Data Quality Check

Deployment Details Quality Check

  • Scan for any errors or mistakes in the Deployment Details spreadsheet (hydrophone, array, recorder, GPS, dates, etc.)
  • Reach out to deployment/retrieval team for corrections if needed

Check Recording Time: UTC vs Local Time

  • Check log files from extraced SUD files for Soundtrap Recording time (UTC vs Local Time), fix if needed
  • Open Taiki’s soundtrapRenamer script and run it to load all necessary functions
  • Run the following code in a new script to check if the files were offloaded in UTC or local time
source('soundtrapRenamer.R')

# change this to whatever file endings you want to change
suffixes <- c('wav', 'sud', 'log.xml', 'accel.csv')

# Change this to whatever folder these files are in
dir <- '../Data/Renamer/orig/'

prep <- prepTzFix(dir, offset=NULL, suffix=suffixes)
  • If the ‘prep’ output indicates that the files are on local time, run the following line of code to rename them
fixStTz(dir, prep)
  • If files are renamed, the code will create a log file in the same folder of all of the name changes it made
  • There is also a reverse option that you can set to TRUE if you want to undo the renaming

Assess Data Quality

  • Scan the LTSA and define overall deployment as good, compromised, or unusable
  • Record Data Quality in Deployment Details spreadsheet

Run Data Check on GPS Data

  • Review drift tracks created by Driftwatch
  • Look at the image that is associated with the drift number and ensure all points are from the drift (no points from when the array was on the boat or land)

Run Recording QAQC 

  • soundtrapQAQC function in PAMmisc
  • Matlab QAQC script
  • If the QAQC shows significant gaps in the files (i.e. >1 second), complete the following steps to account for those gaps

    • If you have a gap in the middle of the drift (with significant amount of data before AND after the gap) continue on with the following steps and add a readMe to the data folder indicating when and how long the gap is

    • If you have duty cylced recordings at the end of a continuous recording period, cut off the data

Scan LTSA for Noisy Data

If drift has been determined to be compromised in the data quality check above, scan the LTSA to log time periods with noisy/compromised data

  • Use the LTSA created with 500 Hz wav files, with 1s and 5Hz parameters
  • View LTSA at 2 hour windows and scan through
  • Use Logger to define start/end times of the noisy/bad data and include the type of noise in the comments (e.g. strumming, impulses, etc.)
  • Split between Anne and Cory