Global grids of gravity anomalies and vertical gravity gradients at 10 km altitude from GOCE gradient data 2009-2011 and polar gravity

The GOCE satellite measures gravity gradients which are filtered and transformed to gradients in an Earth-referenced frame by the GOCE High Level processing Facility. More than 32,000,000 data with 6 components are available from the period 2009-2011. IAG Arctic gravity project data was used north o...

Full description

Bibliographic Details
Main Authors: D. N. Arabelos, C. C. Tscherning, REGUZZONI, MIRKO
Other Authors: D. N., Arabelo, Reguzzoni, Mirko, C. C., Tscherning
Format: Article in Journal/Newspaper
Language:English
Published: 2014
Subjects:
Online Access:http://hdl.handle.net/11311/881382
http://www.isgeoid.polimi.it/Newton/Online_First/newton_online_first.html
Description
Summary:The GOCE satellite measures gravity gradients which are filtered and transformed to gradients in an Earth-referenced frame by the GOCE High Level processing Facility. More than 32,000,000 data with 6 components are available from the period 2009-2011. IAG Arctic gravity project data was used north of 83 deg., while data from Antarctica was not used due to bureaucratic restrictions by the data-holders. Subsets of the data have been used to produce gridded values at 10 km altitude of gravity anomalies and vertical gravity gradients in 20×20 deg. blocks with 10' spacing. Various combinations and densities of data were used to obtain and compare values in areas with known gravity anomalies. The (marginally) best choice was vertical gravity gradients selected with an approximate 0.125 deg. spacing. Using Least-Squares Collocation, error-estimates were computed and compared to the difference between the GOCE-grids and grids derived from EGM2008 to deg. 512. In general a good agreement was found, however with some inconsistencies in certain areas. The computation time on a usual server with 24 processors was typically 50 minutes for a block with generally 40,000 GOCE vertical gradients as input. The computations will be updated with new Wiener-filtered data in the near future.