Johnson Shut Ins     Data Processing notes


1) RAW DATA:
     Raw data is that which is copied from the USB disk. In general it is a full copy
     of what is pn the BALER - however that happens  - note that for some types of BALERS 
     the processes of copying the data can be quite compicated. In this case the copying
     of data is trivial - just copy the USB disk
     
     Raw data is stores on a network storage device, which although slow, is attached
     to many servers (old like terra1 and terra3, new like mantle). It is slow, but 
     we only need to access the data one time. After that it is just kept as an 
     archive.
     
     The network storgae device is /seismonas and the raw seismic data storage  location
     is /seismonas/RAW-SEISMIC-DATA
     Here you will see copies of BALERS from 
          GAMSEIS, POLENET , PATAGONIA, ALASKA,  MADAGASCAR
     and this project Johnson Shut Ins called Johnson_Shut_Ins
     
     In this directory there is a directory for each service run - called 
     serv-1, serv-2, serv-3
     This is true for all progects. The contents of each serv-X directory is a 
     complete copy of each stations's BALER.
     
     (You may start to see a pattern here - Consistancy is good)
     
     In each service directory, there is a directory for each station. 
     THE ONLY THING TO GO IN THIS DIRECTORY IS A COPY OF THE BALER - which
     may consist of a set of multiplexed files or may simply be copies of the 
     two USB disks.
     By convention "disk1" is the first disk written to. In this case - because the 
     service interval was so short - it is the only disk with data on it.
     "disk2" is typically empty.
     
   
2) MINISEED DAY FILES:
      Again, by convention, data is archived as day files (data for one julian day) 
      stored in miniseed format. There are a number of options for file naming and 
      determination of which data to keeep based on timing quality. USUALLY we want 
      to mark that data for which the clock quality was not acceptable for detailed
      seismic work.
      
      We are going to make a database in the dircetory
         /RAID/DB/Johnson-Shut-Ins  called jsp  using all the data
      and one called jsp-good-timing using only the data with good GPS clock quality
      
      the day miniseed files are going into:
         /RAID/DB/Johnson-Shut-Ins/data_files/serv-X/{STATION} - for good timing
         /RAID/DB/Johnson-Shut-Ins/data_files_incl_bad_timing/serv-X/{STATION} - for ALL data
         
      Note that programs used for this processing are in a directory called "programs"
      and that any processing notes are in a directory called "notes"
      (Again you might see a pattern here).
      
      
NORMAL TIMING QUALITY:

serv-1 
   JSP1    2022231 to 2022240,  2022242 to 2022242 
   JSP2    2022230 to 2022268 
   JSP3    2022230 to 2022268 
serv-2 
   JSP1    2022269 to 2022278,  2022280 to 2022296 
   JSP2    2022268 to 2022296 
   JSP3    2022268 to 2022296 
serv-3 
   JSP2    2022296 to 2022335 
   JSP3    2022193 to 2022194 

         
INCL BAD TIMING
         
serv-1 
   JSP1    2022231 to 2022240,  2022242 to 2022242 
   JSP2    2022230 to 2022268 
   JSP3    2022230 to 2022268 
serv-2 
   JSP1    2022269 to 2022296 
   JSP2    2022268 to 2022296 
   JSP3    2022268 to 2022296 
serv-3 
   JSP2    2022296 to 2022335 
   JSP3    2022193 to 2022194 
     

So the only bad timing data is serv-2:JSP1: 2022:279


So do some renaming - since only one day has bad timing and we don't care anyway:
                  rm -r data_files
                  mv data_files_incl_bad_timing data_files
so now data_files is the only data directory



GET RID OF THE LOC CODES
1)    rename all the files so as to NOT have the .00. field
      use   programs re-name-files.pl    on each serv-X and JSPX directory
      
2) then use fixhdr to remove the loc code from the header 
    this only works on terra1-2-3  not on mantle
    
    
    
    
RENAME (1) to (N) and (2) to (E)
   for some reason the database  entry "datalogger Passcal_q330_linear" 
   gives     ELZ  ELN   ELE     instead of     ELZ    EL1     EL2
   and   no  LL1   LL2     LLZ  make LLZ    LLN     LLE
    
    
1) rename all horizontal data file channels to be; ELN   ELE     LLN  LLE
    use  programs/re-name-1-2-chan.pl

2) change the mseed headers  to be  ELZ  ELN  ELE    LLZ  LLN  LLE
   using fixhdr   
   
   NOTE:    fixhdr does not work on "mantle"