observed with ARGO profiles

Transcription

1 Global heat and salt content observed with ARGO profiles F. Gaillard, E. Autret,, LPO,Brest and C. Coatanoan,, SISMER/IFREMER, Brest ARGO science wotkshop, Venice,, march 2006

2 An overview of ARGO status and potential Context: SISMER operates the Coriolis datacenter.. This center is one of the ARGO GDAC, it also collects in real time a large number of in- situ T and S profiles Need to validate the data before distribution Want to offer synthetic products To do this,, Coriolis operates an analysis tool (based on optimal estimation) Products: anomalous data are flagged in database Diagnostics on sensor drift (see( posters) Validated dataset on standard levels 3D weekly gridded fields of temperature and salinity

3 Somes uses of the Coriolis gridded fields Give an overview of the ocean state Starting point for comparisons with model results Reference for satellite measurements (SST, SSS) Useful to understand the context in biology, fisheries Allow to compute climatologies

4 Method: Objective analysis: Gausian covariances (space and time) The analysis system Configuration Reference: WOA01 monthly climatology Grid: 1/2, 59 levels (5 to 2000 m) 190 areas

5 Origin: Data Vertical profiles of T and S ARGO profilers Real time CTD XBT XCTD Time series at fixed points (ex: Pirata) Surface buoys with T sensors Data provider: Coriolis: CT, PF, MO, XB, XC GTS via Coriolis: BA, TE Measurements are interpolated on 187 levels( m) (5, 10, 20 m)

6 Operating modes of the analysis system Data base 1 day Raw data Std_data T and S Maps Residuals Daily analysis Vertical Interpolation on standard levels Merging of over sampled data Optimal Interpolation in horizontal (departure from monthly climatology) Sensors monitoring, Correction estimates 12 months 1 week Weekly analysis: Std_data, Maps Residuals Scientific Re-analysis: Stable raw and std datasets Update of Climatology Ocean variability studies 1-10 years

7 Aims of the work Use the CORIOLIS weekly analysis to 1. validate the processing 2. evaluate ARGO potential 3. Perform heat and salt budgets The analysis system runs in real time since may 2005 We asked Coriolis to reprocess the full ARGO period ( ) 2005) in the real-time mode (daily( daily+weekly) Full years have been reprocessed (presently being validated) will follow period available next April We present here a few results for

8 Data available for the analysis Number of profiles per month (including moorings and buoys) Stations - TEMP ARGO OTHERS Stations - PSAL ARG O OTHERS Number of stations January February March April May June July August September October November Stations - TEMP ARGO OTHERS December 2003 Number of stations January February March April May June July August September October November Stations - PSAL AR G O OTHERS December Number of stations January February March April May June July August September October November Stations - TEMP ARGO OTHERS December 2004 Number of stations January February March April May June July August September October November Stations - PSAL ARG O OTHERS December Number of stations Number of stations January February March April May June July August September October November December January February March April May June July August September October November December

9 Temperature horizontal coverage layer m Map of the error on the annual mean indicate that near surface coverage has been increasing constantly from 2003 to 2005 Masked areas: error is above threshold for at least one season

10 Temperature horizontal coverage m

11 2005 T and S

12 Example: Temperature field as last december

13 Annual mean temperature anomaly: : 0-100m0 Annual mean anomaly is obtained as follows: 1. average weekly analysis over the month 2. compute anomaly relative to WOA01 monthly climatology 3. Compute seasonal anomaly as mean over 3 month 4. Annual mean = mean of 4 seasons A warm anomaly is clearly seen in the North Atlantic during the year 2005

14 Annual mean salinity anomaly: : 0-100m0

15 Zonal averages of heat content anomaly/woa01

16 Zonal averages of salt content anomaly/woa01

17 Time series of heat content of the 0-300m layer Blue: Levitus annual mean Dashed blue: average of Levitus annual mean over (not 0!!!) Dashed green line: linear fit to Levitus over Green line:linear fit corrected for the bias Green stars: our estimates

18 Summary starting in 2005, the ARGO array gives unprecedented access to the global ocean seasonal cycle for temperature and salinity necessary to obtain annual means, and to the deep (down to 2000m) ocean, that carried a different signal. The synergy and coordination triggered by ARGO makes available tremendous and ready to use data (homogeneous, quality controled on soon corrected) We have now access to the full 3D+time dimension of variability.. This opens the way to real monitoring, including determination of signal propagations, understanding of correlations with biological changes.