Home

Eavesdropper Tutorial - the Kansas Geological Survey

1. 1 11 10 12 9 12 10 13 9 13 10 14 9 14 10 15 9 15 10 15 11 15 12 15 13 15 14 15 15 15 16 15 17 15 18 15 19 42 Mod O4 4 1 00 d d OOV CA CA P 7 NE 10 11 10 12 11 11 11 12 12 11 12 12 13 11 13 12 14 11 14 12 14 13 14 14 14 15 14 16 14 17 14 18 14 19 14 20 14 21 1 e oU AW 10 13 10 14 11 13 11 14 12 13 12 14 13 13 13 14 13 15 13 16 13 17 13 18 13 19 13 20 13 21 13 22 13 23 10 15 10 16 11 15 11 16 12 15 12 16 12 17 12 18 12 19 12 20 12 21 12 22 12 23 12 24 Pp 1 fh pm fm pd mt Fmt 00 00 1 E OY Ov C0 1 1 10 17 10 18 11 17 11 18 11 19 11 20 11 21 11 22 11 23 11 24 2 PO C2 PO 5 1 XO NOD 00 d OV ON CA CA 62 10 19 10 20 1
2. time and velocity groups across the expanse of the line for the entire line It should also be noted that at this point a general knowledge of the geologic environment is critical to selecting a meaningful velocity model This is mainly due to the variety of ways some reflection or worse noise energy will stack in at different velocities Distorting creating reflections as a result of incorrectly processing your data is not only possible it is very easy if care is not taken during each step of the processing flow STEP 8b AR DATAS SERERE The display format for the constant velocity stacks groups all the CDPs according to stacking velocity Figure 10 The constant velocity panels will be either stacked or unstacked depending on what you selected The move out velocity of each panel will be displayed in the upper right hand corner of the group of CDPs processed The 75 ms reflection in our sample data set changes velocity from about 1925 m s to 2600 m s across a distance of about 70 m There was no surface feature that would suggest such an extreme variability This in itself should encourage care during the velocity analysis portion of the processing flow The selection of the appropriate velocity function from constant velocity stacks is to some degree a qualitative judgement based mainly on experience and a fundamental knowledge of seis mology and the site geology The odds of avoidi
3. SEGBdemuliplex to KGS SEG22KGS SEG 2 engineering format to KGS ABEM2KGS ABEM seismograph to KGS DG2KGS Data General to KGS DHR2KGS I O DHR 2400 to KGS SCIN2KGS Scintrex seismograph to KGS 1 VAX MAINFRAME 2 TEEE IBM PC Available as separate package Raw unformatted data on your hard disk will be in the form of a sequence of files with identical prefixes and or extensions where each file represents a unique field file recorded on your seismograph and downloaded onto your computer s hard disk The naming process was done during either the downloading of your data to your computer s hard disk or at the time of acquisition and storage of the data in the field The total number of these individual sequential files will be equal to the number of field files copied onto your hard disk Once the data are on the hard disk of your computer in most cases this involves a simple copy command the appropriate conversion routine should be executed to correctly format your data for future processing with Eavesdropper software After completing the formatting operation all the individual field files should be contained in a single MS DOS file STEP 1 EORR EY AMPLE Fe Copy the contents of the floppy onto your hard disk When you list the directory as below the following sequence of files should be present on your hard disk gt dir return dplot exe dseis exe dview exe dengland dat doscn exe dfmai
4. tion of subtle features The actual sorting routine is not particularly difficult from a conceptual point of view but it does require a significant amount of information relating to the acquisition geometries of your multichannel data The sheer number of parameters and geometric configura tions that need to be defined make sorting potentially the most mistake prone part of processing 28 figure 5 seismic data Built into the sorting operation are several ways to cross check the accuracy of the information you have input However they are not completely automated you must request and check the output of these operations to verify the correctness of your parameter and geometry assignments Sorting your seismic data can be thought of as very similar to playing gin rummy The main goal is to order your data or cards into a sequence likely to be of the most use to you later For example in rummy you may be collecting by suit with seismic data you may be collecting by receiver locations In rummy you may be collecting by face value in seismic you may be collecting by common mid points With the card game the identification value numeric or other and suit are displayed on the face of each card with seismic data the identification location and size is contained within each trace header In order for the data to be brought together in a meaningful fashion you must select which partic
5. 1 gt gt 5 see for description of gt gt START 2 EDFM dat see EDKL for description of INPF 3 928 EDMT identifier calls the surgical muting subroutine The two requested parameters are the trace header words that identify the record trace pairs as described during the and EDFM routines The first requested parameter is the trace header word identifying the primary groups or records in this case the SSN trace header location 92 identifies the record portion of the record trace pair The second requested parameter is the trace header word that identifies the secondary group or trace location within the record in this case the channel number of the seismograph trace header location 8 identifies the trace portion of the record trace pair 4 TAPR 10 TAPR sets the taper length as described in Figure 3 The units of taper length are ms The taper slope is linear and is designed to allow a gradual transition from unaffected data to 100 percent mute The taper s 100 per cent point total attenuation is at the user defined first arrival mute value and the 0 percent point no attenuation is at the defined mute time plus or minus depending on the which end of the mute zone the taper length In the case of a 10 ms taper on a mute window starting at 57 ms and ending at 70 ms the data would experience 0 percent signal attenu ation at 47 ms increasing linearly to 100 percent attenua
6. 112 5X 224 225 113 114 6X 7X 115 8X X source location 229 230 231 41 x 51 116 9x 232 1 10 233 1 11 61 117 3 2 1 10X 294 295 11213 71 7 118 119 4 5 3 4 2 3 1 2 1 11X 12X 236 27 238 239 114 145 116 1 7 SSN traces 1 6 120 121 6 7 5 6 4 5 3 4 2 3 1 2 1 13X 14X 122 123 124 8 9 10 7 8 9 6 7 5 6 7 4 5 6 3 4 5 2 3 4 1 2 3 1 2 1 15X 16X 17X 244 245 246 247 248 1 22 1 23 124 is eo sd 224 121 125 126 11 12 10 11 9 10 8 9 7 8 6 ih 5 6 4 5 3 4 2 3 1 2 1 18X 19X 20X 249 250 251 252 3 24 424 144 15 CDP 253 127 128 13 14 12 13 11 12 10 11 9 10 8 9 8 6 7 5 6 4 5 3 4 2 3 1 2 1 254 255 256 257 184 258 7 24 259 191 201 260 8 24 202 261 203 262 9 24 204 263 205 11 station location 264 10 24 206 265 207 10 133 19 18 17 16 15 14 13 12 11 5 266 11 24 208 10 267 209 9 268 12 24 20 10 9 269 270 13 24 271 272 1424 273 274 275 276 16 24 277 278 17 24 279 2041 2012 2043 2044 2045 2046 2017 20 8 2019 2020 2021 8 8 7 7 6 6 5 5 4 4 3 140 280 18 24 281 141 282 19 24 2022 2023 2024 3 2 2 142 143 145 146 147 1
7. EXIT Alt H HELP COMMAND F2 function key At this point the screen will scroll and the following question will appear INPUT DATA FILE dengland raw lt enter gt OUTPUT DATA FILE JUNK dat lt enter gt At this point the following message will be displayed Wait while copying a data file The second level of the program will appear at this point In order to select the trace to be analyzed press the F3 function key and make the changes as indicated below 66 SPECTRUM ANALYSIS PARAMETERS Input Data File dengland raw raw feld data Output Data File junk dat Sample rate 2000 samples sec Sample size 500 samples SSN Trace 5 18 lt enter gt Here we have selected a trace with good data This good data trace will allow us to compare how much and what frequency each particular type of energy represents on an entire trace Window start end 30 lt enter gt 250 lt enter gt The window will allow you to select particular types of seismic energy to analyze It should be noted the larger the window the more representative the displayed spectrum of the real data will be When the sample size gets small the input data program will automatically pad them with zeros to avoid edge effects The result of inputting zeros is an exaggerated smoothness of the spectrum Vertical scale Linear 0 Log 1 0 lt enter gt The log scale option will allow you to see lower amplitude information The linear scale will allow
8. Table 1 continued operation C data description lt sequence lt optional sequence T with the organization as well as some of the rationale for key parameter selections Each person should establish a processing flow that is somewhat tailored to both the needs of the particular survey and to existing equipment Some operations can be reordered removed included used several times etc and will either enhance deteriorate or not change the final section however some of the core operations stacking surface consistent statics and normal moveout do require pre requisite operations All processing operations require proper formatting of the input data Processing seismic data requires a basic understanding of acoustic wave propagation through the earth layered media Attempting to process data without the necessary physics and math background will eventually result in frustration or bogus results To assist the novice or out of practice seismic data processor a sample data set is included with this manual The data were collected during May 1989 in England along the Thames River Several good quality reflectors can be interpreted directly off the raw field files The data set will not present a significant challenge for the seasoned seismic data processor but it will require a variety of standard processing steps properly applied to obtain a stacked
9. lt enter gt 75 Again the program will keep you apprised of its progress through the input data set As with previous operation the list or journal file filt Ist will still possess all the processing history as well as any errors or abnormal terminations Once complete the dview or dplot routine should be used to display at least a couple of files to ensure the results of the filtering operation were what was desired The cleaning up of the data after should be quite evident Figure 18 The bandpass filtering of our sample data set not only improved the signal to noise ratio it also removed DC bias present on the raw field data DC bias is related to wander in the analog side of the seismograph and appears on raw field data as over or under shading of the variable area wiggle trace The removal of DC bias is critical to the effective stacking of seismic data Comparing Figures 11 and 18 it is possible to identify the DC bias The amplitudes of the reflection wavelets have much more trace to trace consistency after the filtering operation The low cut filter is actually responsible for the removal of the DC bias p AMPLITUDE BALANCING The relative amplitude of each sample recorded on a seismic trace is dependent on several properties and parameters The properties that dictate amplitude response are related to the earth and the propagation of acoustic waves through the earth The parameters that influence the amplitude of any one sampl
10. rsrt dat since it is in field file format 3 OUTFT ff5 dat see previous description of OUTF 4 gt gt END see previous description of gt gt END To extract field file 5 from the rest of the data set the previous batch pro cessing file needs to be run as input to the program seis The following sequence must be entered gt DSEIS rsrt dek rsrt lst lt enter gt Now field file 5 is the lone file in the MS DOS file ff5 dat To use the VELP program for a first guess approximation of the stacking velocity of reflections identifiable on field file 5 the program must be executed and questions answered in the following sequence gt VELP lt enter gt enter seismic data field file lt cr gt to quit ff5 dat lt enter gt The input data must have the geometries input into the appropriate trace header locations The program will respond with the following set of messages for our file Data being scanned please wait CDP Numbers from Input CDP gather Data First CDP 5 Last CDP 5 Enter Record cr to quit 5 lt enter gt Enter starting time in ms to display data DF 0 lt enter gt DF means default in this program AGC function 1 0 off DF 0 1 enter t Enter AGC Window length in ms DF 10 150 lt enter gt 54 t The option to choose window length only comes up if the function is ON Otherwise this question will not appear ENTER NUMBER OF BITS TO DISPLAY
11. separating the f in inpf and the first character of the file name Entries following an alpha identifier i e EDKL AUTS need only be separated by a minimum of one space In other words during batch processing any input information need only be in the correct relative sequence The program dseis is insensitive to absolute field location and small or large case i e A or a EDKL 928 EDKL calls the kill trace sub routine The traces to be removed are identified by a record number trace number pair 92 SSN 8 field trace number Record numbers generally identify the primary grouping order In case of raw field data the primary grouping is by field file number which is con tained in trace header location 6 for CDP gathers the primary grouping is according to CDP number which is contained in trace header location 12 Trace numbers generally identify secondary grouping order In the case of raw field files the secondary grouping is according to seismograph channel numbers which is contained in trace header word 8 This group ing or ordering can be changed during the sorting or re sorting operation These record pairs be thought of in a similar fashion to cards in a deck of playing cards that is the suite hearts clubs spades diamonds can be thought of as the record number and the value Queen Jack ten nine etc can be compared to the trace number such that any trace can be identifie
12. the appropriate mute can be designed Once the mute window for each field file has been determined a batch processing sequence similar to the following should be created Line 1 2 5a Description gt gt START see for description of gt gt START EDKL dat see EDKL for description of INPF 92 8 identifier calls the first arrival mute subroutine The two requested parameters are trace header words that identify the record trace pairs as described during the routine The first requested param eter identifies the primary groups or records in this case the SSN s trace header location 92 identifies the record portion of the record trace pair The second requested parameter is the trace header word that identifies the secondary groups or traces location within the record in this case the channel numbers of the seismograph trace header location 8 identifies the trace portion of the record trace pair TAPR 10 tapr sets the taper length The length is in ms The taper slope is linear and is designed to allow a gradual transition from unaffected data to 100 percent mute The taper s 0 percent point total attenuation is at the user defined first arrival mute value and the 100 percent point no attenuation is at the defined mute time plus the taper length In this case if the first arrival mute extended to 30 ms on trace 1 the taper would attenuate 90 percent of the s
13. 108 115 138 surface station Note A piece of the stacking chart is represented above to help equate the station numbers to the actual source and receiver patterns A stacking chart allows you to visualize the relative location of the shots and receivers pn defines each unique relative source to receiver geometry The pattern described on line 5a is designated as pattern number 1 it will be referenced by that number 1 later in this batch job It defines a source location 108 and its associated first receiver location 115 as well as indicates consecu tive receiver locations 24 The fifth value designates the amount each surface station was incremented after each shot In this case stations were incremented by 1 In other words station 115 equates to trace channel 1 of your seismograph station 116 equates to trace channel 2 of your seismograph If your source was occupying being fired every station location but your receivers were located live at every other surface station you would 37 have used a 2 for the fifth input value to indicate the receiver spacing was twice the surface station spacings Then surface station 115 would equate to trace channel 1 and station 117 would have equated to trace channel 2 etc Another possibility is that your roll switch is wired reverse to this sequence which means your trace channel 1 is at surface station location 138 when surface stations 115 to 138 are selected If this is the cas
14. 24 trace channel 24 148 149 150 151 152 24 22 21 20 153 24 22 21 154 155 24 23 156 24 The lower portion of the stacking chart was derived from the upper part The CDP numbers identified along the x axis on the lower part are exactly double the surface location defined directly above on the upper portion of the chart The number pairs beneath each location identify the 55 and seismograph channel number for the trace s sampling this particular midway point between source and receiver CDP As an example locate SSN 5 and seismograph chan nel number 10 on the upper portion of the stacking chart Next find the point which is midway between the shot for SSN 5 and the associated seismograph channel 10 Finally extrapolate that point straight down into the lower portion of the chart and you will find the number pair 5 10 as the fifth set of number pairs beneath CDP location 240 The bottom set of numbers running parallel to the x axis identify the fold of the associated CDP The fold can be figured by simply adding the number of SSN trace number pairs beneath each CDP location The program dseis has an option to print a table which will allow you to directly compare your hand generated stacking chart directly from the field notes with a computer generated chart reflecting the geometries and parameters you defined for the sort operation Our computer generated chart for the sample da
15. 3 10 10 kill 4 4 9 9 24 24 kill 5 5 8 8 23 23 kill 6 6 7 7 22 22 kill 7 7 6 6 21 21 kill 8 8 5 5 20 20 kill 9 9 4 4 19 19 kill 10 10 3 3 18 18 kill 11 11 221717 kill 12 12111616 kill 13 13 15 15 kill 14 14 14 14 kill 15 15 13 13 kill 16 16 12 12 kill 17 17 11 11 kill 18 18 10 10 kill 191999 kill 20 20 9 9 edfm 92 8 tapr 10 farm 51 30 24 65 edmt 92 8 tapr 10 mute 5 1 57 70 24 222 237 sort 12 19 ptrn 2 5 241 pn 1108 115 241 shot 82 260 250 240 230 17 gt rue PC gt gt n S5 Cee 5 gt gt 9 T gt lt a LE ee my eee gt gt TP S C C gt ASS CCT CER gt 999 lt gt gt CC CC GC OR CC CK Conf GS lt lt 1 lt lt lt Xe lt lt lt c e r r er figure 20 11081001 sn 2 1091 sn lt 3191 gt lt 110 1261 gt 1001 20 1271001124 tabl 11 surf 100 alof 105 770 90 se 126 99 100 se 127 100 se 136 100 se 1
16. The trace header locations will always remain the same The actual values within each word in the header can or will be changed in accordance with the described operation and input parameters The standards for formats on seismographs have changed consistently through the years Such familiar acronyms as SEGA 5 SEGY modified SEGY 5 2 etc describe individual variations The recent advances into non magnetic tape storage media have caused introduction of a multitude of formats For the most part each manufacturer of a seismograph with non 9 track tape storage media has a preferred trace header format A standardized format is being considered When a standard has been adopted there will no longer be a need for the format conversions described in a later section The key information necessary to process seismic data using Eavesdropper is contained within the 240 byte 120 16 bit words trace header preceding the seismic data itself which is represented by 2 bytes per sample word i e 500 sample trace data represent a block of 1000 bytes Each header location is identified by a number 1 120 Eavesdropper expects to see a header at the beginning of each trace 240 bytes followed by a data block length dependent on number of samples The header contains the following information at the designated word locations 16 bit Word Number Description 1 Datatype 0 Raw field files 1 gather 2 stacked 3 Record order R
17. Vr 7 i 2 828 BS AE UAR A EEF AA Et t IRA 4 V d Y A A gt 7 7 V A 2 2 oat Y 3 m AML 52 42725 Y 2 eam KA 2 x 1 x v A x 25 7 222204 2 ue Ao n rS 3 2 2 aX gt VACUUM 2 V Md 7 S 273 figure 10 figure 11 3 Time Spatial Varying The velocity function we choose for this data set is listed below CDP LIMITS TIME ms VELOCITY m s 200 232 0 0 50 1850 50 100 2600 100 135 2900 135 250 3000 232 246 0 0 50 1850 50 100 2450 100 135 2900 135 250 3000 246 253 0 0 50 1850 50 100 2450 100 135 2900 135 250 3000 253 278 0 0 50 1850 50 140 1900 140 250 2250 278 290 0 0 50 1850 50 100 2450 100 250 2900 The above described velocity function needs to be applied to the sorted datum corrected data In order to accomplish this a normal moveout operator needs to be used The batch processing file nmot dek used to apply the nor
18. appear as multiple broad band coherent reflection events very dissimilar from the original refraction wavelets on the raw field data The appearance of refraction energy on CDP stacked data can many times entice the creative interpreter and or geo contractor into a fatal pitfall Unmuted refracted energy from a subsurface layer that varies in depth can appear to be a structurally significant coherent reflection event on CDP stacked data This illusion on stacked data results from the changes in the critical refraction distance and time intercept as the depth to the refracting interface varies This pitfall many go unnoticed on some data sets since in some geologic conditions stacked refractions many be representative in a gross sense of actual shallow structure Such stacked refractions typically have lower frequency than shallow reflections Any editing operation that requires the defining of a time window or beginning and or ending points for the zeroing of data will require the defining of a taper length TAPR The taper is intended to allow the gradual attenuation of the trace amplitudes to zero without generating an abrupt truncation point and the associated apparent high frequencies If the trace is filtered at any time in the processing flow without a taper the abrupt truncation of signal resulting from a mute will produce a complex sine function with maximum amplitude at the truncation point decaying to near zero within the muted zone T
19. data set gt Enter ending record number default 32000 gt Do you want auto screen dump 0 No 1 Yes default 0 gt Enter vertical display size in inches second default data dependent gt Plot normal 2 0 off 1 default 0 gt Enter normalize scan delay in ms default 0 gt Enter starting time of plot in ms 0 gt Enter trace spacing in trace inch default data dependent values over 24 degrade hardcopy gt Only applies to dplot program Only an option when plot normalization is selected NOTE data dependent means program calculates a value it considers to be optimum for the input data set 10 NOTE If at any time you wish to terminate the plotting process press the space bar The plot normal option increases the amplitude of each individual trace by multiplying each sample by a normalization constant independent of all other traces in the data set The amount of this whole trace amplitude increase can be different for each trace and is related to the difference between the largest amplitude sample in the trace and the maximum possible amplitude that can be displayed The normalize scan delay time designates the beginning of the amplitude scan window used to determine the multiplication factor for all the samples within the trace The utility of this option can be appreciated on data sets with abnormally high amplitude first arrival information Selecting the beginning scan time after a large amp
20. entered it will operate on the input data as requested in a batch type format STEP 789 DATAS The DVSCN routine is a module operating outside of dseis To generate a constant velocity stack for our sample data set the following series of commands and information must be inputted gt DVSCN lt enter gt KGS Banner and assorted descriptive information will be displayed Input CDP gather data file surf dat lt enter gt Scanning input data please wait 55 Input Data First CDP 223 Last 282 Record Length 250 ms This information is supplied to allow cross checking of what you know about your input data with what the programs finds in the trace headers Output VSCN data VSCN dat lt enter gt First CDP of CDP group 1 2 gt current 223 lt enter gt Last of CDP group 1 current 282 lt enter gt of CDP Groups to be processed gt 1 lt enter gt The number of groups to be processed often allows you to skip through the line and do velocity analysis on sequential groups of CDPs If we wanted to do velocity analysis in an area we deemed to have velocity problems or if we only wanted to do reconnaissance velocity analysis on certain portions of the line several groups of CDPs could be selected for analysis If velocity prob lems were present between CDPs 230 and 240 and between CDPs 260 and 270 a constant velocity stac
21. file name is the output file name the input data will be deleted and replaced with the edited output data gt gt gt gt END identifies the last line of this batch processing deck The actual bad trace edit file just created will look like the following gt gt start inpf dengland raw edkl 92 8 kill 1 1 12 12 kill 2 2 11 11 kill 3 3 10 10 kill 4 4 9 9 24 24 kill 5 5 8 8 23 23 kill 6 6 7 7 22 22 15 kill 7 7 6 6 21 21 kill 8 8 5 5 20 20 kill 9 9 44 19 19 kill 10 10 3 3 18 18 kill 11 112 2 17 17 kill 12 1211 16 16 kill 13 13 15 15 kill 14 14 14 14 kill 15 15 13 19 kill 16 16 12 12 kill 17 17 11 11 kill 18 18 10 10 kill 19 19 9 9 kill 20 20 9 9 outf edkl dat gt gt end The batch processing file you just built to edit bad traces now needs to be run through DSFIS to actually operate on the dengland raw In order to execute the edit job the following sequence is necessary gt DSEIS EDKL dek EDKL Ist The EDKL st file is a journal file created to document all significant infor mation associated with the operation of the EDKL dek file The edited data will be in the file named EDKL dat In order to see the effect of the editing on the raw data you should first use the dview routine Simply type gt return Answer the series of self explanatory questions as described in the preious dplot and dview section and check the format of the screen display If the display is not
22. model The closer to a true near surface velocity depth model you can define for the datum correction the less dependant you will be on iterative numerical statics routines and the more confidence you will be able to put in your depth calculation during the interpretation portion of your survey Determining the appropriate values to input into and the usefulness of the output of the SURF operation requires a thorough understanding of the geologic situation as well as certain physical properties of the near surface materials The input velocity function will be interpolated across the entire expanse of the defined profileline Elevation data input for each station location need only be relative within the survey absolute elevation data are not necessary If the datum was assigned above the highest topographic feature on the survey line the outputted static correc tion for each receiver and shot station will be necessary during the later interpretation stages to calculate depths If the datum was assigned within the subweathering zone time to depth conver sions can be made without knowledge of the amount of the static correction material removed This is because the relative elevation of the datum is known which in turn allows accurate deter mination of time values between the datum and event of interest The necessary datum elevation correction for our sample data set only modifies a small portion of the line The data were collected in a ri
23. nmot dek The following sequence of commands will initiate the processing gt DSEIS nmot dek nmot lst lt enter gt 64 As before the program will keep you advised as to its progress In order to see the results of the velocity correction use the view routine on the CDP gather If you need to have a hard copy then of course use of the plot routine will be necessary In most cases inspection on the CRT will suffice The velocity func tion we just applied will correct the reflection for non vertical incidence The result of this dynamic correction is quite evident when comparing corrected to uncorrected field files Figure 23 I SPECTRAL ANALYSIS The spectral characteristics of a raw seismic data trace are dependent on the acquisition parameters and equipment as well as the surface and subsurface characteristics of the site During the spectral analysis portion of the processing flow the frequency characteristics are determined with respect to the various types of seismic energy present on the data A frequency filter is then designed to enhance the favorable spectral characteristics and attenuate the undesirable ones The amount and type of spectral analysis necessary for a particular data set is totally dependent on that data set 1 Frequency vs Amplitude Depending on the data set the first step in determining the spectral characteristics of seismic data is to define the dominant frequency and bandwidth of the entire data se
24. not require a significant amount of thought or understanding of the physics of the matter way to determine the stacking velocity of particular reflecting events on multi fold seismic data 51 A constant velocity stack simply moves out and stacks all traces within gather at a pre determined velocity Generally a data set will be moved out and stacked at each of a consecutive group of velocities in some cases as many as 20 The analysis technique at that point simply involves visual inspection to determine which velocity is best for a particular reflector The velocity can and many times does vary both horizontally and vertically time The resultant velocity function is defined by CDP time and velocity Multiple reflectors should be identifiable at most CDPs on seismic reflection data collected in a geologic environment conducive to the propa gation and recording of reflection energy The velocity function for each CDP will have an optimum stacking velocity paired with the appropriate time window for each identifiable reflector The velocity function defined for a group of reflecting events at a particular CDP may not be appro priate for the same set of reflecting events at other CDPs across the profile as a result of hori zontal variability Stacking velocities must be analyzed using the constant velocity stack and inputting large groups of CDPs across the entire expanse of the line It is possible and many times advisable to anal
25. obtained about the near surface materials a batch processing file must be created to calculate the appropriate static values for each shot and receiver station This operation SURF does not apply the calculated shifts This routine calculates the appropriate shift and updates the trace headers In order to apply this static shift a STAT 1 operation must follow the SURF operation This can be accomplished either in the same or a separate batch job file The following sequence of entries represent a batch processing file to calculate and apply datum corrections to our sample data set Line 1 2 Description gt gt 5 see for description of gt gt START SORT dat see for description of INPF SURF 100 SURE identifier calls up the terrain correction subroutine from within DSEIS The required input value relates to defining the datum The value of the datum is arbitrarily assigned as 100 In this case think of the data as representing absolute elevation relative to sea level 105 770 90 alvf defines the average velocities and their associated depths For the sample data set we are defining a single velocity across the entire line and therefore the first entry identifies the maximum surface elevation 105 of the defined velocity function The second entry 770 defines the average velocity from the maximum surface 105 to a maximum depth identified by the third entry
26. of data processing procedures are contained within the sub program called SEIS The program SEIS was designed to operate in a batch processing mode requiring an input job file and an output list file The input job file contains all the operation identifiers CAUTS EDKL SORT etc along with the appropriate user assigned parameters The output list file contains all the significant steps and processes completed during the executed batch job The list file also contains bits of information concerning processing time any abnormalities in the flow any error messages etc Operation of Eavesdropper with the assistance of this manual requires a general knowledge of what seismic reflection is as well as a working knowledge of your computer and the MS DOS operating system See information that came with your computer A GENERAL CDP SEISMIC DATA PROCESSING FLOW The goal of digitally manipulating seismic data is to allow the maximum amount of geologic information to be extracted from a minimal amount of data with a minimal amount of effort The processing of seismic data involves a series of steps Each seismic processing step generally has a prerequisite step or operation This means a basic processing flow must be used to effectively process seismic data The exact processing flow you should use for your data set depends mainly on two things 1 the overall quality of your data and 2 the particular information you intend to extract from the final pr
27. of the default values See line 10 SNSN lt 3191 gt 110 1261 1 The SNSN definition permits the assignment of field geometries to large numbers of sequential field files possessing identical shot and receiver patterns as well as source offsets and bad traces The lt gt are treated as entry parameters and therefore need to be separated from other entries by aspace Here we are defining the shot patterns for SSN 3 through 19 which related to shot locations 110 to 126 Both shot location and SSN s increment by 1 as the source and receivers move down the survey line Therefore the first set of bracketed values represent the SSN s and their incrementing while the second set of bracketed values represent the asso ciated shot locations and their incrementation pattern The information within the first set of brackets is equivalent to the first entry on a normal SN definer The second set of bracketed values which are equivalent to the second entry on the normal SN definer parameters beyond the first two of the SNSN definers are identical to those of the ordinary SN definer SN 201271001124 This shotpoint definition has a couple extra entries in comparison to the previous sn definers The SN definition s seventh value is the first dead trace channel number and eighth value 24 is the last dead trace channel number From the field notes for our sample data the 19th and 20th shots were recorded at the same location In order to retain e
28. programs Imagine the resulting seismic section if the programmer of the software designates a particular trace header location as the source station number and the seismo graph manufacturer has designated that location as the receiver location A simple way to think of both the use and organization of a trace header is to compare it to a business letter A business letter will generally have two main parts first is the letterhead and second is the body of the letter The critical part of the letter of course is the body It contains the significant information the information that makes it different from any other letter The letterhead on the other hand is basically the same for every letter sent out by the business The only things that change within the letterhead are date and office of origin within the business The letterhead contains all the information necessary for someone to determine the business name section address and date of the letter A seismic data trace in digital form can be thought of in a similar way The trace header serves similar purpose as the letterhead The data itself are equivalent to the body of the letter Most but not all operations use the header to obtain key information about the data contained within each trace Some operations will update the header with information others will simply use the information to physically change the data set according to the prescribed operation and designated parameters
29. refractions direct wave or air wave SIGNAL TO NOISE S N The signal to noise value is the ratio of the whole trace average amplitude and the average amplitude of signal in the noise window A S N value of 1 will retain any trace with a whole trace average amplitude equal to or greater than the average amplitude of the signal in the noise window Experimentation with this routine is the best teacher A few test runs varying the S N value for a given noise window will give insight into both the utility and the limitations of this routine A batch processing file for doing automatic editing should look about like the following Line Description 1 gt gt START see manual edit EDKL for details on gt gt start 2 INPF dengland raw see for details on INPF 18 3 AUED 20 0 28 111 AUED calls the auto edit sub routine The first requested parameter is the noise window Here 20 ms is used indicating that no source generated signal arrives on any trace before 20 ms The second requested parameter is the signal to noise ratio S N value A signal to noise ratio of 0 28 means any trace not possessing an average whole trace amplitude at least 0 28 times the average pre first break amplitude will be flagged The third requested parameter instructs the program to print 1 or not to print 0 the calculated average whole trace signal to noise ratios S N in the list file The fourth requested parameter instructs the progra
30. satisfactory make the appropriate changes to either view cfg or to the responses provided to the view routine questions Once an acceptable format has been obtained make the appropriate changes to plot cfg and then type gt DPLOT lt return gt Answer the dplot questions with values similar to those used for the pre vious dview routine Using the normalization on plots will improve the usefulness of the display Once the bad trace editing is complete each field file should be missing the trace or traces you selected to remove Figure 2 The trace is not displayed on the 16 figure 2 plot because the dead trace flag trace header word 15 has been tripped in the trace header The file size will not change until the data are sorted or re sorted at which time the trace will be completely removed from the data set The bad trace is still present in the data file after trace editing but not visible on the plot Whole field files should not be removed at this time The sorting operation which will be discussed later requires all source geometries be input in an uninterrupted file sequential format This means that once the data are sequentialized during formatting they must remain in that order without any missing files until the assignment of source and receiver geometries is complete At the time the source geometries are identified snor snsn entire bad files can then be removed This is not a problem when plucking ou
31. section allowing the most realistic geologic interpretation INTRODUCTION mec AW FE nemore AMP LE DATAS Eme wee The raw data include 20 field files chopped to 250 ms of record length at a sample interval of 1 2 ms The data were acquired with an EG amp G 2401 seismo graph processing 24 channels single 100 Hz geophones a 12 gauge buffalo gun both the source and receiver intervals were 2 5 m and 200 Hz analog low cut filters Each step in the general processing flow followed in this manual will use the England data set as the example The field notes as well as the formatted data are included with or within this manual B TRACE HEADERS Next to the data itself the trace headers are the most essential part of digital seismic data All acquisition information essential to future seismic data processing as well as to update information derived from intermediate processing operations is organized and stored within each trace header The organization of trace headers is dependent only on the imagination of the pro grammer of the software and manufacturer of the seismograph The particulars with respect to size and organization of the trace header its location with respect to the rest of the data set as well as the organization and size of each sample of data within the trace itself is what is commonly refer red to as seismic data format The format of seismic data is critical to the effective operation of seismic data processing
32. stat dat gt gt end Either the stat 4 or surf operation are valid ways to correct for variablity across the seismic line Which operation you use with various data sets other than the sample provided with this manual is dependant on the amount of static other information available and probably most important which operation you are most comfortable with H VELOCITY ANALYSIS The compressional wave P wave velocity through an elastic media is related to Lame s constants divided by the density of the media Normal moveout NMO velocity information derived from the curvature of reflection events on seismic data is used to correct each reflection wavelet for a non vertical incident travel path The NMO velocity derived in this fashion can also be used as a rough approximation for the average velocity allowing a first guess approxima tion of reflector depth This correction for non vertical incidence reflectors on seismic data NMO correction is one of the most critical parts of the preliminary processing flow The summing of multi fold data is meaningless unless all traces are corrected for their particular source to receiver offset to effectively simulate vertical incident energy i e source and receiver located at the same point on the earth s surface Determining the appropriate stacking velocity loosely NMO velocity can be accomp lished in a variety of ways A constant velocity stack is the most commonly used probably because it does
33. therefore invisible to the eye If the amplitude of the unwanted noise is low in comparison to signal at equiv alent times on other traces occasionally the multi trace stacking process necessary to generate a CDP stacked section will suppress noise to an acceptable level Also many times various filtering 24 c c c c c gt c gt x 6 x qoa _ 46 MEN n 1 a A M m a aan A P figure 4 operations discussed later in the processing flow can attenuate noise that has unique spectral and or phase characteristics It must be kept in mind that removal of noise and enhancement of seismic signal to allow an accurate geologic interpretation is the ultimate goal There is no replacement for good judgement during the preliminary stages of seismic data processing STEP 5 Tuc CURE Wk DATA ROR ESR ERR BRR REG The plot of the first arrival muted data needs to be carefully studied and appropriate time and trace pairs selected to remove the air coupled wave Once the desired mute windows on the data have been defined and the trace time pairs recorded the mute batch file needs to be created The following sequence of entries is appropriate for the sample data set Line Description
34. values previously deleted If your hand generated stacking chart is correct it will also include the dead or bad traces It will be beneficial at this point to generate a hard copy plot of the entire sorted data set This will allow you to cross check your stacking chart with your sorting table printout as well as the actual sorted data Many times problems not obvious on a sorting table will be quite obvious on a hard copy plot of CDP sorted data STEP 6b OB eee ANE PB 75 54 The principal higher amplitude reflections identified on our raw field file Figure 1 are still interpretable on the gathers of CDPs 245 and 246 Figure 9 These two gathers represent the subsurface points located beneath station loca tions 122 1 2 and 123 Careful examination reveals the zeroed out portions of the traces previously occupied by the air coupled wave The traces are ordered within each CDP according to original source to receiver offset distance trace header word 19 as defined by the sort The location of the traces within each CDP are also identified by a trace header word 14 Header word 14 designates the order within the CDP gather The numbering of traces within the CDP is from left to right with the far left trace of CDP 245 possessing a 1 at trace header location 14 while the trace on the far right possesses an 11 at trace header location 14 Header word 14 is totally unrelated to header word 8 original seismograph c
35. where it is in the processing sequence described in EDMT dek The EDMT Ist file is simply a journal file keeping track of the same information that you will see on the screen during the actual running of seis To see what has been saved in the journal file use the MS DOS type command It is always good to at least briefly look at all the data after any operation To see what the effect of your mute has been on the input data EDFM dat use the view routine as described in previous sections and then if there is a need to carefully inspect a hard copy use the plot routine as described in the display section Our sample data set was muted to remove the air coupled wave Any time a mute is applied to seismic data it should be as well defined and tight as possible This is to avoid removing subdued signal The muting process zeros samples and once a sample is zeroed the information contained within that sample has gone to the great bit bucket in the sky Care should be taken when defining and applying a mute The plot of SSN 5 clearly shows the effect of the mute as well as the narrowness of our defined mute window Figure 5 F SORTING GATHERING INTO CDP FORMAT The way data are organized for display and analysis is the heart of any seismic reflection program Having flexibility to look at data in a variety of ways whether according to receiver location or common subsurface point is critical for future digital enhancement as well as discrimina
36. 0 21 10 22 10 23 10 24 MO NOD AD 00 00 I HL ON Ov CA CA 1 4 Co DY b2 b2 b2 b2 P2 4 ROO 4 4 RR Qo OUO 4 BW QN Ov tA CO 270 T 19 12 18 14 17 16 16 18 15 20 14 22 13 24 271 6 19 13 18 15 17 17 16 19 15 21 14 23 272 6 19 14 18 16 17 18 16 20 15 22 14 24 273 5 19 15 18 17 17 19 16 21 15 23 274 5 19 16 18 18 17 20 16 22 15 24 275 4 19 17 18 19 17 21 16 23 276 4 19 18 18 20 17 22 16 24 277 3 19 19 18 21 17 23 278 3 19 20 18 22 17 24 279 2 19 21 18 23 280 1 19 22 18 24 281 1 19 23 282 1 19 24 gt gt The output from the tabl option will generate a table similar to the above table which contains the SSN trace pairs with their associated RCRD CDP and appropriate fold Checking this table program generated against the one you constructed Figure 8 will help verify the correctness of the sort file as well as assure you the data geometries are input into the trace header correctly It should be kept in mind that the stacking chart generated by the sorting operation tabl does not check the real input data to identify traces designated as bad Therefore the SSN trace pairs output with its associated CDP number will include
37. 02400 MUTE 19 1 57 70 24 222 235 2500 t Not appropriate for this data set used only as an example The mute sequence defined by lines 6a through 6c will operate on only record SSN 5 Mutes defined for records 4 and 6 will terminate the inter polation process The operation of the mute in regard to interpolation is very similar to farm as defined in the previous section EDKL Any traces within this record greater than 24 will be muted as 24 In order to stop the interpolation beyond trace 24 a zero mute window needs to be defined for trace 25 Line 6d defines a mute exactly the same as 6b except it terminates the muting of traces beyond trace 24 OUTF EDMT dat see for description of OUTF gt gt see for description of gt gt END At this point you need to exit your text editor according to the instruction included with your text editor software 27 In order to inspect the batch processing file you just constructed to surgically mute the air coupled wave of the sample data set type the following at the system prompt gt TYPEEDMT dek return gt gt start inpf edfm dat edmt 92 8 tapr 10 mute 5 1 57 70 24 222 235 outf edmt dat gt gt Now order to run the job through the seis program you need to enter the following at the system prompt gt DSEIS EDMT dek EDMT lst return As before while the program is running it will keep you abreast of
38. 2 time ms velocity m s 0 1800 constant velocity zone 45 1800 linear transition 60 _ 26004 2019 linear transition zone 90 2600 constant velocity zone 110 2900 linear transition zone 135 3000 constant velocity zone 3000 end of record As you can see the velocity function as it is defined here is not iden tical to the velocity function determined from the constant velocity stacks listed in the table at the start of section H3 The reason for this is related to the interpolation process operating both vertically in time and horizon tally in space Analysis of the data suggested specific time windows where the stacking velocity seems to be relatively constant and other time win dows where simply from a physically realistic point of view the average velocity through the rock must be changing at a significant rate With this in mind the program must be instructed as to which time windows have relatively constant velocities and which windows have a significant amount of change unit time depth The assigning of a realistic velocity function that possesses significant variability in time and space is a skill knack that will come with time and experience 63 10 11 235 45 1850 60 2450 90 2450 110 2900 135 3000 The velocity function for this definer is for CDP 235 The vertical time interpolation process is identical here as it was with lin
39. 2 X 95 99 122 surface station The sorting routine doesn t care what the actual numbers input for the source 108 and receiver 115 locations are it only looks at relative differ ences What that means is the pn definition on line 5a is identical to the pn definition on line 5c It is not necessary to define existing station num bers Most of the time it is advisable to identify real station numbers just to avoid confusion if problems crop up later This definition of pattern is exactly the same as the one defined on line 5 The station numbers of the shot and receivers need only be relative to each other and do not need to be related to the actual values used in the field Any arbitrary numerical representation will work 1 12 S 13 24 trace channel 101 112 116 127 surface station 38 This definition is of a split spread pattern This definition does not apply to the acquisition parameters and geometries being defined here but is in cluded as an example in case you encounter a split spread source receiver geometry Briefly this is the first pattern defined 1 the shot station number is 114 the first trace channel station number is 101 there are 12 consecutive trace channels defined across 12 sequential surface station locations then a 3 station skip to station
40. 37 103 100 3 se 138 106 100 6 se 139 100 se 140 99 se 141 99 6 se 142 100 6 se 143 100 nmot 0 6 velf 230 45 1850 60 2600 90 2600 110 2900 135 3000 velf 235 45 1850 60 2450 90 2450 110 2900 135 3000 velf 250 45 1850 60 2300 90 2300 110 2900 135 3000 velf 255 45 1850 60 1900 110 1900 150 2250 velf 275 45 1850 60 1900 110 1900 150 2250 velf 280 45 1850 60 2450 90 2450 110 2900 filt 125 400 1 0 60 scal 50 stak 1 outf stak dat gt gt end This batch file is on the demo disk and can be run using the following command line DSEIS PROSS dek PROSS Ist II FURTHER PROCESSING ADVANCED TECHNIQUES Other operations available in Eavesdropper include deconvolution surface consistent statics residual statics f k migration and f k filter In summary a wide variety of processing flow options is available with Eavesdropper after the data have been sorted Inexperienced analysts are encouraged to refer to books on seismic data processing and to experiment with the program prior to processing data that are critical to some project The technical user s manual will provide the necessary mechanical guidance to work through the many program options However it is not appropriate to use the program without some knowledge of why various processes are used 84 Suggested Reading Mayne W H 1962 Horizontal data stacking techniques Supplement to Geo physics 27 p 927 938 Robinson E A and Treitel S 1980 Geo
41. 7 gt lt enter gt Number of bits to display is simply a whole trace normalization routine that rounds to the nearest bit In other words if 7 bits are selected any 7 bit sample will be represented with full deflection of the wiggle trace Whatever gaining was necessary or to produce this full deflection in the 7 bit sample will be equally and uniformly applied to all other samples If the largest word in the data set is 8 and 7 bits is selected as the display amplitude all samples with a word size of 8 will have the most significant bit clipped off if the largest word size is 9 the two most significant bits will be clipped off etc This parameter is necessary due to potential mismatches in resolution between data and display hardware printer and CRT At this point you can use the arrow keys to move the cross hairs around and the f1 function key to select the appropriate time offset pairs You should get a value of approximately 2400 m s for the NMO velocity of the reflection at 75 ms which equates to a depth reflector of approximately 90 m 2 Constant Velocity Stack Once you have established a ballpark velocity for the major reflecting events you should be able to efficiently run the constant velocity stack routine VSCN to fine tune the velocity function for this data set The VSCN program is pseudo interactive that is it will ask you a series of ques tions interactively and then once all the appropriate information is
42. 90 We have determined the following near surface configuration 47 T 4a g 105 100 m datum 95 770 m s ov Sea level For a case with multiple depth velocity pairs 1000 1300 950 1600 900 alvf operation defines the velocity model for the near surface material down to a depth necessary to correct all station locations to the defined datum The values requested for the ALVF operation are the surface ele vation 1000 of the defined velocity function the average velocity 1300 from the surface 1000 to the elevation defined by the third numeric value 950 then the average velocity 1600 from the surface 1000 to the elevation defined by the fifth numeric value 900 This sequence of aver age velocity depth pairs continues until your entire velocity function is defined An important note at this point the velocities defined here ALVF are AVERAGE VELOCITIES FROM THE SURFACE TO THE DEFINED DEPTH The program will calculate interval velocities for the defined intervals 1000 950 950 900 This program works most effectively with uphole downhole check shot velocity information 1000 E Vave 1900 TnT 950 8 900 t Examples not applicable to our sample data set 5 SE 126 97 100 se operation defines each station location and its associated source and receiver elevations The defined values are station location 108 sour
43. Eavesdropper Tutorial SEISMIC REFLECTION PROCESSING DEMONSTRATION USING EAVESDROPPER by Richard D Miller and Don W Steeples KANSAS GEOLOGICAL SURVEY 1930 Constant Avenue Lawrence Kansas 66047 3726 Open file Report 91 27 July 1991 TABLE OF CONTENTS CDP PROCESSING WITH EAVESDROPPER FOR THE NOVICE A GENERAL CDP SEISMIC DATA PROCESSING FLOW B TRACE HEADERS C DATA FORMAT D DISPLAY E EDIT 1 Manual bad trace edit 2 Automatic bad trace edit 3 First arrival mute 4 Surgical mute F SORTING GATHERING INTO CDP FORMAT G ELEVATION CORRECTION DATUM ASSIGNMENT H VELOCITY ANALYSIS 1 Interactive velocity picking 2 Constant velocity stack 3 Time spatial varying NMO D SPECTRAL ANALYSIS 1 Frequency vs amplitude 2 Filtering p AMPLITUDE BALANCING 1 Automatic gain control scale K STACK FURTHER PROCESSING ADV ANCED TECHNIQUES Step No Introduction 1 2 3 3a 4 6a 6b 8a 8b 8c 9a 9b 10 10a 11 TABLE OFEXAMPLE DATA STEPS Operation Data and format Loading data on computer Plotting raw field files Bad trace editing Manual editing procedure First arrival muting Surgical muting Stacking chart Building sort deck Data after sorting Datum correction Interactive velocity analysis Constant velocity stacks Picking appropriate velocity Moved out field file appearance Spectral analysis Reflection information Batch processing file Analysis of spectral plots Applicati
44. TR SN TR SN TR SN TR 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 0 47 1 1 00 OY CA CA HP S 6 59 See ee D ON OY CA CA 4 PR r2 T 0 1 0 0 0 tA 17 10 17 11 17 12 17 13 17 14 17 15 ee T C 9 1 OV CA CA 4 B9 1 1 1 1 00 1 00 1 1 00 1 1 1 14 8 15 7 15 8 16 7 16 8 16 9 16 10 16 11 16 12 16 13 16 14 16 15 16 16 16 17 Qt t PR OU ES Ob
45. VIEW lt return gt Answer the series of self explanatory questions as described in display section and check the format of the screen display If the display is not satis factory make the appropriate changes to either view cfg or to the responses pro vided to the view routine questions Once an acceptable format has been obtained make the appropriate changes to plot cfg and then type gt DPLOT lt return gt and answer the questions in a similar fashion as you view responses that resulted in the appropriate display The first arrival mute defined by the previous batch file will result in a mute on all files of the sample data set First arrival information on file 5 will begin at 30 ms on trace 1 and 65 ms on trace 24 with a 10 ms taper Figure 4 This field file is displayed identical to the file in Figures 1 and 2 allowing direct comparison of before and after 4 Surgical Mute The final editing step involves the surgical removal of bad trace segments Noises resulting from the air coupled wave electronic interference other than power line frequen cies and ground roll are generally constrained to isolated portions of a trace High amplitude noise obviously dominating a well defined time window should be removed However care need be taken when removing these low S N portions of traces since significant seismic signal could be pres ent on unprocessed data at an amplitude level below the dynamic range of the plot and
46. ample data set The primary task associated with sorting your data relates to the assignment of geometries and parameters Trace header information plays a significant roll in this operation The trace header words most important to commit to memory are Header Word Identifies 6 the field file number name under which this trace was stored in after it was recorded 8 this trace s number within the field file it was collected under also is equivalent to the channel number this trace was recorded 12 at this trace s corresponding CDP number usually about twice the associated station location 14 the order of this trace within the appropriate CDP 19 distance this trace is from the source 86 station number of the receiver that recorded this trace 87 station number of the source associated with this trace 92 the source sequence number of this trace which sequentially relates it to individual field files collected for a particular profile A helpful aid in defining the geometries and to double check the accuracy of your field notes is a stacking chart Figure 8 The layout of the stacking chart by design allows visual correlation between the field notes and the sort batch file It will simplify both visualizing and defining geometries for particular Source Sequence Numbers SSN STEP 6 REN AMPLE DATA 55595055 The upper portion of the chart built for our example data set Figure 8 defines station locations on the x axis a
47. as an example If only one file of several is to be first arrival muted farm the series of entries on lines 6a 6b and 6c would be necessary The linear interpolation process is automatic The only way to stop the interpolation is to define 0 mute times just before and just after the defined mutes The farm defined by lines 6a 6b and 6c will mute file SSN 3 only with trace 1 muted from time zero to 30 ms trace 2 from time zero to 32 ms etc out to trace 24 which will be muted from time zero to 45 ms If you wish to stop the mute process after trace 24 therefore retaining all the information in traces 25 to the last trace 48 96 or whatever the number of traces on your seismograph line 6d would be entered in place of line 6b OUTF EDFM dat see for description of OUTF gt gt see for description of gt gt END In order to display on the screen the previously defined first arrival mute batch process simply enter gt TYPEEDFM dek return gt gt start 23 inpf edkl dat edfm 92 8 tapr 10 farm 1 1 30 24 65 outf edfm dat gt gt Now to run the previously defined first arrival mute you need to type the following gt DSEIS EDFM dek EDFM LST As before EDFM LST is simply a journal file The muted data will be in the file named EDFM dat In order to see the effect of your mute on the input data EDKL dat you need to use the dview routine Simply type gt D
48. at stak 1 outf stak dat gt gt end Your CDP stacking file operates on the scaled data as follows gt dseis stak dek stak Ist lt enter gt Plotting your output is a must especially since this is the conclusion of this basic processing flow The plotting parameters used on stacked data are very dependent on interpretation preference That is everyone likes to look at stacked seismic data displayed with particular parameters The larger the num ber of traces per inch the more apparent coherency on the stacked section Also the larger the number of seconds per inch the less apparent subtle variation in reflector depth and the lower the apparent resolution Simply it is good to 81 experiment with the plotting parameters on the finished stacked section until they are aesthetically pleasing to you or your interpreter The final stacked data set you have just plotted should be quite similar except for possible differences in the resolution of your plotter to the one dis played in Figure 20 Simply to show the true power of Eavesdropper and the batch processing mode of operation the following batch processing file could have been set up for the sample data set at the very beginning and executed The input to this batch job is the raw formatted data and the output is a stacked section all in one deck The appropriate sequence would look like gt gt start inpf dengland raw edkl 92 8 kill 1 1 12 12 kill 2 2 11 11 kill 3
49. at least three of the reflecting events within a single window the window size should be somewhere around 50 ms So 50 ms will be used for a first pass trial scaling window Basically the scale window must be small enough that subtle reflecting events don t get overpowered by the more high amplitude events yet large enough that localized comparisons of relative amplitude between reflecting events at various times can be made Final interpretation of the stacked data must be made with the AGC processing parameters taken into consideration A couple of warnings 1 An inappropriate AGC window can generate artifacts on stacked seismic data Most notable are the effects of a window that is around twice as long as the time difference between the high amplitude first arrival and ground roll or air wave information Due to this longer window the contribution of the lower amplitude energy located between the first arrival and ground roll or air wave is insignificant in comparison to the effects of the first arrival and ground roll or air wave The resulting stacked data could possess a high amplitude band of information with spotty coherency that is nothing more than stacked ground roll or air wave 2 An AGC is most effective when the amplitude of noise is high on a few traces relative to signal at an equivalent time on other traces In such a case not using the AGC on unstacked data could result in data with a significantly lower signal to noise ratio than
50. ation disk 1 seismic data extension raw or dat 2 sample batch files extension dek 3 executable files extension exe and 4 assistance files extension hlp or cfg The only files that can be displayed with the DOS type command are the sample decks and set up files To minimize confusion while processing this data it is advised that two sub directories be created The first sub directory should contain the executable files extension exe and should be named something like eav The second sub directory should contain the sample batch files assistance file and data files and called something like demo The path of your computer will need to be modified so the executable files can be called on from the demo directory Once the files on the floppy disk have all been loaded into the appropriate directory you are ready to proceed through the manual I PROCESSING WITH EAVESDROPPER FOR THE NOVICE This document is designed to demonstrate the operation of Eavesdropper by providing step by step detailed instructions and explanations on seismic data processing from raw data to brute stack NOTE This document is separated into large type representing processing steps and smaller type indicative of explanation and background information The format of the text in this document was specifically designed to aid in identifying 1 responses and information supplied to the user by the program upon request 2 informa
51. ay Single Phones 110 003 77 117 10 Source Receiver Geometry 111 004 78 118 5m 112 005 79 119 3 02 Source receivers 113 006 80 120 4444444444444 station location 4 114 007 81 121 Amp 1131517191 11 13 15 17 19 21 23 File 115 1008 82 122 airplane noise 2 4 6 amp to t2 14 16 18 20 22 24 116 009 83 123 airplane noise 2 Test 117 010 84 124 5 Oscil F2 118 011 85 125 m Floating Pt Amps automobile noise ai 09 17 7777 119 012 86 126 ating lt Gains 120 013 87 1127 121 014 88 128 Amp Scan Deay 0 Damper Box In 122 015 89 129 472 Filters High Cuts _200 Hz LowCults 200 Hz 123 016 90 130 12 L1 60 Hz Notch Sample Int 1 2 ______ d 124 017 91 131 11 Alias of Samples _1024 Recordlengih 512 msec 125 018 92 132 10 Rec Start Delay p Amps 155 IAG 126 019 93 133 9 Control Level VA Temp 23 second shot in hole Wind 0 5 KmPH 126 020 93 133 9 Soil Conditions dry sod figure 7 field file 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 108 109 110 2X 3x source station location SSN 108 1 109 2 110 3 11124 112 5 113 6 11457 115 8 116 9 117 10 118 11 119 12 120 13 121 14 122 15 123 16 124 17 125 18 126 19 127 20 figure 1H 4X fold 223 1
52. ce elevation 97 and receiver elevation 100 The receiver elevation need not be defined if it is the same as the source elevation A shot elevation defined at a station where no shot was recorded will be sorted out by the program and properly handled The program interpolates horizontally and vertically between defined values and therefore stations 108 through 125 are not defined since they are all the same as 126 SE 127 100 SE 136 100 48 8 57 137 100 3 9 5 138 100 6 10 5 139 100 11 SE 14099 12 SE141 99 6 13 SE 142 100 6 14 5 143 100 15 STAT1 STAT operation applies static correction values defined in the trace headers These static header words are updated by other operations auts surf or user input values The STAT operation can be executed within the batch job where the actual static correction values are determined or later in a separate batch job either independently or with any other opera tion s The only required input value 1 designates the type of operation used to define static shift values 16 OUTF SURF dat see EDKL for description of OUTF 17 gt gt END see EDKL for description of gt gt END At this point you have completed the building of an executable file for the generation and application of static shifts to sorted data for the correction of elevation and near surface irregularities across the profile In order to quickly check to make sure all the values and operations you just
53. ch Each trace has an allotted area of 1 If line 7 is 100 percent the trace may wiggle as much as 1 If line 7 is say 200 percent then the trace wiggle may be as wide as 2 Values of 150 to 200 generally give pleasing results GAP between field records The plot cfg file has strong similarities to view cfg The text file you just opened has 8 lines each with its associated default value Default Value 0 0 120 144 100 2 Description density 1 high density 0 low density Default is 0 dots line on the attached printer generally either 960 or 1632 Vertical resolution dots inch on the vertical scale time full scale on standard printers is 120 Horizontal resolution dots inch on horizontal scale distance full scale on standard printers is 144 This parameter controls the application of a constant gain in decibels to the data before plotting dplot style 0 variable area wiggle trace 1 wiggle trace Tells dplot what percentage of the allotted trace spacing the trace may occupy For example assume that the trace spacing is 10 per inch Each trace has an allotted area of 1 If line 7 is 100 percent the trace may wiggle as much as 1 If line 7 is say 200 percent then the trace wiggle may be as wide as 2 Inch Values of 150 to 200 generally give pleasing results Controls the number of blank traces inserted between field records After editing plot cfg and or view cfg return to your w
54. ction In areas with alluvium over bedrock generally a direct wave velocity can be used to determine the appropriate time correction In areas with either a thin layer of weathered material or no weathered material at all overlying outcropping or subcropping units a more detailed velocity function is necessary to reconstruct the material between the present ground surface and the datum For shallow reflection data the datum represents a pseudo ground surface composed of actual material beneath topographic highs and the virtual material used to compensate for the distance between the ground surface and topographic lows and the datum The sum is the time between the ground surface and the allowing depth calculations to bedrock or other layers 45 The most inaccurate part of the datum correction for shallow surveys involves the assign ment of depth velocity pairs to each surface station The best way to determine these values is through uphole downhole surveys in strategically placed boreholes The Eavesdropper software will generate a near surface model along the survey line by incorporating all the uphole deter mined velocity and depth information with the spatial distribution of the holes Without uphole data the only information available to approximate the near surface velocity depth structure is direct and refracted wave information In some cases the values obtained using this information are sufficient to generate an acceptable near surface
55. d with the low velocity material in the near surface velf 230 45 1850 60 2600 90 2600 110 2900 135 3000 The velf definers are grouped according to CDP locations The velocity function is input in time velocity pairs The first entry after the VELF definer is the CDP number 230 which this velocity function is defined for The second 45 in units of ms and third 1850 in units of m s entries identify the first time velocity pair of the velocity function at this CDP The program uses the first pair to define a consistent velocity win dow from 0 ms to the first identified time 45 The velocity function is then interpolated between 45 and the third entry 60 which is the time value for the second time velocity pair with 1850 m s defined as the velocity at 45 ms gradually changing sample by sample until at 60 ms the NMO velocity is 2600 m s This process continues in exactly the same fashion for all the defined time velocity pairs down to the bottom of the record In this case the velocity is constant 2600 m s between 60 and 90 ms Then between 90 and 110 ms the nmo velocity changes from 2600 fm s to 2900 m s Finally between 110 ms and 135 ms the nmo velocity is defined to change from 2900 m s to 3000 m s where it remains constant to the end of the record It should be noted that the last time velocity pair initiates a constant velocity correction that begins at the last time identi fied and continues to the end of the record 6
56. d by record number and trace number in the same fashion any playing card can be identified by suite and value The program allows you to select any trace header word to be the record number portion as well as any trace header word for the trace number portion of the record trace pair In this case we wish to use Source Sequence Numbers SSN assigned during formatting header word 92 as the record number portion and the trace numbers within each field record header word 8 which is the seismograph s actual channel numbers as the trace number Assigned by seismograph during acquisition KILL 1 1 12 12 KILL is a command operation that identifies which trace s within the specified records are to be removed In the above case trace 12 record number 1 will be removed 14 KILL 22 1111 KILL 3 3 10 10 KILL 4 4 9 9 24 24 Traces 9 and 24 of record 4 will be removed KILL 5 5 8 8 23 23 KILL 66772222 KILL 77662121 KILL 8 8 5 5 20 20 KILL 9 9 4 4 19 19 KILL 10 10 3 3 18 18 KILL 11 11 2 2 17 17 KILL 12 12 1 1 16 16 KILL 13 13 15 15 KILL 14 14 14 14 KILL 15 15 13 13 KILL 16 16 12 12 KILL 17 17 11 11 KILL 18 18 10 10 KILL 19 19 9 9 KILL 20 20 9 9 OUTF EdKL dat OUTF identifies the destination file name of the edited data The file name can be any MS DOS acceptable name with or without extension The output file name can be the same as the input Of course if the input
57. dded together and their amplitudes divided by 24 a disproportional amount of the final stacked data is from the contribution of the few traces on the inside with significant amounts of high amplitude ground roll and air coupled waves The best way to correct for this inequality is by applying a time varying trace by trace gaining function In the Eavesdropper package this operation is called SCAL meaning scale 76 ccc cc T ON st co st gt OO SM EL EE MR 4545 e SE OBI OB SENE SEM S amp P iO do 6 00 cos figure 18 Selection of the appropriate AGC window is at least partially qualitative requiring experience and a thorough understanding of the mathematics of the operation Assigning the AGC window to be about twice the repetition time of identifiable reflecting events is a rule of thumb that we have found to generally be effective STEP 10 MIRO ER In the case of our sample data set the strong reflecting events at approxi mately 75 100 120 and 150 ms seem to have a repetition time of around 25 to 30 ms This suggests in order for the AGC window to detect the presence of
58. e you would want to designate a negative one 1 incrementation of surface station locations as the fifth value input of the pn definition The correct pn definition for a reversed roll switch would look like line 5b S 24 1 trace channel 5b PN 1 108 138 24 1 X 108 115 138 surface station T 5c This definition indicates the surface station location of the shot was 108 trace channel 1 was 138 and incrementing of station locations is negative 1 with respect to the positive increasing trace channel numbers There fore channel trace 1 is surface station 138 channel trace 2 is surface station 137 channel trace 3 is surface station 136 etc Here the same source to receiver maximum and minimum offsets are defined source 108 mini mum offset 115 maximum offset 138 as on line 5a the exception of course being the decrementing surface station numbers with respect to the incre menting channel numbers This is a critical difference when the source to individual receiver distances are calculated later in this batch job the information necessary to determine exactly how to define the trace channel orders is contained within the trace header and is easily viewed using the trace header dump trhd routine previously discussed S 1 24 trace channel 19599
59. e 4 previous velf definer The significant thing to note here is the interpolation be tween CDP 230 and 235 The interpolation process is uniform between lines 4 and 5 with the velocity defined at 60 and 90 ms gradually changing at each CDP between 230 and 235 from 2600 to 2450 m s This will be true for all velocities defined at each velf definer The velocity will be interpo lated in both time and space VELF 250 45 1850 60 2300 90 2300 110 2900 135 3000 VELF 255 45 1850 60 1900 110 1900 150 2250 VELF 275 45 1850 60 1900 110 1900 150 2250 VELF 280 45 1850 60 2450 90 2450 110 2900 nmot dat see previous description of OUTF gt gt see previous description of 22END It is wise to list each deck after you exit your line text editor to ensure that the batch file has been created and saved the way you had intended Your file when displayed on the CRT should look like the following gt gt start inpf surf dat nmot 0 6 velf 230 45 1850 60 2600 90 2600 110 2900 135 3000 velf 235 45 1850 60 2450 90 2450 110 2900 135 3000 velf 250 45 1850 60 2300 90 2300 110 2900 135 3000 velf 255 45 1850 60 1900 110 1900 150 2250 velf 275 45 1850 60 1900 110 1900 150 2250 velf 280 45 1850 60 2450 90 2450 110 2900 nmot dat gt gt end Applying the defined velocity function making the dynamic correction for non vertical incidence requires the execution of the batch processing file we just created called
60. e are related to acquisition equipment and the settings of that equip ment Therefore to generate a CDP stacked section with relatively equal contributions from all traces summed within a particular gather some trace equalization is necessary It should be noted at this point that for some analysis techniques AVO amplitude vs offset retaining absolute amplitude with increased offset is critical Correction for spherical divergence is necessary but global trace equalization as is suggested for our sample data set would be detrimental to meaningful conclusions 1 Automatic Gain Control scale An Automatic Gain Control AGC should next be applied to the sample data to boost the amplitudes of the reflection information relative to the higher amplitude ground roll air wave and refraction energy The purpose of this scaling operation is to maximize the potential of the stacking process to enhance the reflection information This can be clearly visualized by plotting the raw data without any normalization or scaling Display with no normalization or scaling is called plotting relative This means that true amplitude information is preserved and present on the plot By using the plot cfg or the or keys during plotting you can boost the display gain uniformly for the entire data set until it is up to a desirable level Now observe the amplitude of the reflection event at 75 ms on all 24 of the traces It is quite clear that if all 24 traces were a
61. e line This way is very straight forward from an input point of view but it does require significantly more work prior to interaction with the pro gram DSEIS Described below is the appropriate way to correct your datum using the STAT operation and the sample data set Line Description 1 gt gt start see previous description of gt gt start 2 inpf sort dat see previous description of inpf 3 stat 4 stat identifies a time static operation The only entry necessary to do datum corrections is the first 4 The 4 simply indicates that the 4 option applies station corrections for both source and receiver 4 rece 137 0 3 rece identifies the trace header location that relates to the defined static shift The first entry 137 is the receiver station location of the defined shift 1 which is the third entry The only requirement of the rece definitions is that the numbers increase and that they be in sequential order 50 5 138 0 6 6 rece 139 0 7 rece 140 1 8 rece 141 0 3 9 rece 142 0 6 10 outf stat dat see previous description of 11 gt gt see previous description of gt gt end As with all the batch files it is advisable to type list them out after they have been constructed in a text orline editor The following is what the file stat dek should look like gt gt start inpf sort dat stat 4 rece 137 0 3 rece 138 0 6 rece 139 0 rece 140 1 rece 141 0 3 rece 142 0 6 outf
62. ecord number index and trace number index based on values in trace header word 3 and 4 respectively 4 Velocity scan data 2 Total recording channels 3 Trace Header Word of RECORD number for this data set where 8 common recording channel number 12 common depth point 19 2 common offset 86 common receiver station number 87 2 common source station number 92 common source sequence number 4 Trace Header Word of TRACE number within each record 0 as input order of seismic input data to be sorted 5 Trace direction flag for sorted traces within each record 1 ascending _ 1 descending 6 Original Field Record number 8 Recording channel number 10 Repeated shot number at the same station 12 CDP number 14 Trace number within each record 15 Trace identification code 1 seismic data 2 dead 9 velocity flag 16 Number of vertically summed traces yielding this trace 17 Number of horizontally summed traces yielding this trace 19 Offset distance from source to receiver after multiplied by word 35 21 Receiver group elevation 29 Source elevation 27 Datum elevation 35 Multiplication factor for horizontal distance 50 Source static correction ms floating pt 51 Receiver group static correction ms floating pt 52 Total static correction ms that HAS BEEN applied to this trace zero if no static has been applied 55 Recording delay time ms floating pt 58 Number samples in this trace 59 Sample in
63. ed to CDP gathers we will have 24 traces allowing us significantly more trace to trace coherency More traces will also improve the statistical determination of velocity STEP 8 SORA RU Ue ae eS OA NAP LE DADA S breed oes o ome First we must resort the CDP sample data The following sequence in a batch file will accomplish this Line Description 1 gt gt 5 see previous descriptions of gt gt START 2 INPF surf dat see previous descriptions or 52 3 5 92 8 RSRT operation resorts data according to any trace header word requested The first entry 92 which is the header word defining the SSN defines the primary sort operation that is the major groups to collect traces into e g field files are grouped according to 6 or 92 CDP gathers are grouped according to 12 common offset gathers are grouped according to 19 etc The second entry 8 which is the header defining the trace num ber within the original field file defines the secondary sorting operation that is the order of the traces within each of the major groups defined by the first entry channel number with each original field file are grouped according to 8 traces with each CDP are grouped according to 14 traces within common offsets are usually gathered according to header word 6 87 or 92 4 OUTE rsrt dat see previous descriptions of OUTF 5 gt gt see previous descriptions of gt gt END The resort batch fi
64. ely The CDP stacking of reflection data amounts to the summing all traces with the same midpoint between source and receiver after correcting for different source to receiver distances and dividing by the number of traces summed or some other logical dividing screen This process is conceptually quite simple 79 figure 19 STEP 11 CRGO eee SA ARTA EU AMPLE For our sample data set the following batch processing sequence would result in a CDP stacked section Line 1 2 Description gt gt 5 see previous description of gt gt 5 scal dat see previous description of INPF STAK 1 STAK operator initiates the CDP stacking process The first entry 1 identifies the divisor after the summation process necessary to return the amplitude levels to near pre stack values The divisor can be selected as the actual number of input traces fold or the square root of the fold The square root option will simply result in increase the relative significance from an amplitude perspective of higher fold gathers OUTE stak dat see previous description of OUTF gt gt see previous description of gt gt In order to verify your the correctness of your batch processing deck you should list it out It should look like gt gt start inpf scal d
65. er gt 6velocity 2300 lt enter gt 7velocity 2400 lt enter gt Svelocity 2500 lt enter gt 9velocity 2600 lt enter gt 10velcoity 2700 lt enter gt 11velocity 2800 lt enter gt 12velocity 2900 lt enter gt 13velocity 3000 lt enter gt 14velocity lt enter gt When inputting trial velocity values no set pattern or limits except for maximum number of trial velocities need be adhered to At this point the program will take control and begin to process the con stant velocity stacks in the precise fashion we described above The program will keep you abreast of where it is in the sequence of velocities and CDP numbers as prescribed above The program can be interrupted at any time with a ctrl c 57 command The ctrl c command will terminate the operation however any pro cessing completed at the time of termination will be accessible with view or plot Displaying your constant velocity stacks is critical and will be the first real glimpse of what a stacked section may look like once you have completed your processing flow It is important to generate a hard copy of the velocity scans This will give you a perspective of the relative qual ity of each reflector on each CDP at each of the selected velocities Picking a velocity function is the most significant pure judgement call made yet in our processing of this sample data set You will need to select the best velocity function which could involve several
66. ere which will gather together all traces with equivalent midpoints between shot and receiver The first value header word number input designates the primary sort grouping trace header word 12 CMP or CDP number The second value header word number requested designates the secondary grouping trace header word 19 distance from source to receiver The secondary grouping simply designates the ordering of the trace display within the primary grouping With unprocessed field data the primary grouping is file or source sequence number and the secon dary grouping is according to individual trace or channel numbers PTRN 2 5 241 PTRN operation defines the distance between surface stations and physical seismograph parameters This operation must always precede the SORT procedure definition The first requested input is the distance between station locations 2 5 m The second value to be identified is the total number of recording channels on your seismograph 24 And the final value represents the units of length used during the acquisition of the data 1 meters 0 feet This operation ptrn is the lead in for the next line s pn which will identify all the different source and receiver geometries used throughout the collection of this data set S 1 24 trace channel PN 1108 115 241 X KK
67. f the scaling operation 4 OUTE scal dat see previous description of OUTF 5 gt gt see previous description of gt gt As before the batch processing file just created to automatically gain indi vidual samples relative to nearby samples on the same trace within the defined window will operate on the previously filtered data by using the following commands gt DSEIS scal dek junk lst_ lt enter gt The effect of the scaling operation are probably not worth plotting the entire data set to see The dview display routine will give you a sufficient look at the data to determine if your window is correct and if the operation was com plete and correct The effect of the scaling is actually quite evident on our sample field file Figure 19 K STACK The CDP stacking or summing process is the culmination of most modern digital seismic data processing CDP stacked multi fold seismic data are generally the most useful and most common final form of seismic reflection data A properly processed stacked section can be used to interpret a significant amount of geologic information The stacking process if all previous processing opera tions and parameters have been appropriate and in a logical sequence will enhance seismic reflection information The processing flow prior to stacking is solely intended to manipulate and prepare reflection information to be added constructively while all other energy noise adds destructiv
68. g Gains Amp Scan Delay Damper Box In Out AMPLIFIER GAINS Filters High Cuts 60 Hz Notch Sample Int msec Alias of Samples Record Length Rec Start Delay Amps 1 Control Level Temp Wind Soil Conditions figure 6 i OBSERVATION FORM B 1 tape SEISMIC GEOMETRY Lr EG amp G England 1 Location Purpose Thames River Valley G B s S Flag No of Trace Line EG amp G England 1 ____ C Chris Leach EG amp G l oa ZI 5 2 8 Observer Don S Date __5 10 89 ontractor Coordinator 37 5 5 Split Spread S 2115 0 02 1 12 13 24 AF Time Remarks Energy Source 12 Gauge Buffalo Gun End On Gap 2 No Stacked Per File 108 001 75 115 12 Source Spacing 2 9 m iW P Wave Single Source 109 002 76 116 11 Freq 100Hz S Wave Multi Source Take Out Spacing 2 9 Arr
69. g one shot per surface location 1 is the appropriate sixth value The repeat shot number simply identifies which shot this file represents at this location If multiple shots are recorded individually at this location on different files the program is capable of identifying which shot this file represents and then allows the option to later sort according to particular repetitive shot number at each location SN 2 1091 This SN definition as with the previous one line 7 is required to describe the source station number and it s associated live receiver pattern This must be done for each defined SSN The only difference between this sn 39 10 definition and the one previous is SSN 2 is defined which has shot location 109 whereas the previous one was for SSN 1 which had shot location 108 The last three parameters were omitted on this definition because the default values were appropriate The line format was purposely changed on line 8 simply to demonstrate the flexibility of formatting within the program This shows the program is not field sensitive with respect to the number of spaces between input parameters However it should be kept in mind that the order of the in put values is critical If you wish to input values not identified as default values for the sixth and seventh requested parameters but your fourth and fifth are the same as the default values you must input all values from the first requested to the seventh regardless
70. ganizing trace headers and data bytes into a specific pattern and sequence recognizable by Eavesdropper The formatting utilities available for Eavesdropper require raw unformatted data to be present on hard disk The formatting utilities conversion routines are designed to operate on raw data input from hard disk and output back to hard disk Getting the raw data from the seismograph s preferred storage media floppy disk 9 track tape tape cartridge RAM etc onto the hard disk requires procedures software and or hardware that can be supplied by the seismograph manufacturer Often the transfer of raw unfor matted data to a computer s hard disk requires nothing more than the MS DOS copy command The particular formatting routine necessary for your raw unformatted data depends on the seismograph with which it was collected Until a standardized format can be established and agreed upon by all seismograph manufacturers and software developers a different conversion routine will be necessary for most new and existing seismographs At the time of this writing Eavesdropper facilitates the following data formats Program Description 90002KGS 9000 Bison to KGS BISCONV Geopro Bison to KGS EASI2KGS EASIDISK EG amp G seismograph to KGS SEGI2KGS SEGYinteger to KGS SEGF2KGS SEGYfloatingpoint to KGS SEGPFKGS SEGYfloatingpoint to KGS SV2KGS Seisview EG amp G to KGS GEOF2KGS GeoFlex EG amp G to KGS 24012KGS 2401 EG amp G to KGS DFSDEMUX
71. hannel number and may or may not be the same value 43 number 246 245 mra i A E cms m FER PT A AL PY 54 aA 8 45258 TTA ARNE AA EE NE EL AN ti Ans Fog 12501702 Y Ad ma al a Am T T n pe oe gt Bs A ee E Ar lt 3 a ry LV ig wA y As MR AR Me ka V iv 5 m figure 9 ELEVATION CORRECTION DATUM ASSIGNMENT Correcting your data for a variable near surface can be one of the most difficult problems encountered while processing seismic reflection data Most areas on a seismic reflection section possessing poor signal to noise ratios or that have resulted in an incorrect structural interpretation are pigeon holed as related to static problems These problems can be the result of anything from incorrect compensation for surface topography to uncompensated velocity irregularities
72. he frequency and decay length of the sinusoid is dependent on the defined filter This decaying sinusoid is an artifact of the Fast Fourier Transformation FFT which is part of the spectral filtering process Frequency filtering is often necessary to remove or attenuate unwanted noise Choosing a taper length is a very data dependent undertaking At least one cycle of the dominant reflection frequency or center frequency of the digital band pass filter designed for this data set is a good starting point for defining a taper length Fine tuning of a mute taper generally is not necessary but in certain instances returning to this step in the processing flow to better define a taper length may be necessary During future processing operations involving a taper reference to this paragraph will be made In Eavesdropper the taper is defined according to Figure 3 20 Original Trace Surgical Mute Taper First Arrival Mute Taper taper length ms defined beginning time defined ending time mg taper taper length length ms ms 1 0 1 1 0 1 1 0 1 _ a if E 0 relative amplitude relative amplitude relative amplitude figure 3 STEP 74 ER ANA DT OE DA Ae ee ee S oe direct The first step in the first arrival muting process is to identify refracted and wave energy on your raw field plots Figure 1 Once a definite identifica tion is made
73. high cut low pass frequency The third input 1 identifies the type of filtering you wish to do In our case we are doing a band pass but in certain instances the enhancement of reflection signal may involve the rejecting of a particular frequency window In such a case the band reject filter 0 option for the third entry would be appropriate The forth entry 0 determines whether a 60 dB down 60 Hz notch filter is to be applied The notch filter is design to drastically attenu ate the effects of electrical power lines The notch is only necessary when the effects of 60 Hz noise are noticeable on the field data The final input parameter 60 determines the length of the filter operator Due to the limited amount of time and space in this document a detailed explana tion of this parameter would best be obtained from a basic filter theory text book From a crude practical sense the longer the filter max less than 1 2 record length the better the filter works On the other side the longer the filter the more computer time that is necessary to complete the filter operation A happy medium needs to be determined in order to properly filter your data without taking a significant amount of CPU time filt dat see previous description of OUTF gt gt see previous description of gt gt As with all other batch processing files to operate on the data the file must be run through DSEIS gt DSEIS filt dek filt lst
74. identified during the surf operation type the following at the system prompt gt gt TYPE return You should see the following if the above deck is what you entered gt gt start inpf sort dat surf 100 alof 105 770 90 se 126 97 100 se 127 100 se 136 100 se 137 100 3 49 56138 100 6 56139 100 56140 99 se 141 99 6 se 142 100 6 se 143 100 STAT 1 outf surf dat gt gt The next step is to run your elevation correction job through the DSEIS program At the system prompt type the following gt DSEIS SURF dek SURF Ist return As before while the program dseis is running it will keep you abreast of where it is in the processing sequence If later you want to study each step of the job that you ran simply type out the journal file surf lst If you wish to see what effect your datum correction had on the sorted data follow the series of steps described for plotting earlier in this manual Displaying a data set after surf will probably not be an effective use of time or resources The more advisable way to check to make sure your correc tions were applied the way you had intended is to run the trace header dump routine trhd previously described An alternative way to apply your datum elevation statics is to use the STAT operation and directly input the amount of time shift necessary to compensate for variability in velocity and or elevation to each station location across th
75. ignal at 31 ms 80 percent of the signal at 32 ms 70 percent of the signal at 33 ms etc until at 39 ms 10 percent of the signal was attenuated FARM 1130232333 435 5 36 6 38 7 39 8 41 9 42 10 44 11 45 12 47 13 48 14 50 15 51 16 53 17 54 18 56 19 57 20 59 21 60 22 62 23 63 24 65 FARM defines first arrival mute times according to SSN and trace num ber The farm operation is designed to interpolate both in time and space This interpolation process makes the entry on line 5b the same as the 22 5b T 6a t 6b T 6c t 6d entry on line 5a If only line 5a or 5b farm was defined the entire data set would be first arrival muted according to the defined record number time windows The actual mute defined by line 5a or 5b would include the entire data set and delete all data between time zero and 30 ms on trace 1 from time zero to 32 ms on trace 2 from time zero to 33 ms on trace 3 etc out to trace 24 which will be muted from time zero to 65 ms If more than 24 traces are present in this data set each trace beyond trace 24 will be muted as trace 24 was This means that trace 25 will be muted from time zero to 65 ms trace 26 will be muted from time zero to 65 ms trace 27 will be muted from time zero to 65 ms etc FARM 1 1 30 24 65 This defines exactly the same first arrival mute as line 5a FARM210240 FARM 3 1 30 24 65 FARM 410240 FARM 3 1 30 24 65 25 0 t Not appropriate for this data set used only
76. in the near surface On any data set the first step in eliminating static effects requires adjustment of the data to a datum A datum is simply a reference line with some absolute correlation to sea level or in some cases average slope of the ground surface sloping datum Generally on conventional data the datum is chosen beneath the weathered and within the subweathered zone Defining the datum within the subweathered zone is necessary to insure all non uniform near surface material is above the datum The correction conceptually involves removal of all material above the datum which in turn ultimately allows accurate time to depth conversions relative to sea level on inter preted CDP stacked sections Correcting for both surface topography and a non uniform weathered layer requires at least relative station elevations and a somewhat detailed knowledge of the near surface velocity and depth structure The datum correction is generally perceived as a first guess approximation and is intended to remove major static anomalies associated mainly with relative elevation changes Discrepancies between the calculated datum correction and true datum correc tion are generally addressed using iterative numerical techniques at a later stage in the CDP processing flow Shallow high resolution reflection profiles generally target geologic features within what conventional wisdom would suggest is the weathered zone Correcting for elevation and near surface i
77. inantly reflection energy 2 Filtering Analysis of the spectral plots allows the designing of an appropriate digital filter to en hance reflection energy The most common type of digital filter and the most appropriate for our sample data set is a digital band pass filter This filter by its nature will attenuate all energy with spectral values outside a defined window The window is defined in the frequency domain and is shaped in accordance with predefined front and rear slopes Figure 17 of the filtering options available in Eavesdropper have front and rear slopes that can be thought of in a very similar fashion as the taper previously discussed for the muting operation Without these slopes when the transformation from the frequency domain to the time domain is made after filtering sinusoidal artifacts will be injected into the seismic data The frequency of these artifacts will be related to the frequency of the defined high and low cut filter values of the bandpass STEP 729a Oe SOR esse p AMPLE 73 55 5555 For our sample data set it appears from spectral analysis that the domi nant frequency of the reflection information is approximately 250 Hz The bandpass filter we design must not attenuate any signal within one half octave of that frequency The air wave information ranges from about 50 to 400 Hz Of course that frequency band is coincident with the reflection information The muting operation perf
78. is truly representative of the data STEP 710a EXAMPLE 788 For our sample data set the following batch processing file will most effec tively boost the signal to noise ratio retain some relative amplitude informa tion and improve the interpretability of some of the subtle reflecting events identifiable on the field files Line Description 1 gt gt 5 see previous description of 22START 2 INPF filt dat see previous description of INPF SCAL 50 78 SCAL identifier initiates the scaling operation within SEIS The first entry 50 determines the window length of the scaling operation This value is in units of ms There are other entries associated with the scaling operation but for the data set we are processing here the default values are adequate One of the other potential variable parameters allows you to choose either absolute value mean or root mean square with a user definable reference mean The particular type of statistical technique used to determine the amount of gain necessary for each sample results in subtle difference on most data sets A discussion relating to which type is best and for which type of data is not appropriate for this manual The reference mean is a value set for 16 bit data and variation of this value will effect the amount of gaining necessary relative to maximum possible deflection The third parameter relates to delay time in the initiation o
79. k on just those two sequential groups of CDPs could be per formed using the following sequence t First CDP of group 1 2 gt current 223 230 lt enter gt t Last CDP of CDP group 1 gt current 282 240 lt enter gt t of CDP Groups to be processed gt current 1 2 enter t CDP group increment gt current 20 30 lt enter gt t This information is not necessary for this data set and is provided as simply an example of the program s operation Effectively the program increments 30 from the first CDP of the first group to determine where to begin the process for the second group Therefore in this case the second group would start with CDP 260 and 11 CDPs would be processed 240 230 11 up to and including CDP 270 Likewise if sufficient data existed the third group would have started with CDP 290 and ended with CDP 300 If 15 had been selected as the increment value CDP 245 would be the first CDP of the second group and CDP 255 would be the last As well the third group of the 15 increment example would have started with CDP 260 and ended with CDP 270 Record length ms gt current 250 lt enter gt This allows the option to process a small chunk of data starting at time zero of course This will be especially useful when more data are acquired than is really necessary Allowable sample stretch ratio greater value gives deeper mute 56 min 0 2 max 0 99 current 0 50 0 60
80. le you just created should look similar to the following when displayed on the CRT of your computer gt gt start inpf surf dat rert 92 8 outf rsrt dat gt gt end As with the previous batch jobs the following sequence will initiate the resorting operation gt dseis rsrt dek rsrt lst enter At this point the file rsrt dat is in field file format One option at this point is to peel off only file 5 and input it into the VELP program In the VELP program the option exists to select the desired field file on which to work We will remove only file 5 as an exercise in using the inpf definer The following operation will result in an MS DOS file being created containing field file 5 only Line Description 1 gt gt 5 see previous description of gt gt 5 2 rsrt dat 5 5 100 The three entries after the input data file 5 5 100 identify the beginning 53 and ending file number to operate and record length in ms The pri mary sort order of the data dictates file designation This means that if the data are in field file format you can select particular field files to operate on i e 5 If the data are in CDP format you can select a particular sequence of CDPs to operate on In this case the input data rsrt dat are in field file format allowing us to directly select field file 5 It is not possible however to select a particular CDP or group of CDPs from this particular input data file
81. litude event allows events later in time to dictate the amount of uniform whole trace amplification applied This program is FIELD SENSITIVE which means don t put spaces after a prompt and or before the requested information unless directed by the program Both dplot and dview routines will return system control after the display process is complete STEP 72 EE AMPLE DATA A hard copy plot of raw England field data should contain 20 files identified by the source sequence number in the upper right hand corner of each file The data were collected on a 24 channel seismograph Therefore there will be 24 individual traces within each field file The traces within each field file are identified by original channel numbers starting with channel 1 on the left hand side of the field file and channel 24 on the far right The field file displayed here has all the major types of seismic energy arrivals you will encounter on most seismic data sets Figure 1 Identified on the plot of field file 5 is each trace number time in milliseconds refraction energy air coupled waves ground roll and of course reflection events Of inter est for later processing steps two dead traces are identified at trace numbers 8 and 23 The field file displayed has been scaled to enhance the seismic energy mak ing the identification of various types of arrivals easier E EDIT The next step in a standa
82. location 116 which is trace channel 13 where the remaining 12 channels are defined starting at 116 to 127 inclusive each sequential increase of 1 in trace channel number corresponds to a sequential increase of 1 in the surface station location t Not appropriate for this data set used only as an example 6 SHOT This alpha character description shot simply identifies the succeeding series of sn or snsn definitions as related to the shot location and source sequence numbers SSN This operation must follow the ptrn operation SN 11081001 The SN definition input is required for each field file source sequenced file This operation matches a source sequenced file with its appropriate shotpoint s surface station location and source receiver geometry as described by the ptrn operation In this case the Source Sequence Number SSN is 1 the shot station location is 108 the source receiver geometry is defined in pn 1 The fourth entry is related to inline offset 0 this is simply the distance from the defined surface station location to the actual location of the shot recorded along the line of the survey parallel offset The fifth entry 0 is also related to the actual location of the shot recorded at this surface location 109 and it defines the distance off the line between the shot and actual surface station location perpendicular offset The sixth value 1 identifies the repeat shot number Since we are only recordin
83. lt enter gt Sample stretch ratio defines the amount you are willing to allow your wavelet to stretch as a result of the dynamic NMO correction before it is muted This parameter requires a great deal of care and careful thought before and after application Artifacts will be generated on stacked data if this parameter is not properly selected The result of an improper mute can range from apparent high frequency coherent events to anomalous low frequency shallow wavelets Experience and care will keep you out of trouble with this parameter With the extremely site dependent nature of velocity functions a simple rule of thumb is not possible This is a case where having the appropriate experience or academic background to understand the mathematics of the operation will greatly enhance your ability to properly select mutes and to know from inspection of stacked data if the mute was not properly assigned Stack 0 DO NOT STACK 1 STACK current 20 1 enter This allows you to inspect your data after moveout either in a stacked or unstacked form The unstacked form is very helpful in some cases because it will let you see exactly how much each trace was moved out and the associated contribution of each trace of the final stacked section Enter trial velocities lt cr gt to end max 20 1velocity 1800 lt enter gt 2velocity 1900 lt enter gt 3velocity 2000 lt enter gt 4velocity 2100 lt enter gt 5velocity 2200 lt ent
84. m to print 1 or not to print 0 the flagged bad traces in the list file This fourth option gives you the opportunity to examine the traces the program suspects as being bad The fifth and final option lets you delete the flagged bad traces 1 or save all the traces as inputted 0 This allows examination of the suggested bad traces and readjustment of parameter 1 noise window and 2 signal to noise ratio or both If the 0 option is chosen for the fifth parameter no output file need be named 4 OUTF Aued dat see EDKL for details on OUTF 5 end see EDKL for details on gt gt end The batch processing file AUED DEK you just saved looks like the following gt gt start inpf dengland raw 20 0 28111 outf Aued dat gt gt end In order to execute the automatic editing operation on the raw input data the following sequence again will be necessary gt DSEIS AUED DEK AUED LST The AUED LST file monitors and records the sequence of processing events The following information is included to allow you to check the contents of your list file with what should be there List of example of signal to noise ratio of SSN 4 4 3 24 2 43 3 39 3 10 5 60 3 26 0 62 1 45 0 23 1 42 0 81 1 47 0 32 0 94 0 44 0 92 2 45 1 26 1 42 0 64 2 11 1 14 0 71 0 21 List of bad traces for 55 4 from automated editing 4 924 The edited data be in the file named AUED dat In order to see the effect of the editing on the ra
85. mal moveout correction to our sample data set is created and dissected below Line Description 1 gt gt 5 see previous description of gt gt 5 2 surf dat see previous description of INPF 3 NMOT 0 6 operator adjusts each sample of each trace dynamic correction for a specific velocity function described by the VELF definer The only entry required for the NMOT operator is the value of the sample stretch ratio 0 6 This value specifies the amount of mute to apply to a trace to 61 suppress the stretching of the reflection wavelets resulting from the correction of non vertically incident ray paths By correcting to vertical incidence all traces should be geometrically equivalent and whole trace addition within a particular CDP should result in enhancement of reflec tion energy As discussed during the DVSCN operation a proper nmo stretch mute is critical to generating a realistic stacked section An incor rectly designed and applied stretch mute can generate depending on whether the mute is too extreme or too subtle anything from high frequency spikes to very distorted low frequency wavelets The stretch mute is most evident on shallow low velocity reflections This is because the most extreme stretch results from lower nmo velocities as is intui tively obvious from a hyperbolic curvature perspective The more severe hyperbolic nmo curvature lower average velocity is generally associate
86. n exe dvelp exe fmain hlp view cfg plot cfg nmot dek scal dek edfm dek edkl dek rsrt dek filt dek process dek edmt dek sort dek stak dek surf dek The extension ext be used for quick and easy identification of file type For example files referred to in this document with an exe extension are executable codes cfg extensions are graphics configuration files dat are seismic data files dek are batch processing command files and 1st are journal or list files D DISPLAY Eavesdropper can display seismic data either variable area wiggle trace or just wiggle trace on your CGA EGA or VGA CRT with a hardcopy print option Two routines are contained within Eavesdropper to display data The dview routine is mainly designed as a quick way to see data on the CRT without a hardcopy option This quick display routine is most helpful during preliminary plot parameter design The main plotting routine is called dplot Dplot prepares and displays the data on the CRT with a hardcopy option During your first attempts at plotting we recommend using dview to select the appropriate plotting parameters followed by dplot for final CRT display and hardcopy After your data are in Eavesdropper format examination of a variable area wiggle trace display of all the unprocessed data will allow you to verify proper formatting as well as to get a general feel for the quality and quantity of data The processes involved with getting a di
87. nd field file numbers as the y axis Each individual field record file number y axis has an associated set of 24 recording channels and a shotpoint The shotpoint location for a particular field record is identified by an x located beneath the appropriate station location x axis Along with this x defining the shot location is the assigned source sequence number SSN Each live receiver location is represented along the x axis by the appro priate seismograph channel number Notice the step by step progression of the shot and receivers across the line The lower portion of the chart identifies the locations and SSN original field channel pairs as well as fold redundancy percent coverage for each CDP 31 OBSERVATION FORM Tape tape SEISMIC GEOMETRY Location Purpose ai 5 Contractor Coordinator 5 Flag Noval Trace E 8G Split Spread 556 LT 82 1 121324 GE rime Remarks Energy Source End On No Stacked Per File Source Spacing P Wave Single Source Geophone Freq 0 S Wave Multi Source Take Out Spacing 0 Array Single Phones Source Receiver Geometry 1 3 5 7 9 11 13 15 17 19 21 23 Fiel SA 2141 6 110 12 14 16 18 20 22 24 Test Oscil Oper atin
88. ng bogus reflecting events are increased by studying constant velocity gathers Moved out gathers at this point in the processing flow are for the most part trace by trace the same except for the extreme cases with a slight static shift as the original recorded data The only difference is the whole trace dynamic test moveout compensation which is the variable you are testing Analysis of data in this fashion allows you to go back to your original field files and actually follow identifiable reflection information through the velocity analysis The velocity function that is chosen in this fashion should possess a high level of reli ability once the data are actually stacked and will reduce some of the qualitative aspects of the selection routine STEP 8 AMPLE 1 72 The optimum stacking velocity should remove the hyperbolic curvature of a reflecting event The result of the correct velocity on raw field files is trace by trace consistency in the time and wavelet characteristics of the reflecting event Figure 11 58 velocity value 2600 230 240 250 260 270 280 1900 230 240 250 200 2 0 280 x or iD orco Er GS 21 12 24 VEER BRE EAM x a1 pr gt ry a 2 7 A 8 2 d
89. ocessed section The processing flow we use is structured to maximize shallow high frequency CDP seismic reflection data The general outline of our processing flow is contained in table 1 Table 1 contains all the operations we routinely use to go from raw field data to finished process sections This manual will discuss in some detail all the operations through brute stacking of seismic reflection data The intention of this novice user s manual is to get a person started and somewhat familiar SEISMIC DATA PROCESSING FLOW CHART PDA Formatting Screen Displa 3 epay Raw Formatted Flek Data Screen Display with Option to Generate Hard 0 Bad Trace A A E Eg t amp ES F NES 99 00 00 O6 o 25 5 5 H CRBS 0908 Sorted Data CDP Common Offset Common Trace splay Common Receiver Common Shot H o6 9555 Frequency Main Spectral MA NMO Velocity DaturvE ovation Corrected Function 1 F Main 2nd F Main Spiking Zero Crossing Decon Deco o 5 CGD CD Ca oa Table 1 BLOC SOLOS 2nd Zero Crossing Stack Brute Stack Brute Stack 2 F Main Spking Decon Velocity Function 2 Table 1 of Static Corrections DADEA Intermediate Stack Section HOE 4 5 5 Velocity Static Corrections Function 3 Frequency i 2 ity
90. ollowing commands at the system prompt gt 5 return gt gt start inpf edmt dat sort 12 19 ptrn 2 5 24 1 pn 1 108 115 241 shot sn11081001 sn2 1091 snsn 3191 lt 110 12611 sn201271001124 tabl 11 outf sort dat gt gt end Now in order to run the sorting batch job you just typed out enter the following series of commands at the system prompt gt DSEIS SORT dek TABLE Ist lt return gt As with the other batch operations already run the program will continu ously update the screen with pertinent information regarding the present status of the batch job The TABLE Ist file is again a journal file that will contain not only the blow by blow account of the processing sequence it will also contain the 41 geometries and sorting sequences applied in accordance with the input informa tion This is intended to allow you the option to compare your stacking chart developed from the field notes with the actual information you coded into the various operations in the batch processing file The journal file will be contained in the subdirectory in which you are running the seis program in and can be viewed either in hard copy or on the CRT The stacking chart generated by the program assuming you used the tabl option with the information input during the sort operation is displayed below SORTED RECORD TABLE SHOT 1 RCRD FOLD SN TR SN TR SN TR SN TR SN TR SN TR SN TR SN TR SN
91. on of AGC CDP stacking of data 11 R Figure QD 5 n WN Table 1 LIST OF FIGURES AND TABLES Raw field file SSN 5 England data Bad trace edit file 5 England data Trace and taper First arrival mute file 5 Surgical mute file 5 Page of field notebook Field notes for sample data set from England Stacking chart CDP sort of 245 and 146 Constant velocity stacks 2 Velocity function on file 5 Spectra SSN 5 trace 18 Spectra SSN 5 predominantly air wave Spectra SSN 5 predominantly ground roll Spectra SSN 5 predominantly refraction energy Spectra SSN 5 predominantly reflection energy Shape of the filter for 125 to 400 bandpass Filter of file 5 Scale of file 5 Brute stack of file 5 Processing flow lii o 69 17 21 25 29 32 33 34 35 44 59 60 68 70 71 72 73 74 77 80 83 2 3 INSTALLATION INSTRUCTIONS Programs and data contained on the included disk will operate in a fashion nearly identical to the full Eavesdropper package The demonstration software and manual have been compiled to instruct the novice as well as allow a seasoned processor an opportunity to see and feel the flow of this seismic processing package Only a small sampling of the operations available with the Eavesdropper package are on this demo Four types of data are contained on the demonstr
92. orking directory and at the system prompt execute whichever display routine you desire gt dview return Or gt dplot return The first question the plot or view program will ask is enter filename to plot dengland raw After you have entered the file name dengland raw the program will respond with one or more of the following statements 1 No processing history file available which means basically that no processing has been completed at this point so the file designed to handle the processing history has not been created yet 2 Processing history file name found which means basically what it says and the result will be the printing of the current processing history at the conclusion of the plotting of the data 3 Warning History file C FNAME HO1 not found meaning that there should be history file called FNAME present from previous processing but dplot was unable to find it 4 Plot cfg not found which means it was not able to find the plot cfg file and the predesignated default parameters will be used In the case of raw data such as we are displaying at this time the program should respond with statement 1 only This is assuming plot cfg or view cfg depending on the requested routine is present and has been updated according to the previous instructions If at any time you wish to return to the system prompt hit Crt C Enter starting record number default first record found for
93. ormed early on in the processing flow removed the majority of the air coupled waves The ground roll frequencies on the other hand fall mainly within a band from about 25 to 250 Hz With the amplitude of the high frequency ground roll small relative to the reflection information the low cut side of a band pass filter should attenuate the majority of the ground roll The refraction energy is large amplitude and possesses about the same spectral characteristics as the reflection signal Once again muting was essential this time in removing the effects of the refracted energy From the spectra of the various types of seismic energy arrivals the optimum digital bandpass filter for our sample data set will be something in the range of 125 to 400 Hz band pass As with most seismic data processing selecting a band pass filter whether from spectra plots or directly off raw field data becomes easier with experience and a broader knowledge of the 69 figure 13 1 00 AIR COUPLED WAVE FILE NUMBER 5 TRACE NUMBER 14 0 75 4 SAMPLE RATE 2000 Hz SAMPLE SIZE 500 START TIME _ 140 Ms 050 END TIME 180 Ms LOW CUT 0 Hz HIGH CUT 1000 Hz 0 25 0 00 0 400 600 800 1000 1 00 SURFACE WAVE FILENUMBER 5 TRACE NUMBER 1 0 75 4 SAMPLE RATE 2000 Hz SAMPLE SIZE 500 START TIME 80 Ms E 250 Ms LOWCUT 0 H2 HIGHCUT 1000 Hz 0 25
94. physical signal analysis Prentice Hall Inc Englewood Cliffs NJ 466 p Waters K H 1987 Reflection seismology A tool for energy resource explora tion 3rd ed John Wiley and Sons New York 538 p Yilmaz O 1987 Seismic data processing 5 M Doherty ed in series Investiga tions in Geophysics no 2 Soc Explor Geophys 526 p 85
95. qe 0 200 400 600 800 1000 figure 14 figure 15 1 00 0 75 0 50 0 25 0 00 REFRACTION FILE NUMBER 5 TRACE NUMBER 11 SAMPLE RATE 2000 Hz SAMPLE SIZE 500 START TIME 30 Ms END TIME 70 Ms LOW CUT 0 Hz HIGH CUT 1000 Hz 600 800 1000 figure 16 1 00 0 75 0 50 0 25 0 REFLECTION FILE NUMBER 5 TRACE NUMBER 20 SAMPLE RATE 2000 Hz SAMPLE SIZE 500 70 Ms END TIME 180 Ms LOW CUT 0 Hz HIGH CUT 1000 Hz 600 800 1000 figure 17 1 00 0 75 0 50 0 25 BANDPASS FILTER LOW CUT 125 Hz HIGH CUT 400 Hz FILTER LENGTH 61 0 00 0 200 400 600 800 1000 basic physical principles of seismic data processing There is no substitution for a proper math and physics background STEP 95 EXAMPLE DACIA To build the batch processing file to operate the seismic data from our sample data the following sequence of parameters needs to be defined Line 1 2 Description gt gt 5 see previous description for gt gt START INPF nmot dat see previous description for FILT 125 400 1 0 60 FILT operation initiates the frequency filtering operation The first input parameter 125 designates the low cut high pass frequency The second parameter 400 designates a
96. qual weighting of all shot locations across the line one of the duplicate shots must be removed In this case the second shot was deleted using the 7th and 8th values in the SN definition The program will interpolate between the seventh and eighth input values This means if the seventh value is 19 and the eighth value is 22 the program will omit 19 thru 22 inclusive The previous SN definers have had the same inline and offline offsets as well as repeat shot number The information was not entered because the default values are appropriate in other words if you do not enter anything 40 for the required values the program assumes the default values are correct 11 TABL11 The tabl operation is optional and allows you to cross check your defined geometries with the actual field geometries The two input values are flags The first 1 designates that you wish to have the receiver diagram plotted out and the second 1 signals the printing of the sort table The entire table will be contained within the journalization file i e DSEIS SORT dek TABLE Ist where TABLE Ist is your journal file and it can be typed out or printed using standard MSDOS operations 12 OUTF SORT dat see EDKL for description of OUTF 13 gt gt see for description of gt gt END At this point you need to exit your text editor In order to see the batch processing file you have created to sort seismic data into a CDP format type the f
97. rd processing flow involves the removal of bad traces generally caused by dead geophones bad geophone plants seismograph amplifier failure cultural noise or poor near surface conditions bad parts of traces generally results from the air coupled wave or ground roll and energy arriving prior to the first identifiable reflection signal generally refraction and direct wave energy 11 1 Manual Bad Trace Edit Removal of dead or bad traces is the first editing step This can be accomplished in two different ways The first way the more standard technique involves the manual entering of each trace to be removed into a text editor built edit deck using the procedure The second way uses an automatic whole trace editing routine AUED procedure designed to identify and auto matically remove if specified any trace that doesn t meet the minimum operator specified signal to noise ratio S N In order to develop a good working knowledge of what and how the editing process works it is recommended that command of the manual editing technique be established prior to extensive use of the automatic editing option STEP 3 Www coke e do AMPLE DATAS 995 A plot of the raw field data is critical at this step Careful examination of each trace of every field file will allow you to determine how much and what type of editing will be necessary The object of this stage is to remove all traces and parts of traces wi
98. rregularities on shallow profiles therefore becomes quite complicated in comparison to the pre viously described conventional procedure for datum corrections Detailed knowledge of shallow velocity and depth information is critical for accurate datum corrections on shallow reflection surveys Subtle changes in the velocity of very near surface material can drastically affect the spectral as well as spatial properties of higher frequency data For example if you are looking for structure on the bedrock surface at a depth of 12 m with an velocity of 400 m s and dominant reflection frequencies around 200 Hz an uncompensated 0 3 m variation in elevation or a 10 m s velocity anomaly can result in a 100 degree phase shift in the recorded reflection wavelet A grad ual horizontally increasing near surface velocity can be misinterpreted as a slope on the bedrock surface The selection of a datum for shallow reflection surveys can greatly influence the outcome and validity of the final interpretation of your data Many times reflection information of interest will be removed if a flat datum is chosen within the subweathered layer followed by conventional calculated removal of all overlying material For most shallow reflection profiles a datum should be chosen equal to the highest topographic feature on the survey line Then the time adjustment to compensate for the missing material should be added to each trace in accordance to the velocity depth fun
99. splay using either dplot or dview will be discussed simultaneously This should cause no confusion since dview and dplot generate exactly the same output The dview routine has no hardcopy option and therefore the screen display in dview is several times faster than dplot The few minor differences in requested parameters that occur will be discussed at the appropriate time Line 1 2 3 Line 1 2 3 set the basic plotting parameters you will need to edit the plot cfg plot configuration or view cfg view configuration file NOTE both cfg files are in the EAV subdirectory and can only be altered from that subdirectory Enter your text editor and open either the plot cfg or the view cfg file The view cfg TEXT FILE will have 7 lines with each line requiring a particular parameter as indicated below Default Value 0 350 50 65 0 0 100 2 Description This is a dummy line not used by the dview routine Designates the number of pixels on the EGA screen Approximate vertical resolution for the EGA monitor 13 DOTS INCH Approximate horizontal resolution for the EGA monitor 13 DOTS INCH Whole trace gain applied in dB This value can be either negative or positive Designates either variable area wiggle trace 0 or just wiggle trace 1 display format Tells dview what percentage of the allotted trace spacing the trace may occupy For example assume that the trace spacing is 10 per in
100. t This is most effectively done with an amplitude versus frequency display This analysis technique relies on an FFT to compute the amount amplitude of information on a seismic trace at each particular fre quency This operation will become less necessary depending on the data set as you gain more experience looking at seismic data and identifying particular types of energy and their dominant frequencies But for now you probably should use an amplitude versus frequency plot to determine the spectral characteristics of the air wave ground roll refractions and reflections The spectral analysis routines are within the program called DFMAIN This subset of the Eavesdropper package contains all operations that take place in the frequency domain To obtain a display of amplitude versus frequency of field file 5 trace 18 the following procedure will need to be followed gt DFMAIN lt enter gt A menu will then appear on the screen giving you the option to select any one of several procedures In our case we will select option number 2 which is the spectral analysis routine Once option 2 is selected you will be asked to give the input data file Here the raw data will be analyzed so our input file name will be whatever you named your raw input data For the sample data it was called England raw Once the input data file name is input enter you will skip down to the output file name which for the purpose of spectral analysis is not impor
101. t only a few particular files from a large data set if no future processing i e operations that require source and receiver geometries is planned AT THIS POINT IF YOU HAVE MANUALLY EDITED ALL YOUR BAD TRACES YOU SHOULD PROCEED TO THE FIRST ARRIVAL MUTING PORTION OF THIS MANUAL WHICH FOLLOWS AUTOMATIC EDITING AUED IF TIME PERMITS USE YOUR JUST EDITED DATA SET TO COMPARE AUTOMATIC EDITING TO MANUAL BAD TRACE EDITING IT MAY SERVE TO HELP YOUR CONFIDENCE AS WELL AS SAVE YOU TIME DURING PRELIMINARY BAD TRACE EDITING ON YOUR NEXT DATA SET 2 Automatic Bad Trace Edit Once you have used and feel relatively comfortable with the manual editing routine EDKL using the automatic editing routine AUED will save time in removing obviously dead or very poor quality traces The AUED routine is mainly designed to remove traces that are totally dead or possess a significant amount of background noise i e wind powerline automobile human traffic on line etc The important parameters in this operation are the noise window length time and the acceptable signal to noise ratio S N value At this point definitions will be helpful NOISE WINDOW The noise window identifies a pre first break segment before the arrival of any seismic signal on each trace where the level of background ambient noise is representative of the remainder of the trace The window needs to be selected so as not to include any source generated seismic signal i e
102. ta set will be compared to the hand generated chart Figure 8 after our sort deck is complete A stacking chart may not always be necessary but until a great deal of experience is gained processing seismic data it is a wise aid in assuring the data you are preparing to gather and stack is properly identified and ordered Once you have a complete set of field notes and know how you would like to order your data for future digital enhancement and display you are ready to create a batch process file to sort your data As with other operations start building your batch file by typing the following at the system prompt STEP 6a LEAS 3 7 The following deck is built for sorting of an example data set which is composed of 20 shots recorded on a 24 channel seismograph across 44 receiver stations each separated by 2 5 m The source to closest receiver distance is 17 25 m and the source stations are coincident with the receiver stations The data were collected using a CDP type roll along technique The relative source to receiver orientation was end on Line Description 1 gt gt 5 see for description of gt gt START 2 INPF EDMT dat see EDKL for description of 36 SORT 12 19 The SORT procedure calls the subroutine responsible for collecting traces according to indicated header words We have described a sorting opera tion h
103. tant and you can call it whatever you want Once you enter gt after the output file name has been given the program will indicate it is copying the file When that process is finished a series of parameters will be displayed The F3 function key will put you into change mode allowing any selectable parameter after the total number of samples sample size to be modified After the appropriate changes have been made the F4 function key will initiate the analysis The message please wait will be displayed while the program is working 65 After the spectral information has been calculated it will automatically be displayed on your CRT if you have a math coprocessor on your computer If you wish to plot the amplitude vs frequency information displayed on your screen type the letter p After you are finished studying the spectral plot press enter This will return you to the previous level where you may again select a different SSN trace pair analysis time or display parameters After you have finished your spectral analysis press the esckey twice and you will return to the standard system prompt STEP 9 gc RUN m M PL E DA PAST Eh The spectral analysis for our sample data set should proceed as follows gt DFMAIN lt enter gt MAIN MENU F1 FILTERING F2 SPECTRUM ANALYSIS F3 DECON 4 FILTER CHARACTERISTICS F5 CHANGE FILES F6 COMPARE DATA FILES 7 MANUAL STATIC EDIT CDP FILE ONLY Esc
104. terval in micro seconds for this trace 70 Analog low cut frequency Hz 3 dB pt 71 Analog high cut frequency Hz 3 dB pt 75 Applied digital low cut frequency Hz 76 Applied digital high cut frequency Hz 82 Minimum receiver station number 83 Maximum receiver station number 84 Minimum source sequence number 85 Maximum source sequence number 86 Receiver station number for this trace 87 Source station number for this trace 88 Last trace flag 0 not last trace 1 last trace 89 Surface consistent residual receiver static in number of SAMPLES that HAS BEEN applied to this trace 90 Surface consistent residual source static in number of SAMPLES that HAS BEEN applied to this trace 92 Source sequence number 93 Processing history file flag 0 No history non zero number of characters in file name to follow 94 120 Reserved for processing history file name Packed ASCII Two ASCII characters per word Convention for static corrections POSITIVE value implies static shift DOWN away from zero time NEGATIVE value implies static shift UP toward zero time Elevation can be either absolute i e positively above sea level or relative with reference to fixed altitude In both cases the orientation is such that higher elevation is positive Therefore increasing depth is indicated by the smaller value for elevation Note ms milliseconds C DATA FORMAT Formatting of seismic reflection data involves or
105. th an unacceptable signal to noise ratio S N Determination of useless traces is subjective and your ability to make that determination will drastically improve with experience Traces 8 and 23 of the displayed source sequence file 5 Figure 1 are bad traces It is important to remember all things being equal that 2 or 3 fold of high quality reflection data are better than 48 fold of garbage data The confidence necessary to effectively edit will come with time and exposure to a variety of data sets STEP 3a SERRA A AROSE DADA A SN SRE The manual editing procedure is a batch processing operation and therefore requires a batch processing file constructed around the EDKL identification In order to build a batch processing file you must use your TEXT EDITOR NOTE Any text editor that does not leave embedded commands will work i e EDLIN SIDEKICK NOTEPAD XTREE etc This is the first batch processing deck described in this manual therefore each part will be discussed in some detail and referred to during upcoming processes Line Description 1 gt gt 5 start simply identifies the beginning of this processing deck 2 INPF dengland raw 12 trace numbers LO N reflections figure 1 identifies the input files The alpha character name of the input file including any extension must follow inpf leaving at least one space
106. tion at 57 ms with 100 percent attenuation between 57 and 70 ms the attenuation would then decrease linearly from 100 percent at 70 ms to 0 percent at 80 ms 26 5b T 6a t 6b T 6c t 6d MUTE 5 1 57 70 24 222 235 preferred option and technically identical to line 5b MUTE 5 1 57 70 2 64 77 3 7184 4 78 91 5 86 99 6 93 106 7 100 113 8 107 120 9 114 127 10 122 135 11 129 142 12 136 149 13 143 156 14 150 158 171 16 165 178 17 172 185 18 179 192 19 186 199 20 193 207 21 200 214 22 207 221 23 214 228 24 222 235 The mute procedure identifiers on lines 5a and 5b define a surgical mute for the entire data set The program will linearly interpolate between all defined windows throughout the entire data set The interpolation pro cess is automatic and can only be terminated by entering zeros beginning and ending within the time ranges For the mute defined on lines 5a and 5b the entire data set will be muted with all trace 1 s zeroed all digital information removed and replaced by zeros between 57 and 70 ms all trace 2 s zeroed between 64 and 77 ms all trace 3 s zeroed between 71 and 84 ms etc out to all trace 24 s zeroed between 222 and 235 ms If more traces are present on the records they will be muted according to the trace 24 defined mute i e all trace 25 s will be muted between 222 and 235 ms all trace 26 s will be muted between 222 and 235 ms etc MUTE41002400 MUTE 5 1 57 70 24 222 235 MUTE610
107. tion or commands supplied to the program and 3 key points highlights to remember The information supplied by the program will be in italics and includes error messages file contents displayed on screen using the MS DOS type command messages concerning information being processed notes to the user concerning default parameters etc All italicized text in this manual indicates informa tion generated by the program Information you must supply to the program is always underlined and in bold type and includes execution commands parameters to input spaces necessary etc Key information to remember is always in bold type After processing the sample data set completely through using this manual future data sets could be processed by referring only to the bold type information You should become quite comfortable with the material in this manual after process ing the sample reflection data set The Eavesdropper seismic data processing package is divided into three main categories 1 plotting and formatting 2 filtering and deconvolution FMAIN and 3 the remainder of seismic data processing SEIS The plotting and formatting operations are interactive requiring you to execute the program and then enter the requested information The filtering and deconvolution routines are contained with the sub program called FMAIN and are pseudo interactive Processes in FMAIN will ask a series of questions and then operate on the data set The remainder
108. ular identification parameter is most significant for this data set and future processing routines The data can be gathered together ordered and reordered a variety of times and ways The two most commonly used parameters to sort are CDP common depth point sometimes referred to as CMP common mid point and common source to receiver offset common offset for short Sorting according to common source to receiver offset is exactly what it sounds like AII traces are gathered according to like distances from source to receiver For example each of the 24 traces recorded within each field file to our sample data set is offset from the source by a unique distance Therefore gathering according to a common offset distance will result in 24 different primary groups each with unique source to receiver offset each containing 20 traces In good data areas once the appropriate corrections are made for offset and elevations common offset data if collected within the optimum window can be viewed as a geologic cross section without future digital enhancement However seldom will common offset data yield as much or as detailed information as a properly processed CDP stacked section Eavesdropper is specially designed to enhance reflection information once your data are in a sorted CDP format Good complete field notes are critical to correct and accurate defining of source and receiver geometries surface features and events significant to future anal
109. ver valley River valleys often as with this case do not have significant changes in elevation The majority of the sample data set was collected across a flat lying pasture with little or no surface evidence to suggest areas of potential static anomalies At the extreme east end of the line however the line crossed a wash out extend ing approximately 6 m with a maximum relative depth of 1 m If not for the washout spot on the line no datum correction at this time would be necessary or possible However with the elevation change present as a result of the wash out the SURF operation is necessary on this line STEP 77 OER Ee ae EXAMPLE Determination of the velocity within the very near surface using the information provided with and contained within the sample data set is not straight forward Therefore we will provide you with the velocity information as well as the relative elevation data necessary to properly calculate and apply the datum correction An uphole velocity survey in conjunction with your seismic reflection survey is advisable whenever possible to accurately determine the near surface velocity model AVERAGE VELOCITY FROM THE SURFACE TO APPROXIMATELY 3 METERS OF DEPTH IS 770 5 CRITICAL ELEVATION INFORMATION INCLUDES THE FOLLOWING 46 STATION LOCATION RELATIVE ELEVATION M 136 0 137 0 3 138 0 6 139 0 140 1 141 0 3 142 0 6 143 0 Once as much information as possible is
110. w data you should first use the dview routine Simply type 2 DVIEW lt return gt Answer the series of self explanatory questions as described in the display section of this manual and check the format of the screen display If the display is not satisfactory make the 19 appropriate changes to either view cfg or to the responses provided to the view routine questions Once an acceptable format has been obtained make the appropriate changes to plot cfg and then type gt creturn Answer the dplot questions with values similar to those used for the previous dview routine 3 First Arrival Mute The next step in the processing flow involves the muting of refracted and direct wave energy EDFM This is necessary on most data sets to ensure that refracted and or direct wave energy does not appear coherent on CDP stacked sections The high amplitude as well as the coherent nature of moved out and stacked refraction energy is inviting and in some situations it can easily be misinterpreted as shallow reflection energy Complete identification of refracted energy is sometimes difficult on CDP stacked data Refraction energy has theoretically linear moveout on field files The NMO velocity correction applied to compensate for non vertically incident reflection energy is hyperbolic When refraction wavelets generally non minimum phase and rarely broad band are NMO corrected and stacked they can misalign in such a way as to
111. you to better see relative spectral characteristics Data file to be displayed 1 lt enter gt Inputl0 Outputl 1 Both 2 This option does not apply to this particular application It is designed for comparison of filtered spectra to unfiltered spectra Comparison of before and after filtering permits you to observe the spectral effects of your frequency filter The next step is to press the F4 function key This initiates the processing of the selected trace and time window The screen will display the following message Please wait Once the operation is complete the spectra of trace 18 of SSN 5 will be displayed and will look very similar to Figure 12 67 figure 12 1 004 0 75 0 50 0 25 0 00 WHOLE TRACE SPECTRUM FILE NUMBER 5 TRACE NUMBER 18 SAMPLE RATE 2000 Hz SAMPLE SIZE 500 START TIME 30 Ms END TIME 250 Ms LOW CUT 0 Hz HIGH CUT 1000 Hz 600 800 1000 If a plot is desired press the p key After the plotting is complete enter and return to the second level and begin your entry sequence in the same fashion as previously described The following set of plots are examples of a good series of analysis runs necessary for helping decide on the appropriate digital filter Figure 13 is pre dominantly air wave energy Figure 14 is predominantly ground roll energy Figure 15 is predominantly refraction energy Figure 16 is predom
112. ysis The information that must be contained in the field notes for each recorded shot at each shotpoint includes 1 2 3 4 shotpoint station number live receiver station numbers relative to seismograph channel numbers roll switch number individual digital file name number The remainder of the items listed need to be included but only once unless they change during acquisition of the data 5 sample interval number of samples 6 analog to digital filter settings 7 anti alias filter 8 type number and relative orientation of sources and receivers 9 profile location and purpose 10 any unusual offsets inline or offline 11 space for comments 12 time 13 weather conditions 14 system seismograph error messages 15 reminder to do system QC checks 30 An example of a field notebook that we have used quite successfully for several years is displayed in Figure 6 significant information about the source and receiver geometries as well as acquisition parameters for our example data set are logged in the field notes Figure 7 The 20 field files used for this example data set 615 634 were extracted from a larger data set containing 39 files 601 639 Building a batch processing file to define source and receiver geometries for our example data set requires geometrically relating the 20 shotpoints and the 43 receiver locations used during the acquisition of the section of the line used as our s
113. yze groups of CDP from various spots across a seismic line until trouble spots or areas with need for a more detailed analysis can be located 1 Interactive Velocity Picking The most fundamental way to determine an approximate NMO velocity for a reflector is by using the exact NMO equation and a seismic field record This process involves defining arrival time and source to receiver offset distance pairs and then inputting them into the appropriate equation for the appropriate equation see Telford et al 1976 In order to streamline this process Eavesdropper has an interactive operation that allows you to graphically determine the NMO velocity for a particular reflection event VELP To start the process of determining the appropriate velocity function for our sample data set we first will need to input a file into the VELP program and determine an approximate velocity This will allow us to optimize computer and analysis time during the constant velocity stack portion where we will define our first pass velocity function The VELP program requires the input data to already have the field geometries in the trace headers To be consistent with the previous portions of this manual we will use field file 5 to determine the velocity of our primary reflectors To get field file 5 into the correct format we will need to resort the data from CDP to field file format In this way the required header information is present With field file data as oppos

Eavesdropper Tutorial - the Kansas Geological Survey

Contents

Download Pdf Manuals

Related Search

Related Contents