Key points are not available for this paper at this time.
The global SEASAT altimeter data set has been edited and adjusted using a crossing arc procedure fixing one long arc to provide control. In doing this a set of 168 master arcs were formed from repeat and near repeat arcs. The adjustment, carried out by D. Rowlands, was done in a global net, and four regional areas. The average crossover discrepancy before the adjustment was ±1.5 m, while after the adjustment (solving for a bias and trend) it was ±28 cm. The examination of the crossover discrepancies by area indicated larger than normal discrepancies in certain areas such as the Hudson Bay, the Mediterranean, and the North and Baltic seas. Some of these discrepancies are caused by invalid tide values applied to the altimeter data in these areas. The adjusted SEASAT sea surface heights were used to predict point values at 1° × 1° intersections in 10 different geographic regions for comparison with corresponding GEOS 3 values. A mean difference (GEOS 3 minus SEASAT) of 1.3±0.67 m was found, the mean difference being caused by different ellipsoid parameters. The adjusted data were also used to determine mean surface heights and gravity anomalies in 37,9051°×1°and 1178 5° equal‐area blocks with some of these values being on land. These estimations were done using least squares collocation using covariances functions scaled to each 5° region. The average predicted standard deviation of the 1°×1° anomalies was ±5.1 mGal, and ±2.7 mGal for the 5° blocks. SEASAT values were compared to the GEOS 3 values, finding a difference of ±7.8 mGal and ±0.87 m for the 1°×1° data, and ±2.2 mGal and ±0.76 m for the 5° data. Comparisons with terrestrial data showed no significant difference between the GEOS 3/SEASAT implied anomalies, although a number of significant discrepancies have been resolved with the SEASAT data.
Richard H. Rapp (Mon,) studied this question.