Many types of seismic analysis require ground‐motion data in physical units of displacement, velocity, or acceleration. Because seismometers do not respond uniformly to all frequencies, scientists must deconvolve the instrument response from recorded data. However, popular deconvolution methods violate causality and introduce acausal artifacts to deconvolved traces, which are unacceptable in some applications. In this paper, we discuss a new method using recursive filters that we consider superior for many purposes.

One simple and common method of instrument response deconvolution is spectral division by the analog instrument response  
graphic
(1)
in which S is the instrument’s sensitivity, K is the normalization...

First Page Preview

First page PDF preview
You do not currently have access to this article.