1 / 18

Monitoring the Ocean State from the Observations

Monitoring the Ocean State from the Observations. Stéphanie Guinehut Sandrine Mulet Marie-Hélène Rio Gilles Larnicol Anne-Lise Dhomps. Introduction. Our approach : Consists of estimating 3D-thermohaline and current fields using ONLY observations and statistical methods

tad
Download Presentation

Monitoring the Ocean State from the Observations

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Monitoring the Ocean State from the Observations Stéphanie Guinehut Sandrine Mulet Marie-Hélène Rio Gilles Larnicol Anne-Lise Dhomps

  2. Introduction • Our approach : • Consists of estimating 3D-thermohaline and current fields using ONLY observations and statistical methods • Represents a complementary approach to the one developed by forecasting centers – based on model/assimilation techniques • “Observation based” component of the Global MyOcean Monitoring and Forecasting Center lead by Mercator Océan • Previous studies have shown the capability of such approaches : • In producing reliable ocean state estimates (Guinehut et al., 2004; Larnicol et al., 2006) • In analyzing the contribution and complementarities of the different observing systems (in-situ vs. remote-sensing) (2nd GODAE OSE Workshop, 2009)

  3. The principle The observations The method The products Global 3D Ocean State [T,S,U,V,H] Weekly – 1993-2009 [0-1500m] 24 levels [1/3°]  MyOcean V1 RT/RAN Altimeter, SST, winds Guinehut et al., 2004 Guinehut et al., 2006 Larnicol et al., 2006 Rio et al., 2011 + MDT estimate Intercomparison with independent data sets and model simulations Analysis of the ocean variability Observing System Evaluation T/S profiles, surface drifters

  4. Global T/S  Armor3D - Method vertical projection of satellite data (SLA, SST) combination of synthetic and in-situ profiles 1 T(x,y,z,t) = (x,y,z,t).SLAsteric + (x,y,z,t).SST’ + Tclim (x,y,z,t) S(x,y,z,t)=’(x,y,z,t).SLAsteric + Sclim (x,y,z,t) 2 synthetic T(z), S(z) SLA, SST multiple linear regression 1 optimal interpolation 2 in-situ T(z), S(z) Armor3D T(z), S(z)

  5. Armor3D - 1993-2009 reanalysis NCEP Reynolds OI-SST 1/4° daily - 04/07/2007 SSALTO-DUACS MSLA 1/3° weekly DT - 04/07/2007 Synthetic T’ – at 100m Arivo climatology – July – T at 100 m

  6. Armor3D - 1993-2009 reanalysis In-situ observations – Coriolis data center Synthetic T’ – at 100m Armor3D T’ Argo T’

  7. Armor3D - Hydrographic variability patterns • Temperature variability over the 2004-2008 period (global zonal averaged) : Synthetic SODA 2.2.4 ARMOR3D SCRIPPS • Very similar results for Synthetic/ARMOR3D/SCRIPPS • No bias introduced by the method • Very promising to study the variability of the 1993-2000 period which suffers from poor in-situ measurements coverage 2008 2004 2005 2006 2007

  8. Hydrographic variability patterns • Temperature variability from 1993 to 2008 (global zonal averaged) : 2001 2005 1993 1997 2008

  9. Armor3D - Hydrographic variability patterns • Salinity variability over the 2004-2008 period (global zonal averaged) : SODA 2.2.4 Synthetic SCRIPPS ARMOR3D 2004 2005 2006 2007 • Argo obs sys mandatory 2008

  10. Global U/V/H  Surcouf3D - Method Surcouf : Field of absolute geostrophic surface currents weekly - 1/3° Armor3D : 3D T/S fields weekly - 1/3° - [0-1500]m Surcouf3D 3D geostrophic current fields weekly (1993-2008) 1/3° - 24 levels from 0 to1500m

  11. Surcouf3D - Comparison with model outputs • Vertical section at 60°W, in 2006 *geost. current with level of no motion at 1500m Armor3D* Surcouf3D at 500m Surcouf3D GLORYS at 500m GLORYS

  12. Surcouf3D - Validation of 1000-m currents • Global statistics over the Atlantic outside the equateur (10°S-10°N) • Comparison between 3 different methods (Surcouf3D, GLORYS, Armor3D) and in-situ observations (ANDRO) at 1000 m over the 2006/2007 period (Taylor, 2001) Meridionalcomponent ● SURCOUF3D (weekly, 1/3°) ▲GLORYS = Mercator-Ocean reanalysis (weekly, 1/4°) Ferry et al., 2010 ♦ Armor3D= geostrophic current with level of no-motion at 1500m (weekly, 1/3°) ● ANDRO = 1000-m currents from drifting velocities from the Argo floats (≈10days, ≈50/100km) Ollitraut et al, 2010 skill score • Results are very similar for the zonal component Correlation coefficient Standard deviation (cm/s) Standard deviation (cm/s)

  13. Surcouf3D - Validation Meridionalvelocities (cm/s) Zonal velocities (cm/s) • Comparison with GoodHope VM-ADCP observations from 14/02–17/03/2008 ADCP obs courtesy of S. Speich SURCOUF3D ADCP • Good correlation with independent in-situ obs. • Other time series to be compared 14/02/08 17/03/2008

  14. Surcouf3D - Validation • Comparison with RAPID current-meters in the Western boundary current off the Bahamas from April 2004 to April 2005 Florida Africa 26.5°North 76.5°West SURCOUF3D RAPID (current meters) GLORYS • Good correlation with independent obs., and with GLORYS • Importance of in-situ T/S profiles obs at depth for the inversion of the current

  15. Surcouf3D - AMOC variability at 25°N • Comparison with Bryden et al, 2005(section at 24.5° from Africa to 73°W and at 26.5°N off Bahamas) Floride Strait Transport from electrical cable (Bryden et al,2005) AMOC= Geost + Ekman + Florida (Surcouf3D, Bryden et al., 2005) Ekman Transport from wind stress ERAInterim Geostrophic Transport from 75°W to 15°W and from the surface to 1000m (Surcouf3D,Bryden et al., 2005) • Very consistent with Bryden et al, 2005 • Hight inter-annual variability • Hard to distinguish a long-term trend

  16. Surcouf3D - AMOC variability at 26.5°N • Comparison with RAPID and GLORYSfrom April 2004 to December 2007 • (monthly means + 12-month filtered) • Similar seasonal cycle • Amplitude differences ~ 10% • Higher variability in Surcouf than in Glorys that has to be further analyzed SURCOUF3D RAPID GLORYS

  17. Conclusions / Perspectives • Armor3D/Surcouf3D tools are very useful : • to perform intercomparison exercices • to study the interannual variability of the hydrographic patterns, the AMOC … • Intercomparison studies will be continued between Armor3D and Surcouf3D and MyOcean global reanalysis • Further study the ocean state (T/S variability, MOC, Heat/Salt transport) in key regions and for the 1993-2009 periods • Armor3D/Surcouf3D reanalysis are distributed as part of MyOcean

  18. • …

More Related