ACCURACY ESTIMATES OF GEOPOTENTIAL MODELS AND
GLOBAL GEOIDS
Abstract
Kenneth F. Burke and Scott A. True
Defense Mapping Agency
St. Louis, MO 63118·3399 USA
Milan Bursa
Astronomical Institute
Academy of Sciences of the Czech Republic
Prague, Czech Republic
Karel Radej
Topographic Service of the Army of the Czech Republic
Prague, Czech Republic
The authors have developed a theoretical Geopotential Model Testing (GMT)
methodology and applied it to existing data. The GMT method was developed without
hypothesis regarding the geoid or the internal mass distribution of the Earth. The testing
accuracy is limited only by the observational errors at the testing sites. Therefore, the
GMT method requires that the geocentric positions and normal heights at the testing sites
be determined as accurately as possible. The GMT method is based on knowledge of the
geopotential, W o ' on the geoid. Wo was computed as a function of the four parameters
defining the reference ellipsoid, Wo ( GM, (0, a, f). The theoretical value of Wo is very
close to values computed directly from satellite altimeter data. The error in Wo should be
less than the error in the geopotential W p computed from the geopoltential model being
tested.
A global distribution of GMT sites has been selected to cover land surface areas.
Another GMT procedure is under development for oceans and seas. Numerical solutions
have been computed for the OSU9lA (360,360) geopotential model, as well as for the
corresponding geoid. This paper summarizes and discusses the numerical results of these
solutions.
INTRODUCTION
The geopotential coefficients C nm and Snm determined from satellite and terrestrial
observations are only valid outside the mass of the Earth. Strictly speaking, these
coefficients cannot be used to define equipotential surfaces inside the Earth's mass.
Therefore, any testing method for geopotential models should be based on surface
measurements. In addition, the testing method should avoid using any data that are
50
GLOBAL GEOIDS
Abstract
Kenneth F. Burke and Scott A. True
Defense Mapping Agency
St. Louis, MO 63118·3399 USA
Milan Bursa
Astronomical Institute
Academy of Sciences of the Czech Republic
Prague, Czech Republic
Karel Radej
Topographic Service of the Army of the Czech Republic
Prague, Czech Republic
The authors have developed a theoretical Geopotential Model Testing (GMT)
methodology and applied it to existing data. The GMT method was developed without
hypothesis regarding the geoid or the internal mass distribution of the Earth. The testing
accuracy is limited only by the observational errors at the testing sites. Therefore, the
GMT method requires that the geocentric positions and normal heights at the testing sites
be determined as accurately as possible. The GMT method is based on knowledge of the
geopotential, W o ' on the geoid. Wo was computed as a function of the four parameters
defining the reference ellipsoid, Wo ( GM, (0, a, f). The theoretical value of Wo is very
close to values computed directly from satellite altimeter data. The error in Wo should be
less than the error in the geopotential W p computed from the geopoltential model being
tested.
A global distribution of GMT sites has been selected to cover land surface areas.
Another GMT procedure is under development for oceans and seas. Numerical solutions
have been computed for the OSU9lA (360,360) geopotential model, as well as for the
corresponding geoid. This paper summarizes and discusses the numerical results of these
solutions.
INTRODUCTION
The geopotential coefficients C nm and Snm determined from satellite and terrestrial
observations are only valid outside the mass of the Earth. Strictly speaking, these
coefficients cannot be used to define equipotential surfaces inside the Earth's mass.
Therefore, any testing method for geopotential models should be based on surface
measurements. In addition, the testing method should avoid using any data that are
50