Impact of GOCE on lithospheric modelling in active plate margins ESA Living Planet Symposium 2013 Edinburgh, Michael Hosse (1), Roland Pail (1), Martin Horwath (1), Tetjana Romanyuk (1), Ben Gutknecht (2), Nils Köther (2) (1) Technische Universität München, Institute of Astronomical and Physical Geodesy (2) Christian-Albrechts-Universitaet Kiel, Institute of Geosciences, Department of Geophysics
About IMOSAGA Integrated Modelling of Satellite and Airborne Gravity data of Active plate margins Part of DFG programme „Mass transport and mass distribution in the Earth system“ SPP 1257 Tight cooperation and interaction of geodetic and geophysical expert groups ESA Living Planet Symposium 2013, Edinburgh, September
Project Objectives Investigation of active continental margins for –Density distribution in 3D –Dynamic Structures (Finite Elements) –Rigidity by usage of –Satellite gravimetry and gradiometry (GOCE, GRACE) –Terrestrial gravimetry –Combined gravity models (Least Squares Collocation LSC) ESA Living Planet Symposium 2013, Edinburgh, September
Motivation ESA Living Planet Symposium 2013, Edinburgh, September (Götze et al., 2011) revised gravity field information in geophysically interesting areas impact on static and dynamic lithospheric models
Datasources – Terrestrial data ~ point measurements (g abs, FA, BA) from different sources collected during the past 30 years widely spread over South America No informations available about used instruments and methods No accuracies available ESA Living Planet Symposium 2013, Edinburgh, September
Datasources – Digital Terrain Model ESA Living Planet Symposium 2013, Edinburgh, September Used datasets: - 3“ (land only) - 30“ (with bathymetry) Available at:
ESA Living Planet Symposium 2013, Edinburgh, September Validation of terrestrial data – Heights Differences to ACE2 –Max.: 915 m –Min.: m –Mean: 2.8 m –STD: 53.7m
ESA Living Planet Symposium 2013, Edinburgh, September Validation of terrestrial data – Heights Differences to ACE2 –Filtered at 80m (~25 mGal Free-Air) –Max.: 80 m –Min.: -80 m –Mean: 0.4 m –STD: 25,1 m
Datasources - Satellite gravity data ESA Living Planet Symposium 2013, Edinburgh, September Available at: (Mayer-Gürr et.al: The new combined satellite only model GOCO03s, 2012) The satellite only model GOCO03s GRACE: 7.5 years GOCE: 18 months SLR: 5 years, 5 sat CHAMP: 8 years Constrains: Kaula, deg > 180
ESA Living Planet Symposium 2013, Edinburgh, September Validation of terrestrial data - Gravity Difference terrestrial – GOCO03S (d/o 250)Difference terrestrial – EGM2008 (d/o 2190) [mgal]GOCO03SEGM2008 Max. Diff Min. Diff Mean RMS
Calculation schema ESA Living Planet Symposium 2013, Edinburgh, September
Calculation schema ESA Living Planet Symposium 2013, Edinburgh, September
Calculation schema ESA Living Planet Symposium 2013, Edinburgh, September
Preprocessing terrestrial data Dimension of normal equation depends on number of observations → restriction of points Distribution of points is very inhomogeneous ESA Living Planet Symposium 2013, Edinburgh, September Calculated for each of the stations in study area Selected the stations with the smallest value q within a radius of 3 km
ESA Living Planet Symposium 2013, Edinburgh, September Least Squares Collocation (LSC) … signal covariances, derived from covariance model fitted consistently to (reduced) gravity gradients and (reduced) terr. data … noise covariances: - terrestrial data: empirical - satellite gravity data: covariance propagation (including omission error !!!) BIG advantage: we obtain error estimates:
ESA Living Planet Symposium 2013, Edinburgh, September Input data & stochastic models Satellite data (reduced): Synthetized from GOCO03S, on grid Error information: - commission part: derived by full covariance propa- gation from GOCO03S variance-covariance matrix - omission part (> d/o 250) : derived by variance propagation from EGM2008 variances Terrestrial & altimetry data: Terrestrial data: 9910 points, irregularly distributed Altimetry data (DTU10): 2353 points, on grid Error information: according to quality criterion, ~10 mGal (land), ~4-5 mGal (ocean) mGal mGal
Results ESA Living Planet Symposium 2013, Edinburgh, September Output field [mGal]Error estimates [mGal] Larger errors where only satellite data available (mainly omission part!!!) Realistic error estimates.
Restore & Validation with EGM2008 ESA Living Planet Symposium 2013, Edinburgh, September Restore (GOCO d/o topo-iso) IMOSAGA01C Free-air anomaliesBouguer anomalies IMOSAGA01C EGM2008 mGal IMOSAGA01C EGM2008difference
Restore & Validation with EGM2008 ESA Living Planet Symposium 2013, Edinburgh, September Restore (GOCO d/o topo-iso) IMOSAGA01C mGal difference Improved long-wavelength behaviour Large differences in regions without terrestrial gravity Geophysical interpretation (Kiel) IMOSAGA01C makes more sense terrestr. input Error estimates [mGal]
Lithospheric model (IGMAS+) - Iquique ESA Living Planet Symposium 2013, Edinburgh, September Free-air anomaliesBouguer anomalies IMOSAGA01C EGM2008 mGal IMOSAGA01C EGM2008difference Iquique: terrestrial gravity available
Lithospheric model (IGMAS+) - Iquique ESA Living Planet Symposium 2013, Edinburgh, September Only small variations compared to EGM2008 Lithospheric model could be confirmed
Lithospheric model (IGMAS+) - Beule ESA Living Planet Symposium 2013, Edinburgh, September Free-air anomaliesBouguer anomalies IMOSAGA01C EGM2008 mGal IMOSAGA01C EGM2008difference Beule: low-quality terrestrial gravity data
Lithospheric model (IGMAS+) - Beule ESA Living Planet Symposium 2013, Edinburgh, September Significant deviation to EGM2008 Updated lithospheric model required (currently done)
Lithospheric modelling ESA Living Planet Symposium 2013, Edinburgh, September In many regions due to lack of terrestrial gravity data up to now no lithospheric profile could be modelled by IGMAS+. In these regions, now due to GOCE data lithospheric modelling will become possible for the first time!
Summary and conclusions Significant improvement of regional gravity field models through combination of terrestrial and satellite data Main benefit in areas with terrestrial data gaps, but also concerning the consistency in the long to medium wavelengths Proof of a clear impact of GOCE for lithospheric modelling, especially in topographically challanging regions Methodology also usable for validation of inhomogenous gravity databases Established processing procedure can be applied worldwide (also with low-quality terrestrial data) Updated lithospheric models will be available soon ESA Living Planet Symposium 2013, Edinburgh, September
Final conclusion ESA Living Planet Symposium 2013, Edinburgh, September