The assembly of 5 individual aeromagnetic surveys and 4 geographic regions to create the Central West Antarctica compilation was done in several steps.

DATA PROCESSING STEPS

1. Original unsampled field data were edited and filtered, as required, to remove any anomalous noise. This process was applied at varying times to all of the data transects provided, including those from the aircraft magnetometer, base magnetometer, and longitude, latitude, and elevation from the navigation systems.
2. The edited field data for each transect were then sampled at 1 second intervals and merged.
3. The diurnal component of the magnetic field was removed by first subtracting the base magnetometer values from those of the aircraft magnetometer, and then adding an average base magnetometer value, preferably quiet-time, to adjust the resultant residual field amplitudes to reasonable levels. For the CTZ2 survey, we used the average quiet-time geomagnetic value from the recordings of two ground base station magnetometers located at the base camp near the center of the survey (Saltus and Kucks, 1992). For all other surveys, we used the average geomagnetic value from the base magnetometer recordings for each complete survey. (See the data table for survey specifications, including the average base magnetometer values for each survey, and the data index plot for names and locations of all the geographic regions.)
4. The internal (main-field) component of the magnetic field, represented by the 1995 revision of the International Geomagnetic Reference Field (IGRF), which includes the Definitive Geomagnetic Reference Field (DGRF) for 1990, was subtracted from the aircraft measurements (Sweeney, 1990).
5. The data for each transect were then resampled at a 5 second interval.
6. All data were synthesized into a consistent data base by equating the differing magnetic values of the flight-lines and tie-lines at their intersection locations. The remaining data in each line were then adjusted for consistency with the new intersection values (Mittal, 1984). This process not only reduces the effects of a non-lithospheric component in the magnetic data due to the transects being flown at differing times, but it also makes adjustments due to differences in elevation.
7. Preliminary grids were constructed from the sampled aeromagnetic transects with a cell size of 1.5 kilometers (between 1/3 and 1/5 of the flight-line spacing of each survey), using a minimum curvature gridding algorithm (Webring, 1981).
8. Data in adjacent geographic regions were combined, not by application of some smoothing algorithm between adjacent grids, but by adjustment of all transect data that overlapped into the regions occupied by the adjacent grids, using a modification of the algorithm of Mittal (1984). This adjustment was performed by finding the locations where the overlapping data intersected the orthogonal flight lines in the adjacent area, and forcing them to match the mag values at those adjacent intersection locations. Because of the care taken in selecting the average base magnetometer (datum) levels, the correct application of the IGRF/DGRF suite of coefficients, and the adjustment of all transect data in the regions of adjacent survey overlap, no additional datum level adjustment was applied to any of the regional data to minimize boundary differences.
9. Data quality problems were addressed.
10. Final grids were constructed from the fully-adjusted flight line data using a cell size of 1.5 kilometers.
11. The composite IRE/BSB/WAZ/TKD grid was continued from its original flight elevation surface to a draped surface 1,500 meters above the bedrock elevation surface using the chessboard method of Cordell and others (1992), applied with a moderate lowpass filter.