[an error occurred while processing this directive]
LETTER REPORT: BATHYMETRY OF THE CAPE FEAR RIVER EBB TIDAL DELTA, 2004.

Dr. Jesse E. McNinch
Department of Physical Sciences
Virginia Institute of Marine Science
College of William and Mary

Bathymetry Methodology
Acquisition System:
        234 kHz Interferometric Swath Sonar
        Ixsea Octans motion sensor and gyro
        Trimble Dual-channel RTK GPS

Bathymetry was collected using an interferometric swath sonar system integrated with a motion sensor that removed vessel motion in real-time in January 2004. Dual-channel RTK GPS provided horizontal and vertical control. Horizontal and vertical control during the 2004 survey was obtained from the GPS base and radio station operated by the USACE-SAW at the Oak Island Coast Guard Station (OICG). Geodetic coordinates for the OICG benchmark are 3353'34.23090"N, 7802'05.69403"W, with a height of 14.159m. The base radio is a TrimMark IIE broadcasting at a frequency of 148.725 MHz. Soundings were corrected for water level fluctuations forced by astronomical tides and wind-driven tides using the vertical RTK-GPS measurements. A KTD file, provided by the USACE-SAW for the survey region, was used to correct GPS-measured height to a vertical tidal datum. The KTD file was established relative to MLLW, as that is the datum used by the dredging company. The vertical datum was shifted from MLLW to NGVD29 using the tidal observations from Oak Island (near the inlet) and Baldhead Island. As such, all soundings are provided in state plane, NAD83 and NGVD29. The Coastal Oceanographics Hypack survey program was used to navigate survey track lines and log vertical tide files.

Post-processing using proprietary Submetrix Interferometric swath sonar software, SwathPlus, corrected for errors associated with speed of sound variations and low-frequency vessel motion (portion not removed by motion sensor). Changes in speed of sound, generated by density stratification in the water column, were measured periodically during the survey with CTD casts. Processing algorithms within SwathPlus also provide a continual assessment of speed of sound and vessel motion artifacts and allow removal of these artifacts during processing. Processed swath sonar files were then imported to Gridproc, a proprietary Submetrix program, for data gridding. A nearest neighbor, weighted gridding algorithm determined depths at irregularly-spaced, 2 m grid nodes from swath soundings. Grid soundings were exported in ascii format as x, y, z (m, state plane and NGVD29, respectively).

These highly anisotropic soundings were then imported to IVS Fledermaus and despiked using a standard deviation threshold followed by gridding into a regularly-spaced, rectilinear grid using a kriging algorithm weighted for anisotropic data. Bathymetric data from the USACE LARC cross-shore transects, also collected in January 2004, were imported within the x-y-z spreadsheet and gridded at 20-25 m grid nodes with the soundings collected by the interferometric system. The combined survey track lines from the LARC and interferometric system are shown in Figure 1.

An assessment of quality control was undertaken by comparing the soundings collected by the LARC with interferometric soundings. Individual grids created in Fledermaus from the LARC and interferometric data were compared and the vertical difference between each grid node was determined (Appendix I). Statistics from the 2004 LARC and inteferometric soundings yielded a median difference of 3 cm with a standard deviation of 68 cm. A median was calculated to identify a consistent vertical error That could occur with one of the collection methods such as using different tidal datums or errors associated with the base station(s). The relatively high standard deviation is likely a result of gridding regions over several meters that have significant vertical relief such as the channel edge. Soundings collected just a couple of meters apart yet grouped within the same grid node could easily reveal different grid node averages due to real bathymetric differences in regions of high relief. A clear example of this is seen in the comparison of survey years 2004 and 2000. As expected, this comparison has the largest standard deviation because it reflects the substantial depth changes associated with dredging of the new channel.

Contour and 3-D surface plots were generated using Fledermaus from the combined LARC and interferometric data and exported as *.tiff or *.jpg image files. Finally, x, y, z data in ascii format were exported from the grid files at a 25 m grid node spacing. All image files and ascii data files were digitally transferred (ftp) to the USACE FRF and will be posted on the website (www.frf.usace.army.mil/capefear).

Results

Final processed and gridded soundings from the combined interferometric and LARC surveys are provided as an x-y-z ascii spreadsheet (filename: MMMM.dat) at a spatial interval of 25 m in North Carolina state plane, m, and NGVD29, m. Although these sounding were collected over a time span of several weeks and should not be used as an instantaneous measure of bathymetry or for navigation purposes, gross patterns of seafloor morphology can be seen. A contour plot of all bathymetry and shoreline data collected in 2004 is shown in Figure 2. Bathymetry and shoreline data collected in January 2004 from Baldhead Island, highlighting South and West Beach and Baldhead Shoals, are shown in Figure 3. Figures 4 - 6 show changes in bathymetry between previous survey years. These contour maps illustrate the bathymetric difference between surveys, where yellow to red colors denote negative values or deepening and green shading with positive values indicate accretion. Beyond the obvious channel deepening from dredging, differences most noteworthy from these comparisons include: 1) accretion of West Beach between the 2000 and 2002 surveys 2) accretion along South Beach that appears to remain largely intact as of 2004, 3) loss around the Point of West Beach and accretion of Baldhead Shoals, and 4) loss of initial accretion along much of West Beach by the time of the 2004 survey.

The ebb tidal delta surrounding the mouth of the Cape Fear River is shown in Figure 7. As apparent in previous years, three linear shoals compose much of the ebb tidal delta. Two prominent shoals are present on the west side of the shipping channel, known collectively as Jay Bird Shoal, and a well-developed flood margin channel flanks Oak Island. Figures 8 - 10 show changes in bathymetry between previous survey years. These comparisons reveal deepening of the flood margin channel near Oak Island and accretion on Jay Bird Shoals and Baldhead Shoals. Construction of the new channel is apparent from the bathymetric charts with the 1999/2000 surveys reflecting a pre-dredge chart while the 2001/2002 were conducted at the very end of construction just prior to the cut-in with the old shipping channel. The bathymetry seen in the 2002/2003 and the most recent 2004 surveys, therefore, shows the morphology of this region after roughly 10 months and 2 years (respectively) of the new channel being opened. Little change appears to have occurred on the shoal between the two channels. We expect this shoal to be the most sensitive to changes in mean currents through the channels because it is located between the channels where the magnitude of mean currents around the distal end of the ebb tidal delta are the highest (see associated ADCP reports). In short, the gross morphology of the ebb tidal delta remains relatively static which suggests there have not been substantial changes in the sediment transport pathways around the ebb tidal delta since the 1999/2000 surveys.

A similar set of figures showing the shoreline and nearshore bathymetry along Oak Island are presented in Figures 11-14. A consistent trend of nearshore accretion along much of the south-facing beach is apparent as well as persistent deepening of the flood margin channel at the eastern end of Oak Island.