Conversion from shot to cmp

[nestorkid]

Traces have been read and output as rsf file and header. However, the various programs for conversion from shot to cmp gathers have not yielded the expected cmp gathers. The data set is a 2D data set and regularly spaced. The following modules were tried for sorting the data to cmp gathers, shot2cmp, sfintbin, sftahsort. A look at various reproducible papers and an effort to follow the steps in such did not produce the desired result. Also, editing the header intervals with the 'put' module did not help. Is there anything that might be the issue or is a step missing? Also, observed that most data sets from most reproducible were converted to cmp gathers prior to being run with Madagascar.

• OS: WSL Ubuntu 20

[Boorhin]

Because I had issues with the same problem, I coded it myself in python. If you want I can send you the script. It probably is not the most optimal code but it works and it is accurate. It uses univariate spline to produce your CMP projection curve. Then it is just some sorting. Cheers Julien

On Mon, 15 Jun 2020, 12:49 nestorkid, notifications@github.com wrote:

Traces have been read and output as rsf file and header. However, the various programs for conversion from shot to cmp gathers have not yielded the expected cmp gathers. The data set is a 2D data set and regularly spaced. The following modules were tried for sorting the data to cmp gathers, shot2cmp, sfintbin, sftahsort. A look at various reproducible papers and an effort to follow the steps in such did not produce the desired result. Also, editing the header intervals with the 'put' module did not help. Is there anything that might be the issue or is a step missing? Also, observed that most data sets from most reproducible were converted to cmp gathers prior to being run with Madagascar.

• OS: WSL Ubuntu 20

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/ahay/src/issues/204, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACJCEFXV7HEJJBBHH624FKTRWYDGFANCNFSM4N6DEVEQ .

[nestorkid]

It will be really helpful if you can send me your code for shot to cmp trace sorting, I will really appreciate it. I have actually looked through many flows to appreciate that Madagascar is not optimized for shot to cmp sorting. Please, do let me know how best to run the code together with Madagascar. Regards,

Chisom

[Boorhin]

You can call the function in your Scon like this: #########################Parameters ################################## d1=0.004 d2=50 n1=12500 n2=120 CdP_Bin=100 Dir ='/path/to/your/files' CoX, CoY = 15700000, 3000000 #Coordinate modifier usually 0,0 EPSG_in = 1234 # original projection EPSG_Survey = 2345 # target projection NSF = False #nearsurface velocity correction (static stretch file)

concat your shot-receiver profiles. "All" is a list of names of

shot-receiver profiles Filt_Line ='RSF/Filt_line' Flow(Filt_Line, All, 'rcat ${SOURCES[1:-1]} axis=2') Flow(Filt_Line+'_H', Heads, 'rcat ${SOURCES[1:-1]} axis=2')

Prepare your data for CdP sorting

Prepare.Load(Dir, Filt_Line+'_H.rsf' , CoX, CoY, CdP_Bin, d2, NSF, EPSG_in, EPSG_Survey) ###############Interpolate Missing Data and regularize ################## ########################################################## Regularize.Cubize(Dir, Filt_Line+'.rsf', n1)

load parameters

Params= np.genfromtxt(Dir+os.sep+'Utils/Parameters.txt', usecols=1, dtype= np.int) MinOff, MaxOff, Low_bound, High_bound = Params ################Import CdPs ########################## Near= 1 Far = 62 N2 = Far-Near N3 = High_bound-Low_bound #len(yi) CdPw8 = [] for C in range(N3): cdp= 'CDP_RW8ted/CdP4%s'%C CdPw8.append(cdp) Flow(cdp, 'CDP_R_W8ted/NewCDP%s.dat'%C, 'sfcsv2rsf | put d1=%s label1="Time", unit1="s" d2=%s label2="Offset" unit2="m"| window f2=%s n2=%s | bandpass flo=10 | pow pow1=2'%(d1,d2, Near, N2))

You will probably need to adjust this a bit to make it work for yourself. in python 2.7 you will need m8r pyproj, scipy, numpy and matplotlib. It is probably easy to switch to 3 (it should!) but I didn't find the time. There are multiple options you can activate and actual processing things you could do but I would keep it as it is now as it is without much fuss. I should get myself to integrate that properly into Madagascar but on my own and without financial support I cannot. If someone out there is interested they are most welcome as I believe the 3D approximation of the curve is the way to go, I have done it on several projects and it is pretty accurate. The only parameter that you need to tune and that is still hardcoded is the standard deviation of the CdP cloud line 93 in Prepare.py. I have not found a satisfactory mathematical method to automatically set it up, just bound it. It is pretty important for controling the curvature of your crooked geometry. please read the Scipy.Interpolate documentation. The code displays your future curve so that's what you adjust to get a satisfactory trajectory. Please notice that it interpolates only on 1/6 traces as I used to work on large point clouds (line 95). Most lines are not shot in mountain passes so you probably can forget about it. You can also setup surface velocities to better place your points in 3D, you can extrapolate missing traces (experimental). I hope it is understandable. And the Regularyze is multithreading capable. Also shows your coverage in CdP space

The folder should be organise like this (I hope I have it all) scons RSF/ Methods/ Utils/ CDP_R_W8ted/

Good luck!

Julien

On Tue, Jun 16, 2020 at 9:51 AM nestorkid notifications@github.com wrote:

It will be really helpful if you can send me your code for shot to cmp trace sorting, I will really appreciate it. I have actually looked through many flows to appreciate that Madagascar is not optimized for shot to cmp sorting. Please, do let me know how best to run the code together with Madagascar. Regards,

Chisom

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/ahay/src/issues/204#issuecomment-644629017, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACJCEFVS4A7XJW4KNY7QSCLRW4XB5ANCNFSM4N6DEVEQ .

--

Julien Moreau, PhD, MSc, MSc The NW-Edge http://nw-edge.org Basin Analyst - Independent researcher in geology and applied geophysics Co-Founder of Plastic@Bay http://www.plasticatbay.org - Strategic development and research leader SAS rep - North West Scotland - Scotland & Hebrides Rep https://www.sas.org.uk/rep/julien-moreau/ Breadalbane, Durness, Lairg IV27 4PN United Kingdom Mobile: +44 (0) 746 20 781 62 julien.moreau@nw-edge.org julien.moreau@plasticatbay.org