general overview Comparative experiment of three different methods of measuring rotation rates was provided. Array of twelve 3C geophones was used to compute ADR (array derived rotations). In the center of the array, two types of point rotation rate instruments were installed: Rotaphone and R-1. R-1 instruments were kindly provided by Dr. Chin-Jen Lin from Institute of Earth Sciences, Academia Sinica, Taipei, Taiwan. Thus, the experiment provided both the comparison of the two independent and fundamentally distinct point sensors as well as the comparison of ADR results with the point sensors. We give an overview of the three instruments used, tests provided to check its reliability and final comparisons (Rotaphone vs. R-1 and Rotaphone vs. ADR). Before the experiment, many tests were done. Both R-1 sensors were tested side by side in a distance of 10 cm measuring noise sources (passing cars) in a distance of tens of meters. Lennartz geophones were corrected for individual transfer functions measured during the huddle test as well as for true backazimuths. Quarry blasts were used as the sources. Measurement was held in the distance of first hundreds of meters from the explosions. Before use, array records were compared in time and spectral domains, significantly dissimilar records were disregarded, the others were filtered, corrected for individual instrumental characteristic, backazimuth corrections wer applied and then ADR method was used to compute the rotation rate. conclusions three methods of measuring rotation rate showed significant dependance on instrumental transfer functions; due to their unknown or dissimilar form, ADR can only be used for low number of pairs; comparison of two different types of point sensors showed generally good agreement of measured tilts, however, torsion seems to be much less accurate experiment design N 2 m 2 m Rotaphone R1 BR3 LE-3D Symbols: N 2 m 2 m N 2 m 2 m Rotaphone R1 BR3 LE-3D Symbols: KLECANY Quarry, BLAST 2, Z-component, 6-20 Hz KLECANY Quarry, BLAST 2, N-component, 6-20 Hz KLECANY Quarry, BLAST 2, E-component, 6-20 Hz 0 10 20 30 40 0.9 1 1.1 1.2 AMPLITUDE RATIO 0 10 20 30 40 -0.002 -0.001 0 0.001 0.002 PHASE DIFFERENCE (s) 0 10 20 30 40 0.9 1 1.1 1.2 AMPLITUDE RATIO 0 10 20 30 40 -0.002 -0.001 0 0.001 0.002 PHASE DIFFERENCE (s) 0 10 20 30 40 FREQUENCY (Hz) 0.9 1 1.1 1.2 AMPLITUDE RATIO 0 10 20 30 40 FREQUENCY (Hz) -0.002 -0.001 0 0.001 0.002 PHASE DIFFERENCE (s) Z N E -6.5 -4.4 -0.1 -0.8 -1.2 -0.4 -4.8 0.0 +7.1 -6.4 +1.0 +3.3 KLECANY Quarry 25.7.2012, finite differences dv y /dx and dv x /dy KLECANY Quarry 25.7.2012, finite differences dv z /dx and dv z /dy complete array of 3C geophones, Rotaphones, R1s and recording equipment Rotaphone, R1 and Lennartz LE 3D complete array of 3C geophones, Rotaphones, R1s and recording equipment settings of the array with different instruments preparation of the array Comparative Measurements of Three-Component Seismic Rotation Rate from Proximate Blasts Using Three Different Methods: Point Sensors vs. Small-Aperture Arrays kolinsky@irsm.cas.cz Johana Brokešová, Petr Kolínský and Jiří Málek 3 rd International Working Group on Rotational Seismology Conference, 22-25 September 2013, University of Canterbury, Christchurch, New Zealand Department of Geophysics, Faculty of Mathematics and Physics, Charles University in Prague, Czech Republic Department of Seismotectonics, Institute of Rock Structure and Mechanics, Academy of Sciences of the Czech Republic Rotaphone Rotaphone R1 R1 Rotaphone and R1 comparison rotation rate signals for two pairs of instruments; Z components (torsion) are not in agreement; tilts at N component are very similar and small discrepancies are pronounced at the tilt of E component eentec R-1 Rotaphone Lennartz LE-3D lite left: twelve raw records of LE-3D in the array; records are far to be similar (eigenfrequency of the instruments = 1 Hz) below: spectra of three selected records at E component; records 42 is significantly different from 22 and 32, which are similar to each other - mechanical sensor system designed to measure spatial ground motion gradients; - it consists of highly sensitive geophones connected to a common recording device - the geophones are mounted in parallel pairs to a rigid (metal) ground-based frame - the distance separating the paired geophones is much smaller than the wavelength - the instrument provides collocated records of translational and rotational seismic motions (with the same instrumental characteristics) - rotation rate is determined by more than one geophone pair, which allows to perform 'in situ' calibration of the geophones simultaneously with the measurement - geophones SM-6 (Sensor Nederland b.v) - frequency range 2 Hz – 40 (or 60) Hz - dynamic range 120 dB - least detectable motion 10 -9 m/s, 10 -9 rad/s (theoretical) / 10 -7 m/s, 10 -7 rad/s (in practice) - largest detectable motion 10 -1 m/s, 10 -1 rad/s filtered R-1 records (two instruments) with two different transfer functions applied; black is the eentec's one used for further computation spectra of both R-1 instruments placed side by side measured by the Czech team during the preparation of the experiment; torsion (Z) is significantly different transfer functions measured for both R-1 sensors by the Taiwan team; especially the amplitude responses are different amplitude and phase transfer functions given by eentec for R-1 sensors huddle test was provided to determine individual transfer function of all twelve instruments; it showed significantly dissimilar behavior processing the records: left plot: raw spectra of twelve records, middle plot: dissimilar records omitted; right plot: filtered spectra of used records Amplitude spectra; stations 22 32 42 Rotaphone ADR Rotaphone ADR As a result of all the steps (raw records, spectra comparison, selection of records, huddle test transfer function application, azimuthal correction, filtering) usable records of all 3 components of the array emerged; black are used ones, grey are the records discarded due to inappropriate instrumental characteristics; so only eigth neighboring pairs of geophones can be used for final ADR measurement dashed = raw solid = filtered black = with eentec's transfer function color = with measured transfer function Array Derived Rotations - opposite pairs of differences should provide similar results - differences contributing to the torsion (upper set of plots) are significantly different - differences contributing to tilts (lower set of plots) are much more similar (compare pairs of differences depicted by the same color) Rotaphone and ADR comparison - while torsion (Z) is different both in waveform and amplitude, tilt at N component are similar and tilt at E component has similar waveform while showing differences in amplitudes 2 2.5 3 3.5 -4 -2 0 2 4 VELOCITY (mm/s) 2 2.5 3 3.5 -4 -2 0 2 4 VELOCITY (mm/s) 2 2.5 3 3.5 TIME (s) -4 -2 0 2 4 VELOCITY (mm/s) 2 2.5 3 3.5 2 2.5 3 3.5 2 2.5 3 3.5 TIME (s) 2 2.5 3 3.5 2 2.5 3 3.5 2 2.5 3 3.5 TIME (s) 2 2.5 3 3.5 2 2.5 3 3.5 2 2.5 3 3.5 TIME (s) 2 2.5 3 3.5 -4 -2 0 2 4 VELOCITY (mm/s) 2 2.5 3 3.5 -4 -2 0 2 4 VELOCITY (mm/s) 2 2.5 3 3.5 TIME (s) -4 -2 0 2 4 VELOCITY (mm/s) 2 2.5 3 3.5 2 2.5 3 3.5 2 2.5 3 3.5 TIME (s) 2 2.5 3 3.5 2 2.5 3 3.5 2 2.5 3 3.5 TIME (s) 2 2.5 3 3.5 2 2.5 3 3.5 2 2.5 3 3.5 TIME (s) 2 2.5 3 3.5 -4 -2 0 2 4 VELOCITY (mm/s) 2 2.5 3 3.5 -4 -2 0 2 4 VELOCITY (mm/s) 2 2.5 3 3.5 TIME (s) -4 -2 0 2 4 VELOCITY (mm/s) 2 2.5 3 3.5 2 2.5 3 3.5 2 2.5 3 3.5 TIME (s) 2 2.5 3 3.5 2 2.5 3 3.5 2 2.5 3 3.5 TIME (s) 2 2.5 3 3.5 2 2.5 3 3.5 2 2.5 3 3.5 TIME (s) 22 32 42 Z N E scheme of the array: black are used geophones, gray are discarded ones; numbers represent backazimuth correction applied for respective records to obtain as similar horizontal measurement as possible