10.5880/icgem.2017.006 Wu, Hu Hu Wu Institut für Erdmessung (IfE), Leibniz Universität Hannover, Germany Müller, Jürgen Jürgen Müller Institut für Erdmessung (IfE), Leibniz Universität Hannover Brieden, Phillip Phillip Brieden Institut für Erdmessung (IfE), Leibniz Universität Hannover GFZ Data Services 2017 ICGEM global gravitational model GOCE EARTH SCIENCE > SOLID EARTH > GEODETICS > GEOID CHARACTERISTICS EARTH SCIENCE > SOLID EARTH > GRAVITY/GRAVITATIONAL FIELD > GRAVITATIONAL FIELD Ince, Sinem Sinem Ince GFZ German Research Centre for Geosciences, Potsdam, Germany Reißland, Sven Sven Reißland GFZ German Research Centre for Geosciences, Potsdam, Germany Wu, Hu Institut für Erdmessung (IfE), Leibniz Universität Hannover Dataset https://dgk.badw.de/fileadmin/user_upload/Files/DGK/docs/c-777.pdf CC BY 4.0 IfE_GOCE05s is a GOCE-only global gravity field model, which was developed at the Institut für Erdmessung (IfE), Leibniz Universität Hannover, Germany. The observations with a time span from 1 November 2009 to 20 October 2013 are used for the model recovery. The GOCE precise kinematic orbit with 1-s sampling rate is processed for the gravity field up to degree/order 150, while the three main diagonal gravity gradients are down-sampled to 2 s and used to recover the model up to degree/order 250. With two additional Kaula’s regularizations, the combined model “IfE_GOCE05s” is derived, with a maximum degree of 250.


To develop IfE_GOCE05s, the following GOCE data (01.11.2009 - 20.10.2013) was used:
* Orbits: SST_PKI_2, SST_IAQ_2;
* Gradients: EGG_GGT_2, EGG_IAQ_2.


None any priori gravity field information was used.






Processing procedures:


Gravity from orbits (SST):
* Acceleration approach was applied to the kinematic orbit data;
* PKI data was at 1 s sampling rate;
* Model was derived up to d/o (degree/order) 150;
* VCM (Variance-Covariance Matrix) was derived arc-wisely from the post-fit residuals.


Gravity from gradients (SGG):
* Gradients Vxx, Vyy and Vzz in the GRF (Gradiometer Reference Frame) were used;
* Gradients were down-sampled to 2 s;
* Model was derived up to d/o 250;
* VCM was estimated arc-wisely from the post-fit residuals.


Regularization:
* A strong Kaula-regularization was applied to constrain the (near-)zonal coefficients that are degraded by the polar gap problem;
* A slight Kaula-regularization was applied to improve the signal-to-noise ratio of the coefficients between d/o 201 and 250;
* The regularization parameters were empirically determined.


Combined solution:
* The normal equations for SST and SGG were summed wih proper weighting factors;
* Weighting factors for SST and SGG were determined from variance component estimation;
* A direct inversion was applied on the final normal equation.
-180 180 -90 90