Mean sea surface computation using GEOS‐3 altimeter data

Publisher Website
Google Scholar
Abstract
The mean surfaces of several regions of the world's oceans have been estimated using GEOS‐3 altimeter data. Included in these regions are the northwest Atlantic, the northeast Pacific off the coast of California, the Indian Ocean, the southwest Pacific, and the Philippine Sea. These surfaces have been oriented with respect to a common earth center‐of‐mass system by constraining the separate solutions to conform to precisely determined laser reference control orbits. The same reference orbits were used for all regions assuring continuity of the separate solutions. Radial accuracies of the control orbits have been demonstrated to be on the order of 1 m. In the computation of these surfaces, the altimeter‐measured sea surface height crossover differences were minimized by the adjustment of tilt and bias parameters for each pass with the exception of laser reference control passes. The tilt and bias adjustments removed long wavelength errors, which were primarily due to orbit error. Ocean tides were modeled with the (Estes, 1977) global tide model. For comparison purposes the (Mofjeld, 1975) northwest Atlantic tide model was also used for the ocean tide evaluation. The resolution of the estimated sea surfaces varied from 0.25° off the east coast of the United States to about 2° in part of the Indian Ocean near Australia. The rms crossover discrepancy after adjustment varied from 30 cm to 70 cm depending upon geographic location. Comparisons of the altimeter‐derived mean sea surface in the North Atlantic with the 5' × 5' GEM‐8 detailed gravimetric geoid indicated a relative consistency of better than a meter.