Random noise of sensor mass positions. Integrating over the bandwidth of interest – 1 year (3e-8 Hz) to 1 minute (0.016 Hz) – gives time-domain rms noise.

Verification Testing of Trillium 360 - a New Seismometer for Global Seismology

REFERENCES Berger, J., P. Davis, and G. Ekström (2004). Ambient Earth noise: A survey of the global seismographic network, Journal of Geophysical Research, 109, B11307, doi:10.1029/2004JB003408.

Forbriger T, Widmer-Schnidrig R, Wielandt E, Hayman M, Ackerley N (2010) Magnetic field background variations can limit the resolution of seismic broad-band sensors. Geophys J Int 183(1):303–312

Holcomb, L. G. (1989). A Direct Method for Calculating Instrument Noise Levels in Side-by-Side Seismometer Evaluations, USGS Open-File Report 89-214, 34 pps.

Peterson, J. (1993). Observations and Modeling of Seismic Background Noise, USGS Open-File Report 93-322, 94 pps.

Sleeman, R., van Wettum, A. and Trampert, J. (2006) Three‐Channel Correlation Analysis: A New Technique to Measure Instrumental Noise of Digitizers and Seismic Sensors, Bulletin of the Seismological Society of America, Vol. 96, No. 1, pp. 258-271.

G. Bainbridge*, S. Upadhyaya, B. Townsend, A. Moores *GeoffreyBainbridge@nanometrics.ca

Ottawa, Canada

Trillium 360 Vault at Albuquerque Seismological Laboratory

Trillium 360 Borehole at Nanometrics (Ottawa, Canada)

Trillium 360 Borehole and Reference Sensors at Piñon Flat

Earthquake Spectrum at Piñon Flat

Magnetic Immunity

CONCLUSIONS Availability of the Trillium 360 observatory-grade seismometer in Vault, Posthole, and Borehole form factors enables recapitalization of aging equipment and densification of global teleseismic networks with new stations in a variety of environments, taking advantage of recent advances in seismometry to realize better performance and greater versatility. Performance is enhanced, not only in the essential metric of self-noise, but also in linearity and dynamic range of real earthquake signals, magnetic immunity, and the reduction of surface temperature effects and site noise by means of down-hole or buried deployment.

Figure 7: Magnetic Sensitivity of Trillium 360 Borehole seismometer to vertical magnetic field. Measured at Nanometrics, Ottawa on 11 Nov. 2015. Vertical channel sensitivity is shown measured at various frequencies with error bars and mean sensitivity fit line.

Figure 4: Coherence measurement of vertical noise on two Trillium 360 Borehole seismometers. Measured March 21, 2016 at 30 m depth in nearby boreholes at Nanometrics’ main office in Ottawa, Canada.

Immunity to magnetic fields is needed for low noise performance in the presence of solar geomagnetic activity. The USGS GSN solicitation for a Very Broadband Borehole Seismometer (April 16, 2014) included a specification for magnetic sensitivity < 0.08 m/s2/T. The Trillium 360 betters this specification by an order of magnitude as shown in Figure 7. Mean sensitivity to a vertical field was -41.27 dB = 0.0086 m/s2/T on Z. (Sensitivity was even lower on X and Y so an accurate measurement was not obtained.) Charles Hutt at Albuquerque Seismological Laboratory measured a similar value of 0.0076 m/s2/T on vertical and “little or no obvious response on the horizontals”. (Personal communication, 25 Jan. 2016.

Seismometer response to ground motion must be highly linear, to avoid generating noise in the presence of signal. Figure 6 shows the response to a major earthquake of Trillium 360 and other reference sensors installed at Piñon Flat. Signals are well correlated between seismometers above 1.5 mHz, but at lower frequencies there is a spurious response on some sensors as highlighted in red. The Trillium 360 (black line) is approximately 4x quieter than the reference sensors at the lowest frequencies.

Figure 6: Power Spectrum of March 2, 2106 Mw = 7.9 Indonesian earthquake measured at Piñon Flat on Trillium 360 and Reference Sensors. Courtesy Peter Davis, University of California, San Diego.

Figure 4 shows total signals (in red) and non-coherent noise of two T360 Borehole seismometers installed at 30 m depth in nearby boreholes. The difference of the two signals shown in light blue represents the sum of the noise of the two sensors. Subtracting 3 dB yields the average noise shown in purple. Average noise (purple) is below the Peterson New Low Noise Model (black) down to 300 seconds (0.0033 Hz) and matches well with the Trillium 360 noise specification (green) up to 1.5 Hz. At higher frequencies the distance between the two sensors causes some loss of coherence. Note that site noise is reasonably low despite the fact these sensors are installed in a major urban area. This is the benefit of installation at depth. At this site, noise levels are high within a 15 m surface sediment layer, but good performance is obtained in granite at 30 m.

Trillium 360 Borehole noise performance compares favorably to reference sensors at a quiet site, as shown in Figure 5. Signals are coherent in the range of the microseismic peak but differences in self-noise are visible on Z at lower frequencies. The Trillium 360 (thin blue line) matches its noise spec (red) on Z.

The Trillium 240 (purple) is somewhat below its typical noise spec (thick blue line). The STS-1 (yellow) is above spec, likely due to the age of the unit. The higher background signal on horizontals makes it difficult to distinguish self-noise, however T360 is at or below the level of the reference units at all frequencies.

Figure 2: Long-period Sleeman coherence analysis of three Trillium 360 Vault seismometers at ASL, Jan. 11, 2017. Q330HR gain = 1. Courtesy Charles Hutt and Adam Ringler, USGS Albuquerque Seismological Laboratory.

Figure 3: Short-period Sleeman coherence analysis of three Trillium 360 Vault seismometers at ASL, Nov. 21, 2016. Q330HR gain = 20. Courtesy Charles Hutt and Adam Ringler, USGS Albuquerque Seismological Laboratory.

Figures 2 and 3 above show noise on three Trillium 360 Vault units at ASL consistently matching their specification (red line overlay). Low-frequency noise is reduced 6 dB with respect to the Trillium 240 noise specification (blue line overlay), extending the NLNM crossing point out to 300 seconds.

Figure 2 with Q330HR gain 1 shows an additional digitizer noise slope from 0.1 to 1 s (1 to 10 Hz). Figure 3 with Q330HR gain 20 shows no additional noise, with all three units matching their noise specification from 0.1 to 1 s within the margin of error of the measurement. (The apparently elevated noise at long periods in Figure 3 is an artifact of FFT windowing.)

Figure 5: Power spectra of background signal at Piñon Flat, California. Courtesy Peter Davis, University of California at San Diego

ABSTRACT Test results for Trillium 360 show this seismometer can resolve the Peterson New Low Noise Model down to 300 seconds period. This has been confirmed at multiple sites: Piñon Flat (California), Albuquerque Seismological Laboratory (New Mexico) and Nanometrics (Ottawa, Canada), and in different types of installations using different unit form factors (see Figure 1).

The Piñon Flat deployment captured the March 2, 2016 Mw=7.9 Indonesian event and showed a response coherent with reference sensors including an STS-1 at periods down to 0.0015 Hz. At frequencies below 0.0015 Hz the reference sensors showed a noncoherent spurious response, i.e. noise in the presence of signal, whereas the Trillium 360 was relatively unaffected.

Magnetic sensitivity has been measured to be <0.01 m/s2/T in two independent tests at ASL and Nanometrics. Temperature sensitivity is ~3*104 m/s2/°C. This combination of low sensitivity to both magnetic field and temperature is achieved through magnetic shielding which resolves the side effect of magnetic sensitivity in temperature-compensated spring alloys.

The T360 seismometer components are sufficiently miniaturized for deployment in a 6” borehole. This enables low-noise performance even in an urban environment with thick sediments (at Nanometrics, Ottawa) since the seismometer can be emplaced in bedrock below surface sediments and away from surface noise.

Figure 1: Vault, Posthole and Borehole form factors of the Trillium 360 design.

SSA 2017