  1.  NAME

  GFI ADCP - Quick Reference for ADCP CODAS Databases

  2.  DESCRIPTION

  This manual page describes how to create CODAS databases from ADCP
  data files with emphasis on the instrumentation at GFI.

  If data have been acquired by DAS, refer to the CODAS User's Manual.
  This is file Manual.ps usually found in directory adcp/doc in the
  CODAS package installation directory.

  If processed data acquired with TRANSECT are to be loaded, use
  loadtnbp(1) or loadtbbp(1) for narrow band or broad band ADCP,
  respectively. Refer next to the CODAS User's Manual.

  If raw data from a narrow band ADCP acquired with TRANSECT are to be
  loaded, read the rest of this manual page and refer then to the CODAS
  User's Manual.

  3.  INTRODUCTION

  In 1997, a SEAPATH 200 was installed M/S Ha/kon Mosby to get better
  attitude parameters and therefore improve current measurements over
  areas where bottom tracking was not available. At this time, there was
  no possibility to interface these data to the on board ADCP. Hence, a
  special interface to the synchro board has been developed for this
  purpose. It took however some time before this interface was built.
  During this period, GPS heading was logged separately, together by
  TRANSECT with the navigation data.

  Heading corrections could have been carried on five minutes averaged
  ensembles. However, because most of the data had been acquired in the
  Norwegian Sea, which is at least not famous for its calm conditions,
  it was first decided to write a utility to correct TRANSECT raw files
  and re-average them in playback mode by TRANSECT. However, it was
  found out very fast that this method was not reliable, because
  TRANSECT was not able to read always these corrected files (later it
  was also found that even some untouched files were rejected). Hence, a
  full averaging for these raw files was necessary.

  A simple utility was first written. It showed that data averaged with
  it were in many cases more reliable than those from TRANSECT processed
  files. Therefore, this utility has been further developed and this
  method of averaging ADCP data has been chosen.  Its main disadvantage
  is that the acquired number of data is enormous compared to processed
  files (about 300 times). However, nowadays available hard disks and
  CDROM make it possible to apply this method and transport data without
  too many problems. Once averaged and loaded into a CODAS database, the
  file sizes do not differ.

  If averaging and loading of NBR ADCP is chosen, it is highly
  recommended to scan the raw ADCP and navigation files first. This step
  the same function as running scanping on DAS ping data files. It
  provides in addition the possibility to visualize these data within
  Matlab, and to output a synchronized navigation file for use during
  averaging. Once this step is carried out and data for averaging
  selected, the real process of averaging and loading data into a CODAS
  database can be performed. For further processing of ADCP data, refer
  to the CODAS User's Manual.

  4.  SCANNING RAW FILES

  This step is carried out by scan_nbr(1). Its main purpose is to check
  the consistency of the data files. It also outputs information about
  configuration changes during acquisition, as well as malfunction of
  the transducer (the BIT status). Optionally, one can output some
  information about selected acquired variables and visualize them in
  Matlab with scan_bin(1m). This is useful if only parts of the acquired
  data are to be loaded into the database.

  5.  SCANNING NAVIGATION FILES

  This step has the same purpose as the one for raw files, but it is of
  course designed for navigation files. It is carried out by
  scan_nav(1). As for raw files, selected variables can be selected for
  output and visualized within Matlab with scan_bin(1m).

  6.  LOADING ADCP DATA

  The utility process(1) will merge and synchronize NBR ADCP and
  navigation data, average and load them into an ADCP CODAS database.
  For this purpose, producer definition file adcp_hm.def can be used.
  Some of the log information output during this process can be
  visualized within Matlab with process(1m).

  7.  SEE ALSO

     gfiindex(5)
        Index of all GFI extensions.

  8.  AUTHOR

  Pierre Jaccard, Geophysical Institute, University of Bergen, 1999.

