HR: 14:10h
AN: G23A-03 INVITED     [Abstracts]
TI: Determination of the Canadian Gravimetric Geoid 2005 (CGG05) Using GRACE and Terrestrial Gravity Data
AU: * Huang, J
EM: jianhuan@NRCan.gc.ca
AF: Geodetic Survey Division, CCRS, Natural Resources Canada, 615 Booth St., Ottawa, ON K1A 0E9 Canada
AU: Véronneau, M
EM: marcv@NRCan.gc.ca
AF: Geodetic Survey Division, CCRS, Natural Resources Canada, 615 Booth St., Ottawa, ON K1A 0E9 Canada
AB: The remove-restore technique and the Stokes kernel modification allow the combination of satellite gravity models and terrestrial gravity data for the determination of regional/continental high-resolution gravimetric geoid models. A number of kernel modifications have been suggested over the years to minimize errors in the satellite models and terrestrial data. However, the selection of the proper modification to the Stokes kernel depends largely on data quality. The Vanicek and Kleusberg (VK) kernel provides an efficient way to reduce the far-zone contribution (or truncation) error in the case of poor satellite models and poor remote (far zone) terrestrial gravity data. On the other hand, the degree-banded kernel acts as a filter weight function to remove systematic errors in the long wavelength components of the terrestrial gravity anomalies. It places significant emphasis on the satellite model. Back in 2000, the VK kernel and EGM96 were used in determining the Canadian Gravimetric Geoid 2000 (CGG2000). The kernel was modified to degree 30 to prevent error accumulation from EGM96. The validation of CGG2000 at 430 precise GPS-leveling stations indicates a standard deviation of 21.7 cm. Today, the GRACE mission is advancing the determination of the Earth gravity field to a new level. It poses a compelling question: What is the proper approach to combine GRACE models and the terrestrial gravity data? This presentation investigates methods of combining the high-accuracy of the GRACE gravity models with the high precision of the terrestrial gravity data for optimal estimation of the long- and short-wavelength components of the geoid model. First, a series of numerical simulations are conducted to study the effect of systematic and random errors in the satellite and terrestrial data on geoid modeling. Second, geoid models for North America are estimated from the GRACE gravity model (GGM02C) and terrestrial gravity data using different combination approaches. The Canadian Gravimetric Geoid 2005 (CGG05) is based on GRACE data up to degree 90 and supplemented with terrestrial information for the higher frequencies. A GPS-leveling comparison indicates a standard deviation of 10.2 cm for CGG05, which is an improvement by a factor of two over CGG2000. Finally, the CGG05 error is estimated from the satellite and terrestrial gravity errors, calibrated using GPS-leveling data. This preliminary investigation of the actual accuracy of the geoid model for Canada shows error standard deviation of about 6 cm on average over Canada.
DE: 1214 Geopotential theory and determination (0903)
DE: 1240 Satellite geodesy: results (6929, 7215, 7230, 7240)
SC: Geodesy [G]
MN: Fall Meeting 2005