Sound Velocity Profiles collected in Nantucket Sound Massachusetts in the vicinity of Horseshoe Shoal, during USGS Field Activity 2023-001-FA using AML-3 LGR or AML Minos-X CTDSV sensors (PNG images, SVP text, and ESRI point shapefile, GCS WGS 84)

PROCESSING STEP 1: COLLECT SOUND VELOCITY PROFILES. Sound velocity profiles were collected 1-7 times per survey day using AML-3 LGR or Minos-X CTDSV sensors. Cast timing was determined depending on water column properties and the quality of real-time multibeam bathymetric data acquisition. The profiler sensor was hand-deployed from the after deck of R/V Rafael while the vessel was not making way. Once back on deck, data from the sensor were downloaded using either AML SeaCast (version 4.4; for AML-3 LGR) or Sailfish (version 4.4; for AML Minos-X) software, then files containing the full cast information were saved to the raw AML format (.csv) and downcast portions were saved to the CARIS SVP (.svp) format.

PROCESSING STEP 2: SOUND VELOCITY PROFILE PROCESSING AND EXPORT FROM QIMERA.
Sound Velocity profiles for each day were imported and converted to Qimera (version 2.5.0) format using the Source/import/Caris svp menu. Each profile was reviewed for positional accuracy and erroneous depth/sound velocity values using the SVP editor. Incorrect records near the surface or within the water column were flagged using the "reject selection" context menu.
Edited SVP data were exported from Qimera using the Export/svp/export to CARIS SVP option, resulting in one svp text file that included all 27 svp casts.

PROCESSSING STEP 3: CREATE SVP PROFILE TEXT FILES, PNG IMAGES, AND POINT NAVIGATION FEATURE CLASS.
A python Jupyter notebook was used to read the single svp file exported from Qimera, split it by profile, and write out a CARIS SVP format file for each profile.
A second python Jupyter notebook utilized the python modules pandas (version 1.5.2), geopandas (version 0.12.1), shapely (version 1.8.5.post1), matplotlib (version 3.6.2), and sqlite3 (version 3.40.0) to read the individual svp files and create a PNG plot for each, containing speed of sound (meter/second) on the x-axis and depth (meters) on the y-axis. Position, Date, Time, and Profile ID were extracted from the SVP header line and written into the image as a title line for reference. The image files were named based on the date and time each sound velocity cast was collected. The latitude and longitude position, Cast ID, and Date/UTC time information for each cast, along with fields containing the Survey ID, Device ID, and Vehicle ID were appended as rows by profile to a Pandas dataframe during processing. After all profiles were processed, geopandas and shapely were used to create a point geometry column from the geographic station coordinates and the composite dataframe was imported to a geospatial SQLite database table using the pyspatialite interface.
Finally, the svp location SQLite database table was imported to QGIS (version 3.26) as a point feature layer and exported (Right click on database feature class > Data > Save Features As) to the ESRI point shapefile 2023-001-FA_SVP_data.shp.