Does anyone work with MSEEL Dataset?

[tiago-pitchon]

I am trying to read the dataset from MSEEL, but DASCore shows an error (unknownfiberformaterror: could not determine file format of ...).
Please, tell me if you already used this dataset and could give me a hand.

[d-chambers]

Hey @tiago-pitchon

I haven't worked with this dataset before, and apparently DASCore doesn't know about it either. Could you provide a link to where you got the data from so I can look into it?

[tiago-pitchon]

Hello, @d-chambers!
Thanks for your help.
The files are huge, be careful.

Files: https://data.wvgis.wvu.edu/pub/project/mseel/FiberOpticsData/
Main page: http://mseel.org/

[tiago-pitchon]

@d-chambers, I was able to read the Textual File Header.

C 1 CLIENT North East Natural NNE SCHLUMBERGER CREW NO Pittsburg................
C 2 WELL MIP3H LOCATION Morgantown West Virginia................................
C 3 1 November 2015 OBSERVER Dowdell/Miller.....................................
C 4 INSTRUMENT: SLB HDVS   SERIAL NO ENP26505...................................
C 5 493 TRACES/RECORD   0 AUXILIARY TRACES/RECORD...............................
C 6 SAMPLE INTERVAL 500 SAMPLES/TRACE  60000....................................
C 7 RECORDING FORMAT 1..........................................................
C 8.............................................................................
C 9 GAIN  TYPE: 0...............................................................
C10 FILTERS: ALIAS 0 HZ, NOTCH 0 HZ, BAND 0 HZ, SLOPE 0 DB/OCT..................
C11 SOURCE: 0...................................................................
C12 PATTERN:   LENGTH       WIDTH...............................................
C13 SWEEP: START 0 HZ END 0 HZ  LENGTH 0 MS.....................................
C14 TAPER: START 0   MS END 0 MS 0 TYPE.........................................
C15 CABLE: MIN 0 MAX 0 INTERVAL 0...............................................
C16 UNITS: 2....................................................................
C17 MAP PROJECTION ZONE ID 0 COORDINATE UNITS 0.................................
C18.............................................................................
C19 HEADER FORMATING INFORMATION................................................
C20.............................................................................
C21.............................................................................
C22 NUMBER OF SAMPLES...........................................................
C23.............................................................................
C24   CASE   NUM_SAMPLES(N)               FILE HEADER             TRACE HEADER..
C25                                  TYPE   OFFSET  VALUE    TYPE   OFFSET  VALU
C26   --------------------------------------------------------------------------
C27 1    N <= 32,767                 uint16  3221   N        uint16  115     N..
C28 1    N <= 32,767                 uint32  3507   N        uint32  225     N..
C29.............................................................................
C30 2    32,767 < N <= 65,535        uint16  3221   N        uint16  115     N..
C31 2    N <= 32,767                 uint32  3507   N        uint32  225     N..
C32.............................................................................
C33 3    N > 65,535                  int16   3221 -9999      int16   115   -9999
C34 3    N > 65,535                  uint32  3507   N        uint32  225     N..
C35   --------------------------------------------------------------------------
C36.............................................................................
C37.............................................................................
C38.............................................................................
C39.............................................................................
C40.............................................................................

[d-chambers]

I just downloaded this and had a look. DASCore uses segyio to parse SEGY files. When I try to load one of these files using segyio I get an error.

import segyio

path = "NNE_MIP3H_v00001_NNE_MW3H_TW3H_S02F_20151107_052552_3_SegyPhase.sgy"

segyio.open(path)

# RuntimeError: trace count inconsistent with file size, trace lengths possibly of non-uniform

ObsPy also fails to read these segy files. I don't know if the trace lengths really are non-uniform (I doubt that for DAS) or if the files are incomplete.

Do you have a way to read them now?

[tiago-pitchon]

I had the same problem with SEGYIO. I also tried to open with some softwares. This afternoon, the professor that ownes this data asked me to try to open another stage. He thinks that the file could be corrupted. I am downloading stages 2 e 3 know. This weekend I tell you my findings.

[tiago-pitchon]

Hello, @d-chambers. Good news!
A professor at my university recommended me to try using SeiSee. This software loads binary part the SEG-Y file ignoring the sampling and spacing. It also give the headers and importante information.

By using this software, I was able to discover that the last byte of the file is junk.

Summary information
------------------------------------------
File : (...)\SegyPhase Stages 01\NNE_MIP3H_v00001_NNE_MW3H_TW3H_S01F_20151106_203124_3_SegyPhase.sgy
 
SEG-Y File
 
ASCII Text Header Encoding
 
Big Endian byte order
 
# Traces            : 493
# Trace Samples     : 60000
Sample Format       : 1
                    : IBM Float (32 bit)
Sample Interval (uS): 500
Time Length         : 29999.5
------------------------------------------
Header First trace Last trace
SP               1           1
CDP              1           1
FFID          2189        2189
Warning: 1 bytes of junk at EOF

I removed this byte using Python and both Segyio and DASCore started working properly!

Now I can start my studies on DAS and practice with this data.

Thanks for your support!

[d-chambers]

@tiago-pitchon,

I am glad its working now for you. It is odd there is a junk byte included at the end of the file. It might be worth contacting mseel folks to see if they can fix the data files, or at least add a note somewhere so others don't spend a lot of time trying to figure it out.

[tiago-pitchon]

Thank you!
I already sent an e-mail to them.