SMOS L2 Ocean Salinity Level 2 Ocean Salinity L2OS planning 2 July 2014 ARGANS & SMOS L2OS ESL 1.

SMOS L2 Ocean Salinity

www.argans.co.uk

Level 2 Ocean Salinity

L2OS planning

2 July 2014

ARGANS & SMOS L2OS ESL

1

SMOS L2 Ocean Salinity

www.argans.co.uk

L2OS evolution for metrics: AUX_DTBXY

• To support L1 & L2OS metrics, AUX_DTBXY needs extending to include snapshot statistics for L1c TBs, forward model TBs, and OTT correction TBs: for all polarisations, all 12 FOV sub-zones, & all 3 forward models.

• To support A3TEC metrics, AUX_DTBXY needs extending to include retrieved TEC and associated data.

• No changes to algorithm, configuration, or science products = low risk.

• AUX_DTBXY to increase in size from max ~10Mb to ~13Mb

• We recommend L2OS v620 is delivered September 2014 and used for reprocessing, to generate AUX_DTBXY as a reference data set for L1 & L2OS metrics, and A3TEC analysis.

• Summer 2014 delivery (late – due April) will be L1 Metrics Plan (including ADD) based on draft TRUM (Tools Release User Manual) & PPSR.

• L2OS v620 delivered September 2014 with Tools for L1 Metrics2

SMOS L2 Ocean Salinity

www.argans.co.uk

L2OS evolution for metrics: AUX_DTBXY

3

n x geographical regions:region ID, snap start/stop time/ID3 x forward models

8 x polarisationsott statsdeltaTB[xi,eta]std(deltaTB)[xi,eta]count(deltaTB)[xi,eta]flags[xi,eta]

for each snapshot in region:ID, lat/longpolarisation flagsdeltaTB (0.3 circle)

V612 v620n x geographical regions:

region ID, snap start/stop time/ID3 x forward models

8 x polarisationsott statsdeltaTB[xi,eta]std(deltaTB)[xi,eta]count(deltaTB)[xi,eta]flags[xi,eta]

for each snapshot in region:ID, lat/longpolarisation flagsfor 12 FOV sub-zones

for 3 polarisationsL1cTBfor 3 forward models

forward model TBOTT TBdeltaTBmeasurement count

A3TEC statsfov/geo lat/long, lat TECTEC from L1c & A3, sigma

Max snapshots in global region ~1500, in OTT region ~500

Metrics adds < ~3Mb

A3TEC adds < ~90Kb

SMOS L2 Ocean Salinity

www.argans.co.uk

L2OS tuning for reprocessing

Reference L2OS data sets:

• v620 ref (from ESAC rehearsal) – un-tuned configuration:L1c v620 & L2OS v612 (20110423-20110607 & 20121104 – 20130205)

• v620 run2 (ARGANS) – first tuned configuration:L1c v620 & L2OS v613 (20110423-20110607 & 20121104 – 20130205)

•L1c v620 configuration update changes border width, RFI thresholds•ESL work shows need to revisit L2OS thresholds & filtering, and not to use sun tails flag to filter measurements

•Reprocessing L2 starts September/October 2014?, end Jan/Feb 2015, followed by L1 & L2 catch-up reprocessing of 2014 data => end April/May 2015?

•Process L1c (ESAC) May 2011 (ARGANS) in August with revised thresholds

•New thresholds delivered September 2014 for L2 reprocessing

4

SMOS L2 Ocean Salinity

www.argans.co.uk

L2OS forward model sensitivity analysis – flat sea model

Flat sea SST sensitivity: SST298 minus SST278

Sensitivity to SST is low in XX, but up to 0.1K/degree in YY at high incidence angles

Flat sea SSS sensitivity: SSS37 minus SSS33

~0.6K/psu at low incidence angles~0.9K/psu at high incidence angles

SMOS L2 Ocean Salinity

www.argans.co.uk

L2OS forward model sensitivity analysis – wind speed

General bias between roughness models of ~±0.5K at 7m/s

Model updates 2010-2011: more SMOS data now => better models

M1 XX

M2 XX

M3 XX

M1 YY

M2 YY

M3 YY

SMOS L2 Ocean Salinity

www.argans.co.uk

L2OS forward model sensitivity analysis

Worse case (eg high incidence angle) model sensitivities

7

  Flat sea Roughness model 1

Roughness model 2

Roughness model 3

Galactic noise

SSS (K / psu) 0.9 0.01 XX0.03 YY

0 0 0

SST (K / degree) 0.02 XX0.1 YY

0.005 0 0 0

WS (K / m/s) 0 0.4 XX0.2 YY

0.4 XX0.15 YY

0.4 XX0.2 YY

< 0.05

Wind direction (@ 7 m/s) 0 ±0.1 XX±0.15 YY

0 0 0

  BT (K) SSS (psu)

Flat sea 0.75 1.5

Roughness model 1

Roughness model 2

Roughness model 3 @ SST 17°C, WS 7m/s ±2 ±4

Galactic noise

Worse case (eg high incidence angle) model absolute error estimates

SMOS L2 Ocean Salinity

www.argans.co.uk

L2OS evolution

Major improvements in L2OS performance from v550 to v612 from:• correction of L1 long term drift (daily OTT)• new galactic scattering model• TEC retrieval in descending orbits => better Faraday rotation• multi-threading

Work-plan for 2014-2017 identifies tasks for further improvements:• minimize flaws in L1 over ocean (biases & flagging)• refine forward models• improve filtering retrieval methods• improve quality assessment & user experience

ESL already working on these tasks with promising results (land-sea, filtering) – we expect major improvements in quality of ocean products during next phase.

8

SMOS L2 Ocean Salinity

www.argans.co.uk

L2OS follow-on ESL contract: work-plan overview

9

WP Task Priority Difficulty Objective Lead CommentsMinimize flaws in L1 over oceanWP1.1 Land-sea 1 1 Empirical correction L1c TBs NR/JTWP1.2 OTT strategy 2 2 Matrix OTT/bias correction JLVWP1.3 Filtering 1 1 Improved RFI/sun flagging JLVRefine forward modelsWP2.1 Dielectric 1 1 Evolve dielectric model ? 1psuWP2.2 Roughness 1 2 Evolve models & geophysical dependencies NR SSS 1-14m/sWP2.3 Glint 2 2 Evolve GO model JT ≤1psuWP2.4 Aux data 3 2 Assess impact of geophysical aux data ?WP2.5 Rain 3 1 Identify surface & atmospheric impact XY/JBImprove retrieval methodsWP3.1 Inversion QC 1 3 Evolve OS quality metrics EOWP3.2 Retrieval algorithm 2 2 Stokes 1, 2-step, Monte-Carlo? XY/EOWP3.3 TEC 1 3 Evolve Stokes 3 TEC JLV New productImprove quality assessment & user experienceWP4.1 Validation 1 2 Improved validation tools EO Non-ESL taskWP4.2 Match-up DB 1 2 Match-up DB spec NR Non-ESL taskWP4.3 netCDF 2 3 Fool-proof products MAManagementWP5.1 Planning & delivery 1 2 Planning, documentation, processor delivery PSWP5.2 Coordination 1 2 Teleconf, PMs, QWG, SPCM, web, publications PS

SMOS L2 Ocean Salinity

www.argans.co.uk

L2OS follow-on ESL contract: processor delivery schedule

L2OS v630 spring 2015:• land-sea correction (WP1.1)• improved QC (WP3.1), filtering (WP1.3)• update of roughness/foam model 1 (WP 2.2)• could be use to re-reprocess L1 v620?

L2OS v640 spring 2016:• use of new constraints on roughness models (WP2.2)• better flat sea model (WP2.1)• improved GO galactic glint model (WP2.3)• new TEC product (WP3.3)• netCDF (WP4.3)

L2OS v630 spring 2017:• improved (adaptive?) retrieval algorithm(s) (WP3.2)• using new/alternative geophysical inputs (WP2.4)• mitigation of rain impact (WP2.5)• better OTT strategy (WP1.2)• comprehensive validation (WP4.1)

10

SMOS L2 Ocean Salinity

www.argans.co.uk

L2OS follow-on ESL contract: resources (months)

11

WP Task P D2015 2016 2017 2015 2016 2017 2015 2016 2017 2015 2016 2017 2015 2016 2017 2015 2016 2017

WP1.1 Land-sea 1 1 1 0.5 0.25 0.38 0.25 0.25 1.56 1.56 0.94 0.15 0.15 0.15 5.73 5.23 5.23 NR/JB/NMWP1.2 OTT strategy 2 2 0 0 1 0.13 0.13 0.13 0 1 0.5 0.4 0.4 0.4 JBWP1.3 Filtering 1 1 1.5 0.5 0.5 1.51 1.51 1.51 1.75 0.5 0.5 0.4 0.4 0.4 JBWP2.1 Dielectric 1 1 0 0.1 0.2 0.31 0.63 1.56 0.88 1.05 3.38 ?WP2.2 Roughness 1 2 0.25 0.5 0.25 2.01 0.5 0.5 2 2 2 2.9 2.9 1.9 JBWP2.3 Glint 2 2 0.25 0.2 0.15 1.56 1.25 0.94 3.63 3 2.38 NRWP2.4 Aux data 3 2 0.25 0.5 0.5 1 2 1 ?WP2.5 Rain 3 1 0 0 0.25 0.13 1.25 1 0.4 0.4 0.4 JBWP3.1 Inversion QC 2 3 1.1 0.4 0.35 3.15 3.15 3.15 0.6 0.6 0.6 BECWP3.2 Retrieval alg 2 2 0.75 1 1.25 1.88 2.01 2.51 3 3 3 0.5 1 1 1 1 1 JB/BECWP3.3 TEC 2 3 0.5 1.5 0.75 1.01 1.51 1.26 1.25 0.75 0.75 0.4 0.4 0.4 JBWP4.1 Validation 1 2 4.6 4.6 4.6 JB/NM/BECWP4.2 Match-up DB 1 2 3.1 2.1 2.1 NR/JB/NMWP4.3 netCDF 2 3 0.1 0.5 0.25 0.1 0.1 0.1WP5.1 Plan & deliver 1 2 0.5 0.5 0.5WP5.2 Coordination 1 2 0.5 0.5 0.5 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.5 0.5 0.5 NR

Total person-months 6.7 6.7 6.7 7.28 7.4 7.4 8.4 8.4 8.4 3.69 3.69 3.69 3.9 3.65 3.15 25.6 24.7 24.4

LOCEAN BEC Ifremer/ODL ACRI-st non-ESL

89

ARGANS