Seamount detection and size estimation using filtered geosat altimetry data

Local changes in the marine geoid (<100 nm in size) correspond well with bathymetric features such as seamounts. Thus the marine geoid height may be used to verify existing features, predict the bathymetry of unsurveyed areas, and fill gaps in existing data. The application of matching high‐pass filters to both the geoid and bathymetry data of an area allows the regional trends to be removed so that only the features remain. Filter values that begin to pass data with wavelengths less than 125 miles and all data with wave lengths less than 70 miles were selected. The high‐frequency variations of the geoid can then be correlated to the bathymetry and a scaling factor between the two calculated. The highest correlations (.81) were achieved using a cut‐off value for the filtered geoid data. A gridded synthetic bathymetry file was created by scaling the filtered geoid to the filtered bathymetry and adding the low pass background bathymetry. The gridded historical bathymetry could then be subtracted from the synthetic bathymetry in an automated method to display probable new features. A final selection of 458 previously unreported major features was then made.