Frequencies and sensors for ice-quakes
- use loaned passive seismic sensors as a baseline/control
- implications for power use
- we may have to stop recording much in winter
- NERC Equipment pool -
we would put up to 5 on the surface for the two weeks in the summer.
Geophones - surface and buried - will need amp, adc channel and sampling trials
their commercial ADC/capture device is 24bit!
Walters et al used a GD11D but seem to sample at 1 or 4kHz which is fast! This means they can easily pick up 100Hz signals which we would be filtering out if we sampled slower - but are they important? - conclusion - look at NERC gear data and we need to design a filter so we're sampling at Nyquist frequency.
Their event detection was to look at 800ms of waveform, measure rms and compare that to the previous 80ms window's rms - a difference showing there is an event. This is nice as it takes into account the constant noise (a threshold would not). If we sampled at 200Hz that would mean storing a buffer of 200-400 bytes which is not too bad (even fits in RAM).
Looks like we need a 24dB amp - as they used. (we think we should mount that local to the sensor but this year we'll do cable up to amp). Must look into damping or the sensor (input impedance of amp). Programmable Amp may be best bet - on SPI/I2C so base station can alter gain.
We have a data sample from switzerland - and a copy of SAC to open it and see what its like. SAC data format
back to 2010_designs
