Galperin v. Classical Seismometers

Broad band seismometers come in two different flavors: Classical (XYZ) and Galperin (UVW). You can think of a Galperian seismometer as a classical triaxial seismometer that has been tilted at 54.7 degrees from vertical. Imagine an XYZ-axis balanced on its corner.

Two weeks ago (see What is the best broad band in the world?) we took a look at the self-noise of the following broad bands:

  • Geotech KS-1/2000/54000
  • Guralp CMG-3T/3TB
  • Nanometrics Trillium Compact/ 120P/ 240)
  • Ref Tek 151-120
  • Streckeisen STS-1/2 HG

We saw that some broad bands can vary in consistency between sensors from the same manufacturers, especially in the microseism band. Others can be very consistent.

What role, if any does the sensor design play in this? Are Galperin seismometers better than classical seismometers?

Adam Ringler and Bob Hutt of ASL (see What’s Inside ASL, read the paper here) concluded that they do not see any systematic difference:

galperian-orthogonal seismometers

Of the sensors we looked at last week, only the Nanometrics sensors and the STS-2 have Galperin designs. The rest are classical. Recall how both the Trillium 120 (Galperin) and the Reftek 151-120 (classical) showed similar levels of consistency between sensors, supporting Ringler and Hutt’s conclusion.

Yet despite their comparable self-noise models, Galperin seismometers have some practical advantages over classical designs:

  • True orthogonality guaranteed by design. All axes remain perfectly aligned and mutually orthogonal. In a classic “orthogonal” seismometer, the horizontal axes are, by design, non-orthogonal to the vertical. A slight tilt of as much as 2 degrees is introduced to each horizontal axis to compensate for the levelness of the seismometer or temperature change.
  • Troubleshooting made easier. Are you looking at instrument noise? Or some noise external to the sensor? With UVW you can easily diagnose the sources of spurious signals. Since the axes of a Galperin seismometer are aligned in UVW and not true XYZ and are therefore neither entirely vertical or horizontal but a combination, determining answering these questions is easy. External noise is generally either vertical (e.g., common mode electric noise, high frequency tilt of a pier) or horizontal in nature (e.g., low frequency ground tilt, pops caused by a loose foot, tilt caused by varying rates of temperature change). On a Galperin setup, these signals appear on ALL channels. Instrument noise, however, only appears on a single channel.
  • Single Sensor. The manufacturer only has to produce a single sensor.
  • Identical instrument responses. Shipping indentical seismometers with identical instrument responses month after month, year after year is a tall order. But one Galperian company has made a fine name for itself doing just that. The ability to distinguish between internal and external noise is key to accomplishing this. The instrument developer needs to be able to say with absolute certainty that any noise he is seeing when testing the finished product before shipping it to the end user is external. The Galperin instrument manufacturer is, therefore, better equipped to ensure the sensor meets specs and falls as far below the Nominal Low Noise Model as specs allow. This translates into the highest possible signal-to-noise ratio for the end user.


Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>