1. The SEEGOCE project Michel Diament and the SEEGOCE team (C. Basuyau, S. Bonvalot, C. Cadio, S. Déroussi, H. Duquenne, G. Martelet, V. Mikhailov, G. Pajot, I. Panet, A. Peyrefitte, C. Tiberi …) Bergen 28 june - 2 july 2010

2. SEEGOCE: Solid Earth Exploration with GOCE Bergen 28 june - 2 july 2010

3. Therefore we have to image the inner structure and understand physical processes of the Earth using indirect approaches such as gravity (and other information as seismology) After Hergé

4. Preparation of validation Combination of Goce with ground/airborne data Interpretation of gravity anomalies using dedicated techniques (CWT: continuous wavelet transform) Cooperative gravity-seismology modelling Take best use of gradients Before availability of Goce data?

5. Validation IGN in collaboration with IPGP carried out a series of absolute (with an A10) and relative (CG3/5) surveys over France. This data set is perfectly suited for validating Goce derived gravity anomaly and gradients. Red triangles : Abs. measurements Green points: relative ties.

6. 101001000 Spatial resolution (km) Coverage regional local global CHAMP (2000) GRACE (2002) GOCE (2009) Topex, Jason (currents !!) Airborne gravity Land & sea gravity Supraconducting gravimeter Absolute gravimeter Data on the Earth’s gravity

7. Our goal: to obtain the best accurate and resolved field in areas of interest by combining Grace, Goce and ground data

8. For that aim we use Poisson multipole wavelets  A 3D, harmonic function,  Well localized both in space and frequency,  Two parameters: scale and position. Large scale wavelet Small scale wavelet Multipolar sources e1e1 e2e2  Earth mean sphere  Appropriate to combine data with different spatiospectral characteristics Equivalent gravity sources  Compact representations  Each wavelet is a simple linear combination of non-central multipoles of low orders (Holschneider et al., 2003)  Data at any altitude  Any type of data Chambodut et al., 2005

9. Example: local refinement of a global model Panet et al., 2006 Local zoom-in Regional wavelet model (res. 75 km, 9600 wavelets) Zoom on the Marquesas islands (res. 30 km, 9500 wavelets) GRACE SH model in-situ Data Our method allows to increase the resolution over chosen area of interest: zoom-in.

10. Weighting function F e of the densities  The wavelet models of the gravity potential thus obtained also lead to a multi-scale analysis (CWT)  The correlations between the wavelets  and the potential T provide an integrated, regionalized view of the densities  : Analyzing wavelet Scale Wavelets also allow analysis of the anomalies

11. Application: Central Pacific a paradise… for studying mantle processes!  superswell  numerous volcanic chains  not always linear age progression: hotspot clusters  a fossil alignment displaying ages between 35 and 90 m.y.  no clear linear age progression 6000 4000 2000 0 -2000 -4000 -6000 meters Bathymetrie Gebco

12. Wavelets analysis bring new geodynamical results as to hot-spots Comparison of the gravity anomalies at different scales with the bathymetry ones: Origin of the volcanism: a plume under the Society islands, lithospheric control for Marquesas. Marquesas Tuamotu Society FZ Marquesas CWT of the static geoid Panet et al., 2006

13. Results of the continuous wavelet analysis at longer wavelengths wavelet scale = wavelength of the geoid anomaly Two larges scales geoid anomalies are well isolated:  a geoid low on French Polynesia (-5m)  a positive geoid anomaly 600 km west of the Line Islands chain (12m)

14. Testing physical models such oscillatory domes  The dome loose its thermal buoyancy, become denser again and fall back.  It breaks through the transition zone and may produce traps.  Secondary instabilities at the origin of the short hot spots tracks. Derived from studies of convection in a heterogeneous fluid (Davaille, 1999; Le Bars and Davaille, 2004) where the equilibrium between thermal and compositional effects have been investigated. Cavity plumeDiapiric plume

15. French Polynesia: an upwelling dome We interpret the negative geoid anomaly as the geoid signature of a less dense, therefore rising, dome extending from the CMB up to transition zone. This dome probably created Darwin Rise 100 m.y. ago when it was its ascending phase. This Pacific area could have thus registered a complete pulsation of the mantle. The South Pacific dome now and 100 m.y. ago. This rising dome can explain the volcanism in this area: stopped at the transition zone, secondary instabilities created at this interface then produce South Pacific hot spots: Society, Pitcairn. surface TZ CMB Cadio et al., in prep.

16. Another promising way: Goce data + seismology + GROUND DATA

17. AGU Fall Meeting 2006 The shortsightedness The indesiciveness ?? Cooperative modelling cures the shortsightedness of seismologists and the indesiciveness of gravimetricians.

18. As shown by studies realized with ground data and seismic tomography in many areas as: Baikal rift ( Tiberi et al., 2003 ) Mongolia ( Tiberi et al., 2008 ) Our planned targets with Goce data: Himalaya and French Polynesia

19. Finally we must not overlook the gradients. These new data call for dedicated interpretation methods. After having proposed and tested a new method for gradients denoising (Pajot et al., 2008) and for analysis (Mikhailov et al., 2007), we started working on gradients inversion with application to Africa.

20. Conclusions: As mentionned on Monday by R. Rummel: applications to geophysics using « real » Goce data can start now. Let’s do it and ultimately realize the geoscientists (and Jules Verne’s) dream!!