The GEOSAT Files

This directory includes two files with geoid height and gravity anomaly profiles along the ground track of the Geosat Exact Repeat Mission. The data are the average of 44 repeat cycles as described in the article “High-Accuracy, High-Resolution Gravity Profiles From 2 Years of the Geosat Exact Repeat Mission” by D. T. Sandwell and D. C. McAdoo (Journ. Geophys. Res., v. 95, p. 3049-3060, 1990). Averaging many repeat cycles improves the accuracy (approx. 1 mgal) and resolution (approx. 10 km) of the data as well as the coverage in areas of seasonal sea ice. (For a description of the original, unaveraged data see: Cheney, R.E., B.C. Douglas, R.W. Agreen, L. Miller, and D. L. Porter, Geosat Altimeter Geophysical Data Record (GDR) User Handbook, NOAA, Tech. Memo., NOS-NGS-46, 32 pp., Natl. Ocean Serv., Rockville, Md., 1987).

Along track gravity anomalies were computed from along track vertical deflection by Fourier transformation of the profiles, exchange of the real and imaginary components and inverse Fourier transformation. For a more complete description of the method see: Haxby, W.F., G.D. Karner, J.L. Labrecque and J.K. Weissel, Digital Images of Combined Oceanic and Continental Data Sets and Their Use in Tectonic Studies, EOS Trans. AGU, v. 64, p. 995-1004, 1983. Since Geosat coverage is not dense in 2-dimensions, it was assumed that the gravity anomalies are 1-dimensional and are lineated perpendicular to each profile. Thus, these gravity anomaly profiles will not necessarily agree with shipboard measurements of gravity along the same trackline. Nevertheless, they reveal many previously unknown details in the marine gravity field.

File Format

This directory contains two files:

The data are grouped by passes which consist of continuous profiles having no tome gaps greater than 5 seconds (approx. 33 km). The first record of each pass is slightly different from the other records; the time_offset field is 0, the gravity anomaly field contains the number of records in the pass (as a 2 byte int), and the uncertainty field contains 1000 (as a 2 byte int). The program seehead.c reads the headers and skips the data.