1. The error resulting from using only a finite number of terms of a series, or the error produced by using only a limited operator length in a convolution, or a finite gate length in a correlation. 2. The effect of digitizing an analog signal whose corresponding digital value exceeds the maximum value permitted; see clipped. Loss of information in digitized data because of truncation of high-order or low-order bits causes different kinds of errors. 3. In calculating the total mass from a gravity anomaly, the error resulting from integration not being carried to the limit of the anomaly.