The Ohio State 1991 Geopotential and Sea Surface Topography Harmonic Coefficient Models
Loading...
Date
1991-08
Journal Title
Journal ISSN
Volume Title
Publisher
Ohio State University. Division of Geodetic Science
Abstract
This report starts with the analysis of one year of Geosat altimeter data starting from the orbits computed with the GEM-T2 potential coefficient model and consistent station coordinates (Koblinsky et al., 1990). The first stage in the processing followed the general editing procedures implemented by Denker and Rapp (1990) when working with GEM-T1 orbits. Additional altimeter data, beyond that used by Denker and Rapp, was selected below -63° latitude, in the Mediterranean Sea, and in several areas of high frequency signal. The original radial orbit theory is due to Engelis. The analysis solved for corrections to the GEM-T2 potential coefficient model, coefficients in a degree 10 potential coefficient expansion, and 8 parameters for each of the 76 arcs of data analyzed. The data used included the altimeter data, the GEM-T2 potential coefficients with its error covariance matrix, and surface gravity data represented by 1° x 1° mean gravity anomalies. The root mean square orbit correction was approximately 75 cm with the corrections to the potential coefficients corresponding to geoid changes on the order of 118 cm. After applying the orbit correction terms the adjusted crossover discrepancies were ± 20 cm with a sample point residual of ± 19 cm. The sea surface topography from this solution did not show the slope problem across the northern Pacific Ocean that was seen by Denker and Rapp with the GEM-T1 orbits. Variations of the sea surface over the one year of data were analyzed by fixing the geopotential model and orbit corrections from the one year solution and solving for a monthly sea surface topography representation to degree 15. Variations from the annual degree 15 solutions were analyzed in the time domain to find signatures at different frequencies, for example annually and seasonally. These changes were also studied to detect local variations of the sea surface. In a third stage of analysis a combination solution with the GEM-T2 potential coefficients and the recent 30' mean gravity anomaly data set was carried out using the same procedure as described by Rapp and Pavlis (1990). The global set of adjusted gravity anomalies was used to calculate a potential coefficient model to degree 360. The final potential coefficient model was formed by taking the coefficients from degree 2 to 50 from the first combination solution with the coefficients from degree 51 to 360 of the last solution. The standard deviations of each coefficient were computed from the adjustment process and by error propagation. The cumulative geoid undulation commission error of the 91A model to degree 10, 50, and 360 is 5 cm, 25 cm, and 49 cm, respectively. The OSU91 model was tested through orbit predictions and data fitting; through comparisons with geoid undulations computed from Doppler and GPS located stations, and with comparisons to geoid undulations implied by Geosat altimeter data. In the latter case the root mean square difference between the Geosat undulation (after orbit and sea surface topography correction) was 34 cm for OSU91 as opposed to ± 53 cm with OSU89B.