От: Misha Barmin Кому: theoph@gsras.ru Тема: Re: CSS_PAZ Дата: понедельник 01 марта 1999 г. 20:39 Viktor Dmitrievich, Po vtoromy voprocy. Ja vam visilaju starii mail J. Coyne (1994), sm. Attachment 1 v konze meila s opisanien CSS formata bazi dannix polusov i nulei. Mi zdes dannimi GSE stancii ne ispolzuem. Ja ne v kyrse menjali oni s togo vremeni u sebja ili net no dlja nashei basi (ctaroi) on goditsja, format GSE nemhogo ot nego otlichaetsja no ideja ta je samaja. Vse novie izmenenija navernjaka Chernobai znaet. Primeri dlja hekotorix ctanzii mojno naiti na ring computere v directorii /disk3/SUPERSTAR/geotool.res Bolee cvejix dannix u menja net. Pro programmu zapolnenija ja nichego ne znaju. Vsego nailuchego, pishite kak dela, Misha Attachment #1 ============ Instrument Response File Format The Center databases have been set up to incorporate cali- bration parameters and to point to an instrument response file. This memo describes the calibration and response file pointer fields in the databases as well as the first version (1.0) of the format for the response files. The calibration information is stored in three different tables of the Center databases; wfdisc, sensor, and instru- ment. The wfdisc table contains the calib and calper fields which give the calibration in nm/count at the calibration period. This is usually the best estimate of the calibration at the time of recording and does not change as better esti- mates are obtained. If a better estimate is obtained at a later date, the correction is made in the sensor table using the calratio and calper fields. Calratio is a dimen- sionless ratio of the correct calibration to the calib value in the wfdisc table. Calper in the sensor table is the same as it is in the wfdisc table. The correction is made in the sensor table to avoid making hundreds or thousands of changes in the wfdisc table. There are many wfdiscs for each sensor and they are linked through the sta, chan, time, and chanid fields. The sensor table is also linked to an instrument table through the inid field. The instrument table contains the nominal calibration factors in the ncalib and ncalper fields, pointers to the directory and file containing the instrument response, and a field giving the response type (e.g. paz, fap, fir, and mult for poles and zeros, frequency amplitude phase, finite impulse response, and multiple response types, respectively). Like the relationship between the wfdisc and sensor tables, several sensor entries can be linked to the same instrument. This structure allows a small number of instrument responses and calibrations to be used for a great number of stations and waveforms. By defining the various "calibration" values in units of nm/count at a specific period in the Center databases, the scaling of the response curves is explicitly defined. Thus, the responses stored in the external files need only preserve the true shape of the response curve, not the amplitude. The responses defined by poles and zeros, how- ever, do include a "normalization" factor in the format. It is included primarily to remain consistent with the response information as it is received at the Center. Although the Center will include these normailzation factors in the response files, we will not vouch for their appropriateness. We strongly recommend using the calibration and calibration period values to scale the response curve properly. The format allows the complete response to be given as a series of response groups that can be cascaded. Each response group can have a different format or representation including frequency, amplitude, phase; finite impulse response filters; and poles and zeros. Other representa- tions can easily be added in the future. Modern instruments are composed of several different components, each with its own response. This format can mimic the actual configura- tion of the instrumentation. One of the benefits of this design is that the response shapes from standard instrument components can be kept separately and combined into complete response files as the need arises. In addition, one will be able to choose which parts of the complete response curve they wish to remove from their data. For example, it may be preferable not to remove the anti-alias filter when removing the instrument response from waveform data. Of course, responses are sometimes given as frequency, amplitude, phase triplets that represent the response of the entire system, and in these cases, the advantages of the cascading responses will not be realized. In most cases, theoretical responses are given as poles and zeros, finite impulse response filters, or a combination of the two. Measured responses, on the other hand, are given as frequency, amplitude, phase triplets. The format labels each response group as either "theoretical" or "measured" which allows both types to be stored in the same file for retrieval as needed. When frequency, amplitude, phase values are given, interpo- lation routines are usually used to fill in the missing points of the response curve. Unless points are included in the response file at very low and very high frequencies, extrapolation may be required to generate some of these points. The following policy will be adhered to concerning fap responses. When the fap values are "theoretical", ampli- tude and phase values will be given at frequencies of 0.000001 and 1000.0 Hz. For "measured" fap responses, only the values reported will be included in the response file. We suggest that the "theoretical" curve be used to fill in any response values at frequencies outside the "measured" band. The format for the response curves is given below. The data will be stored in ASCII. In the version 1.0 format only three response groups are defined; paz, fap, and fir. Suggestions for other groups were made and included gen (as in the SEED generic format), pol (polynomials), but (butter- worth filter), har (harmonic oscilator), and others. Ver- sion 1.0 will not include most of these types of responses, although a butterworth response and a harmonic oscilator response will probably be added as response parameters are accumulated. By limiting the number of response groups in the format, the software required to read the format is simplified. Future versions may, however, incorporate them as our experience dictates. All response parameters received at the Center will be saved, and if it was necessary to con- vert from formats not defined in the format to one that is, an audit trail will be placed as comments in the response file pointing to the original data and to the conversion software. To get the response of a particular instrument, the calibra- tion and calibration period values must be known. The response shape curve defined in the external file is adjusted so that its displacement value is one at the cali- bration period. The calibration value can then be used to scale the curve to the appropriate value. If the displace- ment response is desired, this would be nm/count. Velocity or acceleration responses can also be obtained by multiply- ing the response curve by iw or -w2, respectively. The best estimate of the response at the time of the recording will be obtained using calib, calratio, and calper in the wfdisc and sensor tables. The nominal response is found using ncalib and ncalper in the instrument table. Table 1: _____________________________________________________________________________ | Response File Format | |_____________________________________________________________________________| |Line #| Position | Field | Format | Description | |______|__________|_______|__________|________________________________________| |1-L | 1-80 | - | a80 | General comments preceded by a # | |______|__________|_______|__________|________________________________________| |L+1 | 1 | 1 | a1 | # | |______|__________|_______|__________|________________________________________| | | 3-80 | 2 | a78 | instrument type/description (KS36000, | | | | | | GS-13, etc.) | |______|__________|_______|__________|________________________________________| | Instrument Response Group Using Poles and Zeros (paz) | |_____________________________________________________________________________| |L+2-K | 1-80 | 1 | a80 | comments (preceded by a "#") | |______|__________|_______|__________|________________________________________| |K+1 | 1-12 | 1 | a12 | response source | | | | | | (theoretical or measured) | |______|__________|_______|__________|________________________________________| | | 14-15 | 2 | i2 |sequence number | |______|__________|_______|__________|________________________________________| | | 17-28 | 2 | a12 |description (instrument, | | | | | |anti-alias, etc.) | |______|__________|_______|__________|________________________________________| | | 30-35 | 3 | a6 |response type (fir, paz, fap, etc.) | |______|__________|_______|__________|________________________________________| | | 37-80 | 4 | a44 |author or source of information | |______|__________|_______|__________|________________________________________| |K+2 | - | 1 | f or e |normalization factor (A0) | |______|__________|_______|__________|________________________________________| |K+3 | 1-8 | 1 | i8 |number of poles | |______|__________|_______|__________|________________________________________| |K+4-N | - | 1-4 | 4(f or e)|complex pole and complex error | |______|__________|_______|__________|________________________________________| |N+1 | 1-8 | 1 | i8 |number of zeros | |______|__________|_______|__________|________________________________________| |N+2-M | - | 1-4 | 4(f or e)|complex zero and complex error | |______|__________|_______|__________|________________________________________| | Instrument Response Group Using Frequency, Amplitude, Phase (fap) | |_____________________________________________________________________________| |L+2-K | 1-80 | 1 | a80 | comments (preceded by a "#") | |______|__________|_______|__________|________________________________________| |K+1 | 1-12 | 1 | a12 | response source (theoretical | | | | | | or measured) | |______|__________|_______|__________|________________________________________| | | 14-15 | 2 | i2 | sequence number | |______|__________|_______|__________|________________________________________| | | 17-28 | 2 | a12 | description (instrument, | | | | | | anti-alias, etc.) | |______|__________|_______|__________|________________________________________| | | 30-35 | 3 | a6 | response type (fir, paz, fap, etc.) | |______|__________|_______|__________|________________________________________| | | 37-80 | 4 | a44 | author or source of information | |______|__________|_______|__________|________________________________________| |K+2 | 1-8 | 1 | i8 | number of fap triplets | |______|__________|_______|__________|________________________________________| |K+3-N | - | 1-5 | 5(f or e)| frequency (in deg), amp, phase | | | | | | (in deg), amp err, phase err | |______|__________|_______|__________|________________________________________| | Instrument Response Group Using Finite Impulse Response Filters (fir) | |_____________________________________________________________________________| |L+2-K | 1-80 | 1 | a80 | comments (preceded by a "#") | |______|__________|_______|__________|________________________________________| |K+1 | 1-12 | 1 | a12 | response source (theoretical | | | | | | or measured) | |______|__________|_______|__________|________________________________________| | | 14-15 | 2 | i2 | sequence number | |______|__________|_______|__________|________________________________________| | | 17-28 | 2 | a12 | description (instrument, | | | | | | anti-alias, etc.) | |______|__________|_______|__________|________________________________________| | | 30-35 | 3 | a6 | response type (fir, paz, fap, etc.) | |______|__________|_______|__________|________________________________________| | | 37-80 | 4 | a44 | author or source of information | |______|__________|_______|__________|________________________________________| |K+2 | 1-12 | 1 | f12.4 | input samples/sec | |______|__________|_______|__________|________________________________________| |K+3 | 1-8 | 1 | i8 | number of numerator coefficients | |______|__________|_______|__________|________________________________________| |K+4-N | - | 1-2 | 2(f or e)| numerator coefficient and error | |______|__________|_______|__________|________________________________________| |N+1 | 1-8 | 1 | i8 | number of denominator coefficients | |______|__________|_______|__________|________________________________________| |N+2-M | - | 1-2 | 2(f or e)| denominator coefficient and error | |______|__________|_______|__________|________________________________________| | Additional Response Groups as Needed | |_____________________________________________________________________________| Example Response File (Fictional) # # ** CAUTION ** CAUTION ** CAUTION ** # # All responses in this file are displacement curves and have # arbitrary scales. The scaling information required to use # this file is contained in the calib (or ncalib) and calper # (or ncalper) fields of the wfdisc (or instrument) tables. # The calib value defines how many nm/count there are at the # calper period. Scale appropriately. # The convention followed for the Fourier transform is that the # forward transform (from the time domain to the frequency domain) # is defined with a negative exponent and the inverse transform # (from the frequency domain to the time domain) is defined with # a positive exponent. # # S-750 borehole instrument with GS1400 amplifier # # Response shapes with poles and zeros are defined by: # # T = A0 * (s-z1)(s-z2)....(s-zn)/((s-p1)(s-p2)....(s-pm)) # # where T = unscaled transfer function, # A0 is the normalization factor, # s = j*omega (imaginary angular frequency), # z1 through zn are the n complex zeros (in radians/sec), and # p1 through pm are the m complex poles (in radians/sec), # # # The response of this instrument is considered excellent up to # about 20 Hz where instrument noise can become a problem at quiet # sites. # # Jeff Stevens of S-cubed compiled and verified this data # theoretical 1 instrument paz Teledyne Geotech manual 0.46678E+22 20 -.78828E+05 0.0 0.0 0.0 -.500E+05 0.0 0.0 0.0 -.990E+04 0.0 0.0 0.0 -.672E+04 0.0 0.0 0.0 -.263E+03 +.4067E+03 0.0 0.0 -.263E+03 -.4067E+03 0.0 0.0 -.530E+03 0.0 0.0 0.0 -.625E-01 0.0 0.0 0.0 -.997E+00 +.7653E+00 0.0 0.0 -.997E+00 -.7653E+00 0.0 0.0 -.12566E+04 0.0 0.0 0.0 -.628E+03 0.0 0.0 0.0 -.28270E+01 0.0 0.0 0.0 -.28270E+01 0.0 0.0 0.0 -.28270E+01 0.0 0.0 0.0 -.28270E+01 0.0 0.0 0.0 -.862E+02 +.2584E+02 0.0 0.0 -.862E+02 -.2584E+02 0.0 0.0 -.6264E+02 +.791E+02 0.0 0.0 -.6264E+02 -.791E+02 0.0 0.0 13 -.3737E+03 0.0 0.0 0.0 -.1148E+04 0.0 0.0 0.0 -.6505E+04 0.0 0.0 0.0 -.78344E+05 0.0 0.0 0.0 -.2112E+06 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 # # Response shapes with frequency, amplitude, phase triplets have # units of Hz, displacement (arbitrary units, usually nm), and degrees. # # This response was derived from the complete calibration done # on June 16, 1987 at RSNY # measured 1 instrument fap Sandia report S-1425 21 0.1 +.740E-04 538.0 0.0 0.0 0.15 +.724E-03 495.0 0.0 0.0 0.2 +.502E-02 444.0 0.0 0.0 0.3 +.535E-01 357.0 0.0 0.0 0.4 +.105E+00 326.0 0.0 0.0 0.5 +.212E+00 290.0 0.0 0.0 0.6 +.331E+00 264.0 0.0 0.0 0.7 +.449E+00 246.0 0.0 0.0 0.8 +.664E+00 221.0 0.0 0.0 1.0 +.100E+01 193.0 0.0 0.0 1.2 +.142E+01 168.0 0.0 0.0 1.4 +.171E+01 154.0 0.0 0.0 1.7 +.210E+01 140.0 0.0 0.0 2.0 +.262E+01 124.0 0.0 0.0 2.5 +.337E+01 105.0 0.0 0.0 3.3 +.455E+01 82.0 0.0 0.0 4.0 +.544E+01 67.5 0.0 0.0 5.0 +.667E+01 48.3 0.0 0.0 8.0 +.840E+01 20.2 0.0 0.0 10.0 +.104E+02 -29.7 0.0 0.0 20.0 +.650E+01 -146.0 0.0 0.0