Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Interpretation of geopotential data using inverse spectral modeling techniques eliminates the ambiguous results produced by the standard forward-modeling approach. However, applying the spectral modeling procedures by hand is time-consuming and lacks precision. This study expands on current spectral analysis techniques by adding least-squares statistical rigor to the fitting of the chosen model. Automating the model fitting procedure allows an investigator to examine quickly many overlapping sections (windows) of mapped data. Successive overlapping of neighboring windows suggests a correlation between the ensemble-average depth solutions unique to each window is possible. This technique was shown to be especially useful in Louisiana and west Mississippi, where the thick sedimentary cover and relatively smooth anomalous geopotential fields insure correlation between successive windows. The correlated depth solutions appear to correspond well with horizons located by seismic refraction and inferred from hydrocarbon exploration studies.