Seismic station monitoring using deviation from the Gaussianity

Degradation of the seismic signal quality sometimes occurs at permanent and temporary stations. Although the most likely cause is a high level of humidity, leading to corrosion of the connectors, environmental changes can also alter recording conditions in different frequency ranges and not necessarily for all three components in the same way. Assuming that the continuous seismic signal can be described by a normal distribution, we present a new approach to quantify the seismogram quality and to point out any time sample that deviates from this Gaussian assumption. We introduce the notion of background Gaussian signal (BGS) to characterize a set of samples that follows a normal distribution. The discrete function obtained by sorting the samples in ascending order of amplitudes is compared with a modified Probit function to retrieve the elements composing the BGS, and its statistical properties (mostly its standard deviation sigma (sub G) ). As soon as there is any amplitude perturbation, sigma (sub G) deviates from the standard deviation of all samples composing the time window (sigma ). Hence, the parameter log(sigma /sigma (sub G) ) directly quantifies the alteration level. For a single day, a given frequency range and a given component, the median of all log(sigma /sigma (sub G) ) that can be computed using short-time windows, reflects the overall gaussianity of the continuous seismic signal. We demonstrate that it can be used to efficiently monitor the quality of seismic traces using this approach at four broadband permanent stations. We show that the daily log(sigma /sigma (sub G) ) is sensitive to both subtle changes on one or two components as well as the signal signature of a sensor's degradation. Finally, we suggest that log(sigma /sigma (sub G) ) and other parameters that are computed from the BGS bring useful information for station monitoring in addition to existing methods.