We describe the problem of combining hydrography with marine geodesy and satellite altimetry for the purpose of determining the general circulation of the oceans, defining the eddy field, and improving the marine geoid. The critical problem is to understand the error budgets of four fields: orbit, height measurement, geoid, and ocean water density. Corrections must be made for atmospheric load, tides, tropospheric water vapor, wave height, and other parameters. A general formalism for deducing the geoid and ocean circulation is obtained in terms of inverse theory and applied to some limited examples.

No abstract

In satellite tracking using ground-based radars, an estimate of the total electron content (TEC) along the path to the satellite is required to measure accurately the range of the satellite. This estimate is necessary because the radar wave travels at a slower speed as it propagates through the ionosphere. The range error AR that is introduced is dependent on the radar frequencyf and on the TEC along the propagation path, and can be expressed bywhere Ne is the local electron density and R is the radar range. The TEC can vary significantly with the time of day, geomagnetic activity, and look direction. A real-time synoptic ionospheric monitoring system has been developed using data acquired from a T14100 Global Positioning System (GPS) receiver for use at the Millstone Hill satellite-tracking radar. The T14100 GPS receiver can track up to four GPS satellites at any one time. Each GPS satellite transmits signals at two different L-band frequencies: LI (1575.42 MHz) and L2 (1227.6 MHz). The TEC along the path to each satellite can be determined by combining both frequencies using the pseudorange and the integrated phase data. At Millstone, the TEC is measured every 3 s foreach GPS satellite in view. The data are input into a Kalman filter that is used to predict the coefficients of a simple TEC model with azimuth and elevation dependence. This model takes advantage of the real-time knowledge provided by the GPS data of the variations in TEC around the Millstone location. The coefficients for this model are then sent to the satellite-tracking computer, and the model is applied in real time to account for the ionospheric path delay to whatever satellite is currently in track. The preliminary results of using this ionospheric monitoring system at Millstone will be discussed. The zenith value of the TEC predicted by our GPS model will be compared with another ionospheric measurement, the foF2 obtained from the collocated University of Lowell Digisonde. The TEC values predicted by our GPS model during both geomagnetically quiet and disturbed time periods will be discussed, as well as those associated with high solar flux time periods. Finally, the improvement in our radar system calibration due to the use of the new GPS model will be demonstrated. This improvement is evident in observations of the average range residual over a pass of the Lageos satellite. The standard deviation of these residuals drops from approximately 20 TEC units to about 5 TEC units when the new GPS model is used to estimate the ionospheric refraction correction. A TEC unit is 1016 electrons/m 2 . At the Millstone radar frequency of 1295 MHz, this correction corresponds to an improvement in the standard deviation from 5.3 to 1.4 m. It is our opinion that, other than upgrading all satellite-tracking radars to dual-frequency capability, the GPS is currently the best monitoring system of the TEC available. NTIS GRA&I •" DTIC TAR ]Utiv;i no inced Ju:; 1; . ACKNOWLEDGMENTSWe would like to thank Pat Doherty and J.A. Klobuchar for use of their H...

Geodetic parameters describing the earth's gravity field and the positions of satellite‐tracking stations in a geocentric reference frame have been computed. These parameters were estimated by means of a combination of four different types of data: routine and simultaneous satellite observations, observations of deep‐space probes, and measurements of terrestrial gravity. This combination solution gives better parameters than any subset of data types. In the dynamic solution, precision‐reduced Baker‐Nunn observations and laser range data of 21 satellites were used. Data from optical cameras, in addition to those from 19 Baker‐Nunn stations, were used in the geometrical solution. Data from the tracking of deep‐space probes were used in the form of relative station longitudes and distances to the earth's axis of rotation. The surface‐gravity data in the form of mean anomalies for 300‐n.mi. squares were provided by Kaula. The adopted solution from each iteration was a combination solution and was chosen to improve the residuals of all types of data. In addition to these four data sets, astrogeodetic data, surface triangulation, and some recently acquired surface‐gravity data not included in the set used for the combinations were used for an independent test of the solution. The total gravity field is represented by spherical harmonic coefficients complete to degree and order 16, plus a number of higher‐degree terms. The half‐wavelength resolution of this global solution subtends about 11° at the earth's center. The accuracy of the global field has been estimated as ±3 meters in geoid height, or ±8.7 mgal. Coordinates of many of the stations are determined with an accuracy of 10 meters or better.

Geodetic parameters describing the earth's gravity field and the positions of satellite‐tracking stations in a geocentric reference frame have been computed. These parameters were estimated by means of a combination of five different types of data: routine and simultaneous satellite observations, observations of deep space probes, measurements of terrestrial gravity, and surface triangulation data. The combination gives better parameters than does any subset of data types. The dynamic solution used precision‐reduced Baker‐Nunn observations and laser range data of 25 satellites. Data from the 49‐station National Oceanic and Atmospheric Administration BC‐4 network, the 19‐station Smithsonian Astrophysical Observatory Baker‐Nunn network, and independent camera stations were employed in the geometrical solution. Data from the tracking of deep space probes were converted to relative longitudes and distances to the earth's axis of rotation of the tracking stations. Surface gravity data in the form of 550‐km squares were derived from 19,328 1° × 1° mean gravity anomalies. The surface triangulation data consisted of the datum coordinates of each tracking station. Coordinates and potential coefficients were derived separately for each iteration. The adopted solution in each iteration was a combination solution chosen to improve the residuals of all data types. In addition to these five data sets an independent test of the solution utilized sea level heights plus satellite‐tracking and surface gravity data not used in the combination. The total gravity field is represented by spherical harmonic coefficients complete to degree and order 18 and a number of higher degree terms. The half‐wavelength resolution of this global solution subtends about 10° at the earth's center. The accuracy of the global gravity field has been estimated as ±2.5 m in geoid height, or 64 mGal2. Coordinates of the fundamental laser stations are determined with an accuracy of 2–4 m, and those of the fundamental optical network, with an accuracy of 5–10 m. The best‐fitting ellipsoid has a flattening ƒ of 1/ƒ = 298.256 ± 0.001 and a semimajor axis ae = 6378140.4 ± 1.2 m.