The Effect of Finite Data Length in the Spectral Analysis of Ideal Gravity Anomalies

The interpretation of gravity data which has been transformed from the space domain to the frequency domain is subject to limitations imposed by the transformation process and the nature of the data. One potential source of error in the spectral analysis of gravity anomalies is the incomplete definition of the anomaly due to insufficient data length. Comparison of theoretical Fourier transforms and the transforms of finite‐length segments of gravity anomalies due to two‐dimensional vertical slabs and horizontal cylinders indicates that the interpretation error can be held to less than ten percent if the profile length is at least six times the maximum depth to the source of the anomaly. This result was achieved with a rectangular data window in the transformation process. Greater error resulted from the use of the Bartlett, Parzen, and Tukey data windows.