS L2 Ocean Salinity

delta OTT1 – OTT3, ascending 20100704 (event 21)

Strong gradient in HH?

www.argans.co.uk

SMOS L2 Ocean Salinity

delta OTT2 – OTT3, ascending 20100704 (event 21)

Strong gradient in HH seems to come from model 3?

www.argans.co.uk

SMOS L2 Ocean Salinity

delta OTT1 – OTT2, cross-pol, ascending 20100704 (event 21)

Strange pattern in Stokes 3?

www.argans.co.uk

SMOS L2 Ocean Salinity

delta OTT1 – OTT2, cross-pol, ascending 20100704 (event 21)

Strange pattern in Stokes 3, hhv & vvh?

www.argans.co.uk

SMOS L2 Ocean Salinity

delta OTT3 – OTT2, cross-pol, ascending 20100704 (event 21)

Strange pattern in Stokes 3, from model 2?

www.argans.co.uk

SMOS L2 Ocean Salinity

Reprocessed v550 L2OS

20110704T014904, filtered by poor(retrieval or geophysical)Problem in SSS1 at edges in EAFFOV: wind/TEC/???

www.argans.co.uk

SMOS L2 Ocean Salinity

Reprocessed v550 L2OS

20110704T141932, filtered by poor(retrieval or geophysical)Problem in SSS1 at edges in EAFFOV: wind?

www.argans.co.uk

SMOS L2 Ocean Salinity

Conclusions:

• Does delta OTT1/2/3 indicate any algorithm issues?

• Does delta SSS1/2/3 in EAFFOV indicate problem with WS in SSS1?

• Why only in EAFFOV?

• Also at leading/trailing edges of snapshot inside swath, but hidden?

• How can we work towards unification of SSS1/2/3?