Digital Elevation Model from Single-Beam Bathymetry XYZ Data Collected in 2015 from Raccoon Point to Point Au Fer, Louisiana

DGPS Navigation Processing:The GPS base stations were established by USGS personnel for the purpose of this survey. Benchmark PAF1 was located to the east of the survey area on a marsh island to the west of Fish Bayou, PAF2 was located in the middle of the survey area on the eastern shore of Oyster Bayou, and PAF3 was located in the western portion of the survey on the middle shore of Point Au Fer Island (http://pubs,usgs.gov/ds/01041/html/ds1041_overview.html). The base stations were equipped with Ashtech Proflex GPS Receivers recording 12-channel full-carrier-phase positioning signals (L1/L2) from satellites via Thales choke-ring antennas, recording at a rate of 0.1 seconds (s). GPS instrumentation was duplicated on all survey vessels (rovers) with the exception of the R/V Shark (15BIM07) and the R/V Chum (15BIM08) on which Ashtech Global Navigation Satellite Systems (GNSS) were used instead of Choke-ring antennas. The base receivers and rover receivers recorded positions concurrently at all times throughout the survey. Rovers R/V Shark and R/V Chum recorded at 0.1 s throughout the survey while the R/V Jabba Jaw and the R/V Sallenger recoded at 0.2 s. The coordinate values for each of the GPS base stations (PAF1, PAF2, PAF3) are the time-weighted average of values obtained from the National Geodetic Survey's On-Line Positioning User Service (OPUS). All base station sessions of recorded data are decimated to 30 s, and then submitted to OPUS via the online service. All solutions are then returned to the user and entered into a spreadsheet where time-weighted ellipsoid values are calculated for each station, for the entire occupation. Any individual ellipsoid value that falls outside three standard deviations for the entire occupation was excluded and the final coordinate values were then determined. The final base station coordinates were imported into GrafNav version 8.5 (Waypoint Product Group) and the kinematic GPS data from the survey vessel were post-processed to the concurrent GPS data from the base stations. In all cases the closest base station to the roving vessel was utilized to help keep vertical error to a minimum. During processing, steps were taken to ensure that the trajectories between the base and rover were clean and resulted in fixed positions. By analyzing the graphs, trajectory maps, and processing logs that GrafNav produces for each GPS session, GPS data from satellites flagged by the program as having poor health or satellite time segments with cycle slips were excluded, and the satellite elevation mask angle was adjusted to improve the position solutions when necessary. The final, differentially-corrected, precise DGPS positions were computed at their respective time intervals (0.2 s for 15BIM05, 15BIM06 and 0.1 s for 15BIM07, 15BIM08), and then exported in ASCII text format. The file was then used to replace the uncorrected real-time rover positions recorded during acquisition. The GPS data were processed and exported in the World Geodetic System of 1984 (WGS84) (G1150) geodetic datum UTM zone 15N.

Single Bream Processing: The raw HYPACK® data files were imported into CARIS Hydrographic Information Processing System (HIPS) and Sonar Information Processing System (SIPS) ® version 9.0.17. The corrected DGPS positions exported from GrafNav were imported into CARIS using the generic data parser tool. After parsing, the navigation data were scanned using the Navigation Editor, which allows the user to view multiple types of plots including trackline orientation, timing, and course direction. This check verifies if the parsed data corresponds to the processed DGPS. Next, Speed of Sound Profile (SVP) casts were entered, and edited, using the SVP editor tool, and then applied as nearest in distance within time. All soundings are referenced to the WGS84 ellipsoid during processing. This involved a step in CARIS to compute the GPS tide. The GPS tide represents the ellipsoidal surface. GPS tide and GPS height are then compared against each other to ensure correct computation by the program and applied GPS antenna height provided in the vessel file. All bathymetric data components; position, motion, depth, GPS tide, and Speed of Sound (SOS), were then merged and geometrically corrected in CARIS to produce processed x,y,z point data. Once merged, the dataset is reviewed for erroneous points using the Single Beam Editor. The points that are visually obvious outliers are often related to cavitation in the water column obscuring the fathometer signal, tight turns in the surf zone affecting the tracking of the incoming GPS signal, and/or false readings due to general equipment issues. Data showing these are either discarded or adjusted to surrounding sounding depths. Also, data points in areas of extremely shallow water (0.30 m to 0.50 m) such as on shoals or within seagrass beds were reviewed against the surrounding data for overall consistency. Finally, a Bathymetry with Associated Statistical Error (BASE) surface was created. Using the Subset Editor, the BASE surface was used as a color coded guide to pinpoint crossings that are visually offset from one another. If an offset was identified, it was further examined and was reprocessed if necessary. The geometrically-corrected point data are then exported as an x,y,z ASCII text file referenced to WGS84 (G1150), equivalent to ITRF00, UTM 15, and ellipsoid height in meters. The single-beam bathymetry datasets combined (15BIM05, 15BIM06, 15BIM07, and 15BIM08) consists of 4,626,476 x,y,z data points with an ellipsoidal height range of -35.096 m to -25.904 m.

Quality Control and Quality Assurance (QA/QC) and Datum transformation: All single-beam data found in the ASCII exported from CARIS were imported into Esri ArcMap version 10.2, where a shapefile of the individual data points (x,y,z) was created and plotted in 0.5-m color coded intervals. First, all data were visually scanned for any obvious outliers or problems. Then, the data were run through an in-house script similar to a Python script. The script was created for the purpose of evaluating elevation differences at the intersection of crossing tracklines by calculating the elevation difference between points at each intersection using an inverse distance weighting equation. Elevation values at line crossings should not differ by more than the combined instrument acquisition error (per manufacturer specified accuracies). GPS cycle slips, stormy weather conditions, and rough sea surface states can contribute to poor data quality. If discrepancies that exceed the acceptable error threshold were found, then the line in error was either removed or statically adjusted only by the average of the crossings within the line. The script was run on all vessel point data first on a vessel by vessel basis and then run a final time for all data points from all vessels collectively by use of a merged shape file with all point data (2015-320-FA_SBB_Level_04_1041_ITRF00). For this dataset, the collective run is the most important as the survey structure did not allow for sufficient vessel crossing, rather vessels crossing other vessels. Once the dataset passed all QA/QC procedures and manual editing steps, the data were considered final, pending datum adjustments. The single-beam bathymetric data were transformed horizontally and vertically (via the transformation software VDatum version 3.2) from their data acquisition datum WGS84 (ITRF00) to the NAD83 CORS96 reference frame using the NGS GEOID12A. The NAD83 x,y,z data points were exported in ASCII.

Gridding Bathymetric data and computing grid error: The single-beam soundings were imported into Esri’s ArcMap version 10.2 and gridded using Spatial Analyst tools "create tin," "tin to raster," and "extract by raster mask." First, a bounding polygon representing the extent of survey tracklines was created and converted into a raster mask using the ArcGIS "polygon to raster" tool. Next, the final point data were merged into a single shapefile and converted into a triangulated irregular network (TIN), using the "create tin" tool and then the data were converted into a raster DEM by use of the "Tin to raster" tool using the natural neighbor function with a cell size of 200 m. Finally, the interpolated DEM is clipped to the survey extent utilizing the "extract by mask" tool and using the raster mask created in the first step of this process. The final product is a DEM of the entire survey extent, with grid elevation values ranging from -0.40 m to -7.25 m. In order to evaluate how well the final DEM represents the final sounding data both spatially and quantitatively, a comparison of the DEM versus the sounding (x,y,z point) data was plotted in ArcGIS by use of the “Extract values by points” spatial analyst tool. This tool extracts the value represented by the underlying grid and compares it to that of the overlying point data. By use of the generated shape file and associated attribute table, the root mean square (RMS) error, quantified as the difference between the measured depth and the grid depth values, and calculated. The overall RMS Error in meters is 0.16.

Added keywords section with USGS persistent identifier as theme keyword.