Loading ...
Global Do...
News & Politics
3
0
Try Now
Log In
Pricing
The GLOSCAL project and observations of near sea surface salinity variability GLOSCAL : Global ocean sea surface salinity : calibration and validation of SMOS P.I. F. Gaillard, IFREMER / Brest Participant laboratories : IFREMER/LPO, LOS, SISMER IPSL/LOCEAN, LATMOS, LEGOS, Météo-France, CLS, ACRI-st See also presentation by N. Reul and poster by B. Picard Outline • Large scale variability of the SSS from in-situ measurements • Sea Surface Salinity drifters • Salinity vertical gradients near sea surface GLOSCAL: Large scale variability of the SSS from in- situ measurements Fabienne Gaillard, Emilie Brion, Karina Von Schuckmann Laboratoire de Physique des Océans / IFREMER Goals • Produce monthly gridded fields of near surface temperature and salinity based on in-situ data over the Period: 2002-Present • Compute the corresponding mean annual cycle and variances before SMOS launch date. To prepare the SMOS data validation • Keep producing the monthly fields during the SMOS mission using as many validated data as possible ISAS: In-Situ Analysis system/ Near Surface module ( ) ( ) ( ) 1 1 1 o f a f ao o a f T ao o ao o ao o d y y x x C C R d P P C C R C y y R C R d − − − = − = + + = − + − = + Method: Optimal Interpolation •yo: data vector (observations) •xa: state vector (T or S at grid points) •Pa: covariance matrix of analyzed field •R: data error covariance matrix •Co and Cao: covariance matrices of obs-obs and analyzed-obs •Xf: first guess for x at analyzed points •Yf: first guess at data point Grid: •½° Mercator •Analysis performed at 5 m (representing 0-10m layer) •Variances and Covariance scales defined at each grid point (see below) Data type and depth: •Argo profilers (5-15 m) •GOSUD: Thermosalinometer (3-20 m) •Drifting buoys (0(1m)) psu 0 1 Standard deviation of SSS wrt climatology Argo dataset • Argo measurement starts between 4-5 m and 10m, • Coverage is nominal since fall 2007 (see error map below) • Data downloaded from Coriolis are checked against climatology before use % of a- priori variance Annual mean salinity error 2002 2007 2006 2005 2004 2003 0 100 0 100 Producing validated TSG datasets A Processing protocol has been defined for TSG data from French research vessels and VOS and will be implemented during 2009 by Coriolis datacenter •Conversion to GOSUD – NetCdf format •Adding water sample and instrument calibration information •Visual check •Correction using water samples or Argo data (TSG-QC software) In order to produce salinity fields including TSG data and check the analysis tool we have processed the 2006, 2007 research vessels dataset Results from ARGO data: Global ocean T-S variability The mean 2003-2007 anomaly • Argo data have been combined to produce gridded fields of temperature and salinity over the period 2002-2007 (ARIVO product). Horizontal maps of the difference between our estimation of the mean temperature (b) and salinity (d) at 10m depth during 2003-2007 and WOA5 shows the global warming of the northern hemisphere. The salinity change has a more complex pattern: warming of the north atlantic correspond to a saltening, except in the tropical band. IN the North Pacific we observe strong zonal variability. • On the left: zonal integrals of ARIVO T and S changes relative to WOA05 (blue curves) and differences between NOAA optimum Interpolation SST short-term (NSST, 2003-2007) and long-term (NSST_L, 1990-2007, red). SST(2003-2007) – WOA5 Climatology SSS(2003-2007) – WOA5 Climatology Schedule • January-April 2009: Preparing ISAS-V4 surface module, testing • January-April 2009: Preparing 2002-2007 TSG dataset • Mai 2009 : Analysis of surface layer from 2002-2008 V1 dataset. • June 2009: Computing climatologies and variances • September 2009: Full depth analysis and estimate of mixed layer depth • October 2009: Real time SST/SSS/MLD and SMOS validations • November 2009: Upgrade of in-situ dataset (adding surface buoys and new TSG data) • SIO (SBE 37 SI) • Metocean (SBE 47) Both conductivity/salinity sensors from SeaBird • SURFACT (ASD C/ T sensor) (‘low cost’) Salinity drifters G. Reverdin, J. Boutin, F. Gaillard, J. Rolland, T. Delcroix LOCEAN, LPO, Meteo-France, LEGOS Salinity drifters Characteristics 28cm (recent 10cm) 40 cm 50cm & 17 cm Temp. depth 1-2months >12months About 6months Duration 40cm SBE 37 SI 30 minutes SIO/PacificGyre 26cm (recent 10cm) ASD 15 minutes SURFACT/LOCEAN 50cm Mean depth for conductivity measurement SBE 47 CT sensor 60 minutes Time resolution Metocean Manufacturer Validation of SSS from drifters • Extensive validation of Metocean/Pacific Gyre drifters in the Gulf of Biscay: SSS accuracy and drifts after 3 months < 0.03pss78 (Reverdin et al., JAOT, 2007); C and T sensors not at the same depth may create artificial diurnal S signal (corrected); after 8 months at sea, no bias on a Pacific Gyre in the N. Atl. • Validation of new SURFACT near Banyuls (to be used for process studies): Surface salinity measured by 2 SURFACT floats; samples taken close to the floats are indicated in blue. Low salinity values are due to freshwater coming from the coast, as confirmed by low salinity values on samples taken after the SURFACT time series. Accuracy < 0.05pss over 1.5 months with strong biology 7 Drifters deployed since Sept. 2007 in the North Atlantic (21 months of meas.) in the Gulf of Biscay (30 months of meas.): 4 Drifters deployed since October 2007 in the tropical Atlantic (20 months of meas.) Temperature measurements of a SURFACT tied up to a Metocean drifter Temperature measurements performed by two surface floats in the southern part of the Gulf of Biscay (44.25N, 3.63W), at three depths (56cm in red, 30cm in black, 15cm in blue). The closer to the surface the measurement is, the larger the temperature peak recorded in the afternoon is. Also the increase of temperature starts earlier at 15cm than at 30 and 56cm. Reverdin, Boutin et al. 2009, in prep. Trajectory of Metocean drifter (Oct 07-May 08) superimposed on rain rates from a) F14 SSM/I satellite (29 December 2007 at 2030) and b) F13 SSM/I satellite (29 December 2007 at 2111). Large dark dot = drifter position on 29 and 30 December 2007. Salinity decrease recorded by Metocean buoy at 7°N, 48°W. The black curve represents the salinity measured at 66 cm depth and the gray curve the temperature at 56 cm depth. Hénocq, Boutin et al., in revision, 2009 Freshening event observed by a METOCEAN drifter in the tropical Atlantic Ocean Date Area Number Responsible Campaign/ ship 2009 May 2009 Gulf of Biscay 2 PG 1 MO LOCEAN CAROLS- Gogasmos/ Antéa September 2009 Atlantic NE 3 MO LPO Podorange October 2009 Atlantic eq.W 3 MO LPO Podorange September 2009 Atlantic eq.E 2 MO 1 PG LOCEAN Pirata/ Antéa End 2009 Pacific W 5 PG LEGOS Commercial ship 2010 May 2010 Indian equatorial 2 PG LOCEAN Tara June 2010 Atlantic NE 2 MO 1 PG LPO Ovide/ Thalassa Total 2009-2010 22 (11PG + 11MO) Future deployments of PG and MO (depending on SMOS launch date) Need to be coordinated with J. Font/J. Pelegri and D. Stammer 17 remaining drifters to be deployed in 2011 Salinity vertical gradients near sea surface Claire HENOCQ*, Jacqueline Boutin, Gilles Reverdin, Sabine Arnault** *LOCEAN / ACRI-St **LOCEAN Objectives : - quantify the vertical variability of the salinity above 10m depth - rain influence ? - determine highly variable specific areas • TAO mooring in the equatorial Pacific (5S, 95W); hourly measurements of salinity at 1, 5 and 10m , sea surface wind speed, rain rate measured every 1à minutes Observation of rain impact at local scale Salinity vertical difference between 1m and 10m close to 1pss in case of rain Henocq et al (JAOT, 2009, in revision) S10m – S1m = 0,93 pss Rain Rate Wind speed Salinity at 1m Salinity at 5m Salinity at 10m 10m-1m salinity vertical difference averaged over 10 days: still large differences (>>0.1psu) • 4 TAO moorings (Eastern Equatorial Pacific, 95W ) • In case of large rain rate, differences >>0.1 psu remain What happens at tropical scale ?? Henocq et al (JAOT, 2009, in revision) • ARGO floats ( Coriolis data base) various depth (>4m) • Thermosalinographs : from ships 5 and 11 m depth from fixed moorings (TAO/PIRATA) 1, 5 and 10 m depth • CTD / XCTD ARAMIS project SISMER data base World Ocean Database 2005 • Correction and sorting of data according to quality coefficients ( eg. CORIOLIS data flagged 3 ou 4 discarded) • Classification of the differences according to vertical depths : ΔS 10-5 = S([8; 11]m) – S([4; 6]m) ΔS 5-1 = S([4; 6]m) – S([0; 2]m) ΔS 10-1 = S([8; 11]m) – S([0; 2]m) Data Measurements every meter between 4 and 10m Large differences (>0.2psu) observed close to intertropical convergence zones : CNFGG Mesure de la salinité par satellite: le projet SMOS 17/11/08 Henocq et al (JAOT, 2009, in revision) 1. Colocation of salinity vertical differences with satellite rain rates measured the same day and the 2 days before in a 12.5km radius: Satellites : AMSR-E, TMI, SSMI (F10, 11, 13, 14, 15) 2. Compute a « 3 days accumulation of maximum rain rate» ( = sum of daily maximum rain) (« 3D max rain rate ») parameter 3. Correlate vertical salinity differences with 3D max rain rate ; compute and mean and std of salinity vertical differences in classes of 3D max rain rate Influence of rain Influence of rain for 3D max rain rate < or >=10mm/hr 94,60 75,90 4,92 20,13 0,10 0,14 0,40 3,83 < -0,1 [0,1; 0,1[ [0,1; 0,5[ >= 0,5 Percentage of salinity differences[- ,1; 0,1[ Henocq et al (JAOT, 2009, in revision) Salinity differences (pss) « 3D max rain rate » < 10 mm/h « 3D max rain rate >= 10 mm/h Statistical relationship between vertical salinity differences and 3D max rain rate • Vertical difference increases with 3D max rain rate • Larger dependency obtained for vertical difference computed with salinity at 1m depth — linear regression lines computed from colocated (∆S, 3D max rain rate) • Mean of ∆S computed in 3D max rain rate classes Henocq et al (JAOT, 2009, in revision) [0; 1[ [1; 2[ [2; 3[ [3; 4[ [4; 5[ [5; 6[ [6; 8[ [8; 10[ >= 10 « 3D max rain rate » en mm/h All depths ∆S 10-5 ∆S 5-1 ∆S 10-1 All depths ∆S 10-5 ∆S 5-1 ∆S 10-1 ∆S 5-1 + ∆S 10-5 0,18 0,16 0,14 0,12 0,1 0,08 0,06 0,04 0,02 0 Salinity vertical differences (pss)∆S10-5 ∆S5-1 ∆S10-1 • On average over large data set, vertical variability (up to 1m) is less than 0.1pss • In rainy regions, signature of rain history on surface salinity is expected to be on average >0.15psu => this suggests to classify the differences between SMOS SSS (at 1cm depth) and 5m or 10m in situ SSS as a function of 3Dmax rain rate : – To look for better than 0.2 psu accuracy in rainy regions – To validate SMOS SSS by testing its sensitivity to rain history. This criteria could be used to validate SMOS SSS. • Need measurements of salinity close to the surface (depth < 50cm), especially in tropical regions Conclusion
