CRUISE REPORT: RV COASTAL SURVEYOR CRUISE C-1-99-SC
MULTIBEAM MAPPING OF THE LONG BEACH, CALIFORNIA, CONTINENTAL SHELF
April 12, through May 19, 1999
Long Beach, California
By James V. Gardner1, John E. Hughes Clarke2, and Larry A. Mayer2
Open-File Report 99-360
Conducted under a Cooperative Agreement between the US Geological Survey and the Ocean Mapping Group, University of New Brunswick
This report is preliminary and has not been reviewed for conformity with U.S. Geological Survey editorial standards or with the North American Stratigraphic Code. Use of trade, product, or firm names in this report is for descriptive purposes only and does not imply endorsement by the U.S. Government.
U.S. DEPARTMENT OF THE INTERIOR
U.S. GEOLOGICAL SURVEY
1 Formerly at MS 999 345 Middlefield Road, Menlo Park, CA 94025; now at University of New Hampshire, Center for Coastal and Ocean Mapping, Jere A. Chase Ocean Engineering Lab., 24 Colovos Road, Durham, NH 03824
2 Ocean Mapping Group, University of New Brunswick, Fredericton, NB E3B 5B3
The Multibeam Mapping Systems
Kongsberg Simrad EM3000D
Pulse Width and Bandwidth
Sound Velocity Measurement
Sidescan Strip Compilation
Cruise Report for C-1-99-SC
Multibeam Mapping the Long Beach Continental Shelf
The greater Los Angeles area of California is home to more
than 10 million people. This large population puts increased pressure
on the adjacent offshore continental shelf and margin with activities
such as ocean disposal for dredged spoils, explosive disposal,
waste-water outfall, and commercial fishing. The increased utilization
of the shelf and margin in this area has generated accelerated
multi-disciplinary research efforts in all aspects of the environment
of the coastal zone. Prior to 1996 there were no highly accurate
base maps of the continental shelf and slope upon which the research
activities could be located and monitored. In 1996, the United
States Geological Survey (USGS) Pacific Seafloor Mapping Project
began to address this problem by mapping the Santa Monica shelf
and margin (Fig. 1) using a state-of-the-art, high-resolution
multibeam sonar system (Gardner, et al., 1996; 1999). Additional
seafloor mapping in 1998 provided coverage of the continental
margin from south of Newport to the proximal San Pedro Basin northwest
of Palos Verdes Peninsula (Gardner, et al., 1998) (Fig. 1). The
mapping of the seafloor in the greater Los Angeles continental
shelf and margin was completed with a 30-day mapping of the Long
Beach shelf in April and May 1999, the subject of this report.
The objective of Cruise C-1-99-SC was to completely map the broad
continental shelf from the eastern end of the Palos Verdes Peninsula
to the narrow shelf south of Newport Beach, from the break in
slope at about 120-m isobath to the inner shelf at about the 10-m
Mapping the Long Beach shelf was jointly funded by the U.S. Geological Survey and the County of Orange (CA) Sanitation District and was conducted under a Cooperative Agreement with the Ocean Mapping Group from the University of New Brunswick (OMG/UNB). The OMG/UNB contracted with C&C Technologies, Inc. of Lafayette, LA for use of the RV Coastal Surveyor and the latest evolution of high-resolution multibeam sonars, a dual Kongsberg Simrad EM3000D.
Figure 1. Areas mapped in 1996 (Santa Monica Bay), 1998 (Long Beach to Newport Beach slope) and this survey (Long Beach shelf).
The Multibeam Mapping Systems
The Kongsberg Simrad EM3000D Mounting
The EM3000D consists of two EM3000 transducers; one mounted facing to port and the other to starboard (Fig. 2), each tilted 30° from the vertical. Each transducer provides up to 127 receive beams that span from +65° to -65° to the normal to the transducer (Fig. 3) with a 60° overlap zone. As will be discussed below, this overlap results in some noticeable cross talk at a decimeter scale, even though the two head are separated in frequency by about 6 kHz. Although the transducer mountings were located at exactly the same level, the transmit and receive arrays were located off center within the housings, resulting in the starboard array being physically ~10 cm higher than the port array. Initially, in order to avoid this problem, the starboard array was mounted backwards so that the centers of the two arrays were mounted at exactly the same elevation. However, even though a 180° mount
Figure 2. Simrad EM3000D transducers mounted on bow ram.
offset was entered into the Simrad software, this orientation caused an unexpected problem. The pitch steering algorithm is coded to perform properly only if the arrays are mounted in the usual fore-aft direction. But, because of the switched starboard transducer orientation, the pitch steering actually translated the starboard side beam solutions forward during a nose-up pitch event. Consequently, both transducers were remounted in a fore-aft common alignment. The physical offset between the two acoustic centers was accounted for by entering the mounting coordinates into the Simrad control software. To minimize interference, each transducer uses a different frequency range. The port transducer initially used the lower frequency of 297 kHz, and the starboard transducer used 303 kHz. This configuration produced poor bottom detection in low-backscatter environments on the starboard transducer. However, when the frequencies for the two transducers were switched, the two transducers appeared to perform identically.
Figure 3. Schematic of the port and starboard beam coverage of EM3000D with 60° overlap zone at nadir.
The beams from each receiver are spaced according to the angular equivalent of the bins in a spatial Fast Fourier Transform (FFT) across the receive array. This results in an unusual beam spacing whereby the beams are spaced neither equiangular nor equidistant. Rather, the beam angular spacing increases away from the broadside direction. The beam spacing at nadir is 0.9° and increases with the secant of the beam steering angle. For a single EM3000 transducer, the resulting beam spacing at 60° is 1.8° apart, whereas for the dual transducer EM3000D with a mount angle of 30°, the beam spacing is only 1.04°. At the maximum use angle of 75° from vertical, the beam spacing for an EM3000D is 1.27°, with a resulting beam spacing that corresponds to 36% of the water depth (compared to only 1.6% at nadir). The spatial resolution in the bathymetric data is highly variable across the swath, ranging from excellent at nadir to poor in the outermost swath. This variable spatial resolution results in unusually poor small-scale resolution in bathymetry for the low grazing-angle solutions. However, as will be discussed below, the backscatter imaging actually improves toward the low grazing angles. Because the beam spacing is fixed with reference to the array (because the FFT beamforming is accomplished according to the array element spacing), the physical beams rotate with the transducer. Consequently, no electronic roll stabilization can be attempted. In practice, because the combined beams from both transducers cover a sector from +95° to 95°, the beams within the usable sector are retained. Also, depending on the sound speed at the transducers, not all the beam locations (spatial frequency indices in the FFT) actually correspond to physically realizable angles in front of the receive array. In fact, only rarely will 127 beams actually be preserved.
The transmit beam width of the EM3000D is 1.5° by ~140° (one each for each transducer). The receive beams are all formed from a linear receive array so the narrowest beams are formed at broadside (in this case at +30° and 30° off the vertical). The beam widths grow with the secant of the transducer-relative steering angle. The dual-transducer EM3000D mounted at 30° requires only 30° of beam steering to project to 60° off vertical and generates a 1.73° beam at this angle, whereas a single-transducer EM3000S beam at 60° from the vertical would generate a beam width of 3.0°. At the widest-angle data used (~75° off vertical,) the receive beam widths of the EM3000D are only 2.12°.
Pulse Width and Bandwidth
The EM3000 allows user-selectable pulse widths from 0.05 to 0.20 ms and receive bandwidths from 4 to 8 kHz. In principle, these could be varied with water depth, going from shortest pulse length and broadest bandwidth at shallow depth, etc. However, it is clear that these changes are not properly compensated for in the EM3000 backscatter strength estimates. To avoid this problem, the pulse width was set at 0.15 ms and bandwidth at 8 kHz for all operations, as suggested by Kjell Neilsen of Kongsberg Simrad.
Each transducer of the EM3000D, mounted as described above, should capture any echo over an arc from +95° to -95°. Obviously, for most of the shelf, no echoes exist above the horizontal. Operationally, if the sonar were to wait for echoes that might occur above 80° off the vertical, the ping rate would be greatly reduced. A new feature recently added to the EM3000 allows the sonar to wait for arrivals out only to an operator selectable incidence angle (measured with respect to the vertical). This change allows the ping rate to be kept to reasonably fast limits. The change also prevents acquisition of data at grazing angles that are unrealistic because of signal-to-noise levels. Note that as the water depth gets deeper, attenuation quickly reduces the signal-to-noise level and the swath angular sector rapidly narrows. One drawback of the user-selectable sector is that the sonar waits for, and tries to process, echoes out at the operator-selected angle even though in deeper water there is nothing but noise at the far range. This results in very low ping rates and a high percentage of unreliable bathymetric data that must be edited out. For example, below ~80m, the only data recovered was within ~+40° to -40° of nadir, resulting in complete overlap between the two sectors.
The EM3000D provides a single depth telegram corresponding to the summed 254 possible beam locations. In most cases, each of the central 40 beams overlaps adjacent beams. The sonar achieved ping rates faster than 7 Hz in water depths shallower than 10 m.
The EM3000D was interfaced with an Applied Analytics TSS POS/MV 320 vehicle motion sensor. Initially, a newer version of POS/MV hardware and software was used, but a technical problem required the older instrumentation to be reinstalled. The POS/MV provides roll, pitch, heading (yaw), and heave. The three angles are rated to ~0.05° and the heave is rated to better than 5 cm or 5% of the short-period vertical displacements. However, all these angles are only as good as the accuracy of the instrument alignments. Despite extensive patch testing, doubt remained throughout the survey of the validity of the mounting angles. It was easy to confuse changes in surface sound speed with false offsets in the head mounting locations. Several patch test values were entered at different stages of the survey to investigate these motion artifacts (Fig. 4). The biggest uncertainty remained about a yaw misalignment of the vehicle motion sensor with respect to each of the sonar transducer. A separate yaw alignment was calculated for each transducer during the patch test but a deviation in yaw alignment angles of one or both of the EM3000 transducers resulted in a strong roll-pitch cross talk in the data (Fig.5). We decided to use zero yaw misalignment angles for the entire survey and will investigate this problem during post-cruise processing. However, this decision resulted in a roll-pitch cross-talk artifact in the logged data that manifests itself in a systematic 10 to 15 cm rippling in the bathymetric data (Fig. 6). This will eventually be removed in postprocessing.
Figure 4a. Profiles of all 254 beams from 80 pings stacked in profile, looking along track. The 60° zone of overlap between port and starboard receivers (from red arrow to green arrow) of the swathes of the port (red) and starboard (green) receivers shows good agreement. Compare to Figure 4b. Vertical scale 1 m between lines.
Figure 4b. Profiles of all 254 beams from 80 pings stacked in profile, looking along track. The 60° zone of overlap between port and starboard receivers (from red arrow to green arrow) of the swathes of the port (red) and starboard (green) receivers shows poor agreement. Compare to Figure 4a.
Figure 5. Example of processing screen showing a strong correlation
between ship pitch
and an apparent artifact in the data (lower and middle left panels). The vertical dashed lines
show the correlation.
Figure 6. Ripple artifacts induced in the bathymetric data
by a subtle sensor misalignment.
Both the POS/MV inertially smoothed DGPS solution and the Trimble 4000 SEDGPS solution was logged by the Simrad system. The POS/MV solution corresponds to the location of the vehicle motion sensor whereas the Trimble solution corresponds to the position of the DGPS antennas. Only the POS/MV solutions were used for processing because they maintained very high quality positioning throughout the survey. DGPS corrections were obtained from a satellite down link to the vessel.
Sound velocity was measured at the transducer and fed directly into the Simrad software. In addition, vertical sound-speed profiles of the water column were obtained at least once for each 12-hr sortie. Additional deployments were made if significant refraction-induced errors were noticed aboard ship. Transit lines to and from the survey area required particularly severe refraction corrections because the transits crossed several water masses. To avoid these problems and the more subtle problems due to slowly evolving water masses throughout the day, the Univ. of New Brunswick Ocean Mapping Group refraction software tool was used. This tool allows the operator to apply qualitative adjustments to the bathymetric data based on the user's perception of a systematic residual in the water-column refraction. Although these refraction corrections are entirely empirical, they are completely reversible at any time.
The EM3000D provides a time series of backscatter-intensity measurements through the centers of each of the beam footprints. These time series only overlap from beam-to-beam if the seafloor slope is steeper than the expanding acoustic wave front. As long as the time series are properly registered according to the beam locations, a complete record is obtained of the received time series, equivalent to a conventional sidescan image. The EM3000 imagery also has full across-track topographic correction. However, the system has two disadvantages; (1) it has a high aspect ratio because the transducers are hull mounted rather than deep towed and (2) it has a 1.5° beam width rather than a typical 0.75° or less for a towed sidescan sonar. In principle, multibeams record the backscatter-intensity values using a time-varying gain (TVG) that is designed to compress the logged data into measures of the bottom-backscatter strength. Also a number of simplifying assumptions are used, such as locally flat seafloor, Lambertian response, etc. These do not work well with the EM3000D because changes in the ensonified area related to changes in the pulse length were not being properly compensated.
For the first 3 days, a bug in the Simrad sidescan collection software resulted in the beam traces being truncated at 12 samples requiring a new Simrad software version to be installed. Because the sidescan data retain the full digital-sampling rate through each of the beams in turn, the across-track resolution is defined by the pulse width. This translates to about 5 to 12 cm across track. As with any conventional sidescan, the time sampling of the EM3000D results in poor spatial resolution near nadir but excellent across-track resolution in the far range. However, this characteristic, together with the much lower grazing angle in the far range, resulted in excellent target definition in the far range. By comparison, the EM3000S, which only looks out to about 65° (~2x water depth to the side, compared to ~4x water depth to the side for the EM3000D), has a much poorer target detection capability in the sidescan imagery.
All data processing was accomplished in the field. The raw data collected on day 1 were processed on day 2 so that any data-acquisition problems could be immediately corrected. The basic data-processing flow (Fig. 7) consisted of: (1) the editing the 1-Hz navigation fixes to flag bad fixes; (2) examining each ping of each beam to flag outliers, bad data, etc.; (3) merging the depth and backscatter data with the cleaned navigation; (4) reducing all depth values to mean low low water using observed NOAA tides recorded at the entrance to Los Angeles Harbor; (5) performing additional refraction corrections, if necessary, for correct beam raytracing; (6) separating out the amplitude measurements for conversion to backscatter; (7) gridding depth and backscatter into a geographic projection at the highest resolution possible with water depth; (8) regridding individual subareas of bathymetry and backscatter into final georeferenced map sheets; (9) gridding and contouring the bathymetry; and (10) generation of the final maps. Nearly finalized maps were completed in the field one day after the completion of the cruise.
Figure 7. Data processing steps used in Long Beach survey.
All bathymetric data were passed through a number of simple filters that (1) flagged all beams that form a spike when compared to their neighbors where the across-track slope anomalies exceed 25° and (2) flagged all beams that form a spike when compared to their neighbors where the along-track slope anomalies exceed 55°. These two criteria removed the worst of the spikes. All beams without immediate neighbors fore and aft were rejected and all beams without immediate neighbors port and starboard were rejected. The two criteria removed isolated beams but the remaining outliers had to be examined and subjectively flagged using the UNB/OMG SwathEditor tool in a line by line mode (Figure 5 is an example of a screen of SwathEditor).
Sidescan strip compilation
Slant-range-corrected images were created for all EM3000D data using a 0.25-m pixel resolution in the across track direction. The along track inter-ping spacing was 50 cm to 2m, based on an 8-kt ship speed, and the pulse-repetition rate varied between 2 and 7 Hz.
All amplitude (sidescan) data were initially processed at 0.25-m resolution and then resampled to the appropriate resolution for the individual map sheets (see below). The individual strips of backscatter were automatically desampled to match the backscatter area map being mosaicked (see below). In general, the resolution of the mosaics was chosen to be about 5% of the local water depth (50 cm in < 10m depths, 1m > 20m depths, etc.). Although the beam spacing increases in the far range, the full-resolution sidescan retains across-track sample density that corresponds to the bandwidth of the signal that makes possible target detection of fine scale in the far range. All sidescan data were automatically stencilled into the mosaic based on horizontal range from nadir, thus maintaining an equidistant seam for all the area.
All of the accepted EM3000D bathymetric solutions were combined into a mean surface rectilinear grid. The grid-cell spacing was chosen to approximate 5% of the water depth and is identical to the backscatter mosaic scale. Because the footprint dimension of the sonar increases with the incidence angle, a Butterworth weighted filter was used whose dimension was dynamically adjusted according to the slant range from the sonar to the seafloor. In addition, because the confidence in the bathymetric solutions decreases with increasing incidence angle (because of motion uncertainty, refraction errors, and bottom-detection noise), the weight of the outermost beams was systematically reduced so that for the case of overlapping data, the mean surface is strongly biased toward the more near-nadir soundings. This weighting scheme results in a natural blending of the inter-swath regions.
Six-minute tides measurements from the NOAA Los Angeles Harbor tide gauge (station number 9410660) were downloaded via ftp (wlnet2.nos.noaa.gov) at the end of each day to correct each bathymetric sounding to the mean low low water datum.
A series of 710 area maps was generated to cover the entire region (Fig. 8). Each area map represents 1600 m by 1000 m on the ground and together they form a regular grid covering the entire continental shelf. Adjacent area map sheets overlap one another by 0.5 minutes of latitude and 1.0 minutes of longitude. The area-sheet resolution varies from 50 cm in water depths less than 10 m to 2 m for water depths greater than 40 m. The map-sheet dimensions were chosen to reflect ~5% or less of the regional water depth to retain the highest justifiable resolution that the sonar could resolve. Groups of adjacent area map sheets were compiled into subregion maps at 4-m pixel resolution (Fig. 9). The subregion maps are titled.
All maps in this report are in Mercator projection, using 33° latitude for the projected latitude and the WGS84 spheroid. The entire area is within UTM zone 11. The maps of backscatter and shaded relief that accompany this report were generated from the highest resolution permitted by the average depth of each area map. The regridding reduces resolution in the shallower areas but allows the entire area to be mapped at a constant grid size.
Figure 8. Area maps generated for the Long Beach mapping survey.
Figure 9. Outlines of subregion map sheets. Labels are NPH (Newport Harbor), Hunt
(Huntington Beach), LAH (San Pedro Bay), GAB (San Gabriel), and PVE (Palos Verdes east).
Contour maps represent the traditional method of displaying bathymetry. For this project, the contours were derived from the gridded, tide-corrected bathymetry. The resultant contours were smoothed by a 3-point running average for the overview map. However, even at the original contour grid, more than 90% of the data must be discarded to only show some chosen contour interval. There are no data values between the contours so the viewer must imagine what might be there. A much easier way to view and comprehend bathymetry, using 100% of the data, is a shaded-relief map.
Shaded-relief maps represent a non-quantitative interpretation of the surface but use 100% of the bathymetric data. A shaded-relief map (Fig. 10) is a pseudo-sun-illumination of a topographic surface using the Lambertian scattering law (equation 1), where B is the pseudo-sun brightness, I is the maximum brightness, and F is the angle between the pseudo sun and a normal to the bathymetric surface.
If you overlay a shaded-relief map with contours, then you get both the complete image of the surface and selected quantitative contours.
The backscatter map is a representation of the amount of acoustic energy, at 300 kHz, that is scattered back to the hull-mounted receiver. Backscatter can be thought of as albedo; that is, the actual reflectance of the seafloor to 300-kHz sound. The Simrad EM3000D transducers have been calibrated to an rms pressure referenced to 1 mPa at 1 m from the transmitter. All gains, etc. applied during signal generation and detection are recorded for each beam and removed from the backscatter amplitude. Consequently, the backscatter is calibrated to an absolute reflectance of the seafloor, measured in decibels (equation 2), where I1 is the measured backscattered amplitude and I2 is the reference pressure of 1.
However, the amount of energy is some complex function of constructive and destructive interference caused by the interaction of an acoustic wave with a volume of sediment (Gardner et al., 1991) or, in the case of hard rock, the seafloor. The backscatter recorded by the EM3000D from a sedimented area represents volume reverberation to at least 1-m subbottom depth caused by seabed and subsurface interface roughness above the Rayleigh criteria (a function of acoustic wave length), volume inhomogenieties larger than about half the wavelength (50 cm), the composition of the sediment, and its bulk properties (water content, bulk density, sound velocity, etc.). Although, it is not yet possible to determine a unique geological facies from the backscatter value, reasonable predictions can be made based on the local geology.
It cannot be stressed too strongly that one of the great advantages of this survey is that the bathymetry is completely georeferenced with the backscatter. That means that each pixel on the map has a latitude, longitude, depth, and backscatter value assigned to it.
Figure 10. Map of the area surveyed during this cruise (area between land and break in slope
from the flat continental shelf to the incised continental margin.).
Daily Logs (all times in local time, GMT-7 hr)
April 08, 1999
The RV Coastal Surveyor (Fig. 11) and the C&C team arrived Seal Beach, CA. from Lafayette, LA at mid morning. The boat as assembled and in the water by the late afternoon.
Figure 11. RV Coastal Surveyor. The multibeam transducers are submerged on either side
of the bow ram (A). One DGPS antenna is located directly above the bow ram at B and two
DGPS antennas are located above the cabin directly below the letter B'.
April 10, 1999 (JD 100)
The first team of processors (JVG, JEHC, and Hou) arrived in Seal Beach, CA and set up the processing computers and plotter in the field headquarters. No problems were encountered. The processing equipment consisted of one SGI R10000 Indigo2, one SGI R4000 Indigo2, one Apple Macintosh G3, and one Apple Macintosh G3 PowerBook, all networked via ethernet and sharing 100 gbytes of available hard-disk space. In addition, a 1200-dpi HP 4000TN laser printer and a 600-dpi 36" HP 2500CP color plotter were also on the network. All data were archived on magneto-optical and DLT disks medium.
April 11, 1999 (JD 101)
A major rainstorm hit during the night and continued all day. Work was conducted on the boat at dockside to calibrate the POS/MV vehicle motion sensor and worked out ethernet problems.
April 12, 1999 (JD 102)
The first patch test run in mid-morning. Problems immediately cropped up with the new POS/MV and with sampling of sidescan records. The Simrad recording system is truncating the outer beams to only 12 samples/beam, which is adequate for the inner beams but is undersampling the outer beams. The result is that there are significant gaps in the sidescan record in the far range. Simrad was contacted overnight for a solution to the sidescan problems. An Applied Analytic engineer came to the boat and spent the day working with the POS/MV. The problems with the motion sensor appeared to be a formatting issue with the data string. The new POS/MV was replaced by the spare POS/MV. An additional problem was that the C&C technicians had mounted the starboard transducer 180° from the Simrad recommendation so that the two Mills Crosses would be in mirror image to one another. However, Simrad has software flags to account for the asymmetry so that the improper mounting was working against the software. This problem was corrected by remounting the starboard transducer in its recommended position.
April 13, 1999 (JD 103)
A second patch test was run and both the POS/MV and the Simrad software appeared not to be properly compensating for the vehicle motion and for adequately sampling the outer beams amplitude data. Everything was put on hold until new software was received from Simrad.
April 14, 1999 (JD 104)
The latest version of the Simrad software was received and loaded into the system. The system was rebooted and it appeared the software upgrade eliminated the problem we had with the backscatter and the vehicle-motion compensation. The third patch test began at 0930 hr in the morning and continued until 1900 hr. Preliminary processing of the data was accomplished during the patch test and the results were satisfactory. Thorough processing of the patch test was performed in the evening and all systems were performing up to spec.
April 15, 1999 (JD 105)
The survey began at 0600 hr with a transit south to the Newport area and the mapping of the narrow Newport shelf. Although there were some minor problems with the vehicle motion sensor (POS/MV) during the transit from Seal Beach harbor to the Newport area, the systems were all properly performing by the time the boat was on the first line. The surveying was routine and the sea state was calm.
Suddenly, at about 1830 hr, all the power went off in Seal Beach. Apparently, there was an underground explosion in one of the nearby underground power conduits. Although the processing computers were connected to two UPSs, the smaller UPS failed to supply power to the second SGI and it crashed. The power was down for about 1 hour.
April 16, 1999 (JD 106)
Routine day of mapping the narrow shelf south of Newpor. The effects of sewage discharge changing the salinity and temperature of the water column began to cause refraction problems. Although the processing tools can correct for the effects in the sonar data, the discharge causes a careful evaluation of the water velocity measured at the transducers.
April 17, 1999 (JD 107)
Routine day of mapping the outer shelf off Newport. Another patch test was run mid day and the statistics showed new offsets might improve the slight motion error we were seeing in the first days of data. The new offsets are:
|vertical offset||0.8||starboard head|
There continues to be problems with the POS/MV locking up. Because the inertial navigator of the POS/MV is the primary navigation sensor, the lock ups essentially turns off navigation. This leaves gaps of a few seconds. Our processing software detects the navigation gaps and performs a linear interpolation between the last good position before the lockup and the first new location after navigation is reacquired.
April 18, 1999 (JD 108)
Routine day of mapping to complete the narrow shelf south of Newport and begin mapping the inner shelf directly west of Newport. We observed a semiconsistant offset between the port and starboard beams in the zone of overlap. The offset is on the order of 25 to 50 cm. The C&C crew aboard Coastal Surveyor also noticed the offset and, after ending line 97, ran an additional patch test and came up with a new set of offsets. They commenced using the new offsets on line 110 and beyond. They thought the new offsets improved the data overlap. However, because we were processing about 24 hours after the data collection, an entire day was run using the new offsets before we could see the effects. It appeared during the processing that the new offsets made the overlap worse, so we called the boat and had them revert back to the original offsets at the beginning of line 120.
April 19, 1999 (JD 109)
The C&C crew decided to continue surveying for 48 hours without a break. They commenced surveying but had some engine problems that cost about 4 hours. Other than that, it was a routine day of mapping in the outer shelf area off Newport. Occasional POS/MV lockups continue to occur. Also, rapid changes in water properties require several CTD casts each day.
April 20, 1999 (JD 110)
Mapping continued throughout the night and day and completed the Newport Harbor margin part of the survey. The remainder of the day was spent mapping the mid shelf south of Huntington Beach. All surveying was routine. During the processing, an error in the predicted tides was found by comparing the predicted tides to the observed NOAA tides (Fig. 12). This required contacting NOAA to acquire their tide data immediately after the data were recorded at the tide gauge. NOAA provided edited observed tides with a 3-hour delay on their server (telnet) wlnet2.nos.noaa.gov. The tide gauge is at the entrance to Los Angeles Harbor (station 9410660), essentially Long Beach Harbor. The tide delay in time between Long Beach and Newport Beach varied between 1 minute for high water and 9 minutes for low water. The tide correction between Long Beach and Newport Beach is 0.96 times the tidal height. Consequently, we used the Los Angeles harbor observed tides for all tide corrections.
Figure 12. Plot of predicted tides (solid line) versus observed tides (dashed line) at Los Angeles
Harbor entrance. Observed tide from NOAA data on http://wlnet2.nos.noaa.gov. Predicted tide
from XTIDE program.
April 21, 1999 (JD 111)
Routine mapping on the mid shelf south of Huntington Beach. The crew arrived at the dock at about 1200 L for a break in surveying. The day was spent regridding and remosaicking the mapsheets to correct for the new tides and various minor processing glitches. An apparent dumping ground was mapped on Line 161 between San Gabriel and Newport Canyons in about 46-m depth (Fig. 13). The high-backscatter features are about 9.5 m long, suggesting they may be discarded drill pipe.
Figure 13. Backscatter from Line 161 on outer continental shelf between San Gabriel and
Newport canyons at 33° 36.333414' N, 117° 57.783319' W in 46 m water depth.
April 22, 1999 (JD 112)
The C&C crew took the day off because they surveyed continuously for the last 48 hours. Processing continued to catch up with the 48 hours of continuous data collection. It became immediately obvious that the new offsets were performing much better than the original ones, which will require some post-cruise reprocessing to correct the first four days of data. Refraction because of changes in water properties (probably both temperature and salinity) alter the sound speed in water over very short distances. This is apparently the result of the numerous sewage outfall pipes on the shelf. Although our processing refraction tool can compensate for this effect, it has required the shipboard crew to stop mapping and run a CTD cast several times a day. Tianhang Hou left and Jose Martinez arrived.
April 23, 1999 (JD 113)
The boat departed at 0530 hr to begin surveying immediately south of Palos Verdes Peninsula because of an announced sailing regatta planned for today and the weekend. However, during the transit to the operational area the engine began to overheat. The engine had to be turned off so they threw out the anchor. Nevertheless, the Coastal Surveyor drifted up against the Long Beach Harbor seawall. The Coast Guard was call and the boat was towed back to port. A mechanic was called and the rest of the day was spent fixing the engine. The temperature gauge was determined ruined and it will take 2 days to get a replacement. The Captain determined that the engine temperature warning light was sufficient to make it safe to continue operations the next day. It was the best day for the engine to fail because the wind was brisk and the swell was from the southwest, making surveying difficult. Processing continued unabated.
Fields of high-backscatter bedforms are found all along the inner shelf in depths between 5 and 10 m. The bedforms have wavelengths of less than 1 m and wave heights of less than 50 cm (Fig. 14).
Figure 14. High-backscatter, small-scale bedforms
found on Line 141 in 6-m water depths in the area
between Newport Beach and Huntington Beach.
April 24, 1999 (JD 114)
Routine day of mapping the mid shelf south of Huntington Beach. No problems occurred with the boat engine overheating. All the map sheets were cleaned and rerun for new grids and new mosaics to ensure uniform treatment. Line 108 did not get written to the raw data and Line 219 had to be broken into two pieces at record 2568.
April 25, 1999 (JD 115)
Routine day of mapping the head of San Gabriel Canyon. Small motion artifacts occur throughout the data. The artifacts are on the order of 5 cm in amplitude but they show up in the shaded-relief images. More troubling is the spacing of the outer beams, especially when the beams are looking downslope. There is a noticeable separation of the beam footprints in these outer beams that, if not overlapped by adjacent swaths, create a track-parallel artifact. Mark Paton arrived.
April 26, 1999 (JD 116)
Routine day of mapping the outer shelf west of San Gabriel Canyon. The survey was terminated at mid afternoon so that the boat could be cleaned up for Media Day. An experiment was run on the data using a median filter on the raw data rather than using the standard manual swath editing (swathed). The median filter will pick the median depth value within each cell, regardless of its location within the cell. This technique effectively eliminates the outliers, especially in cells with lots of samples. Because the EM3000 provides such a large data density within cells, and because the Butterworth filter we use is constructed with a 3-cell taper, almost all of our data can be processed with the median-filter technique. This speeds up the processing and allows us to devote more time to the map products. The experiments continued throughout the day.
April 27, 1999 (JD 117)
Today was Media Day. Several reporters from the local newspapers and TV crews attended the briefing, as well as local government officials. The briefings were over by noon but, because of stiff winds and swell, no surveys were run. The processing experiments with the median-filter technique continued.
April 28, 1999 (JD 118)
The boat departed for surveying the outer shelf between San Gabriel and Newport Canyons but by 1000 L the swell was 10 ft and the wind had freshened and marine advisories were posted. The sea state was outside the limit of acceptable data collection and the boat handling became increasingly difficult. The Captain decided to terminate the survey for today. It was a good decision because once the Simrad transducers are elevated into the surface zone of bubbles, the data is degraded to the point of uselessness. The processing experiments with the median-filter technique continued. Larry Mayer and Mark Paton left.
April 29, 1999 (JD 119)
The seas and swell continued to be too large for operations. The downtime allowed us to catch up on the data processing.
April 30, 1999 (JD 120)
The storm front passed through the area during the night and by morning the seas and swell were down. Routine day of mapping the outer shelf between San Pedro Sea Valley and San Gabriel Canyon.
May 1, 1999 (JD 121)
Routine day of mapping along the outer shelf in the vicinity of San Pedro Sea Valley. High-backscatter outcrops occur along the outer shelf.
May 2, 1999 (JD 122)
The boat departed at 0300 L to escape all the pleasure boaters during the weekend. Mapping continued off the eastern Palos Verdes Peninsula. Toward the end of Line 392 it appears that something got wrapped around the starboard transducer. The data from the starboard transducer had to be completely edited out. Kelp that was ripped up during the strong swell and sea during the past three days was seen floating in masses all along this area. Just prior to completing the eastern Palos Verdes area the boat sucked something into the water intake causing the engine to overhead. The boat limped back to the dock at 1430 L. The Captain bought a wet suit and repeatedly dove down to pull out a large wad of plastic jammed in the intake. The NOAA tide server (wlnet2.nos.noaa.gov) crashed so we were unable to get tides for JD 120 and 121. This delayed our gridding and mosaicking. Jose Martinez left but Miguel Pacheco, a Portuguese Naval Officer and graduate student at UNB who was due to arrive late this evening, was stopped at the US Immigrations Office, Toronto Airport because he lacked the proper visa to enter the USA.
May 3, 1999 (JD 123)
Mapping continued to complete the Palos Verdes East section but by noon the winds were very strong and the swell was 5 to 8 ft. The strong swell together with the local chop created by the strong winds reduced the data quality to unacceptable so the day was terminated.
May 4, 1999 (JD 124)
Routine day of mapping along the continental shelf between Palos Verdes Peninsula and San Gabriel Canyon. The POS/MV continued to randomly shut down causing premature end of lines. Weather was good and everything appeared normal. Out crops are apparent all along the shelf at ~35 m water depth.
May 5, 1999 (JD 125)
Routine day of mapping in the area west of San Gabriel Canyon. The POS/MV continued to shut down so a new one was ordered from C&C. The problem was isolated to the power supply and a new one will take 2 days to get to us. The CTD began to give problems and a new one was flown to us and arrived late this evening.
May 6, 1999 (JD 126)
Routine day of mapping in the area west of San Gabriel Canyon. An induced pitch/heave artifact continued to be apparent throughout the data because of a 6 to 8 ft swell that quartered the track lines has accentuated the effect. We suspect the induced pitch/heave artifact is one of the results of the POS/MV problems. The artifact will require post-cruise processing to eliminate. We reached the 50-gbyte level of processed data today, requiring a massive backup of files. The magneto-optical drive on the processing computer (steelhead) failed during a write cycle and is no longer able to be used as the primary backup medium. The DLT tape is now the primary backup. This will cause a major headache back in the lab because the tape is sequential, rather than the random access of the magneto-optical disk, causing long search times (3 to 4 hr) for even a single file.
May 7, 1999 (JD 127)
Routine day of mapping in the shelf west and north of San Gabriel Canyon. The POS/MV continued to hang up causing loss of vehicle motion corrections. The new POS/MV computer controller and power supply arrived late in the afternoon.
May 8, 1999 (JD 128)
Routine day of mapping in the shelf west and north of San Gabriel Canyon. The new POS/MV controller and power supply were installed but the new configuration did not allow the heading calibration of run to completion. The old controller and power supply were put back into the system and the random hang-ups continued throughout the day.
May 9, 1999 (JD 129)
Mapping continued throughout the day in the area immediately south of the San Pedro Harbor seawall. The boat's autopilot failed at mid day requiring the boat to be steered by hand throughout the remainder of the day. A large, unidentified, man-made target was mapped (Fig. XX). The data processing continued to have problems of data volume. The lack of a workable magneto-optical drive meant that when files had to be retrieved from the backup DLT tape because of a processing error, more than 6 hours were wasted waiting for the DLT file retrieval.
May 10, 1999 (JD 130)
Mapping continued in the area south of San Pedro Harbor. The replacement autopilot part arrived and the autopilot was repaired. The new POS/MV controller board and power supply was calibrated and appeared to be working. Two hours were used at the end of the day to perform a POS/MV test and all appeared back to normal.
May 11, 1999 (JD 131)
Routine mapping in 20-m water depths directly south of San Pedro Harbor. Results from data collected with the repaired POS/MV show the roll/pitch artifact is still strong.
May 12, 1999 (JD 132)
Routine mapping in 20-m water depths directly south of San Pedro Harbor. A dredge site for the approaches to San Pedro Harbor was surveyed (Fig. 15).
Figure 15. Shaded-relief image of a 1997 (?) dredged area to deepen the approaches to
San Pedro Harbor. The illumination is from 310°. The center of the dredged area is
33° 40.0030'N 118° 13.6890'W. The depression is 0.75 to 1.1 m lower than the
May 13, 1999 (JD 133)
Routine day of mapping in the area directly south of San Pedro Harbor.
May 14, 1999 (JD 134)
Routine day of mapping in the area directly south of Huntington Beach. The winds became very stiff in the mid afternoon, with gusts approaching 40 kts. The sea conditions finally exceeded the limit to collect good data by 1530 hr so the mapping was suspended for the day.
May 15, 1999 (JD 135)
The day was spent rerunning line 990511/583 with an EG&G model 272-TD 100-kHz sidescan sonar (Fig. 16). The purpose was to perform a definitive test to compare the multibeam-backscatter map of a small area with a similar-frequency sidescan sonograph of the same areas. The line started at 33° 39.87759'N 118° 04.223642'W and ended at 33° 38.806138'N 118° 00.297223'W. The sidescan sonar was acquired with a Codex acquisition system and the only recording medium was a 1024 byte/sector magneto-optical disk and the only format was "SEGY". The data had to be transcribed into another format on another medium before it could be read. In addition, a Seatex vehicle motion sensor was set up to compare with the POS/MV sensor that was used for the entire survey. No mapping was accomplished this day. It must be remembered that the sidescan data are not georeferenced nor are the data adjusted for across-track bathymetry. The sidescan data can be roughly georeferenced to the multibeam data but they can not be corrected for the flat-bottom-at-nadir assumption.
Figure 16. EG&G 272-TD 100/500-kHz sidescan-sonar fish used for comparison test.
May 16, 1999 (JD 136)
Routine day of mapping in area directly south of Seal Beach. By the end of the day virtually all of the area shallower than 20 m has been mapped.
May 17, 1999 (JD 137)
The day got off to a late start because the fuel dock did not open when scheduled. To make up for the lost time, the boat stayed out until 2000 hr. Routine day of mapping in area directly south of the San Pedro breakwater. The NOAA tide server crashed so that tides for May 16 and 17 were not available. Consequently, the data for these days could not be gridded or mosaicked.
May 18, 1999 (JD 138)
Routine day of mapping to complete the entire survey. The day was spent mapping the shallow area directly south of the San Pedro Breakwater.
May 19, 1999 (JD 139)
Entire day was spent processing the data from May 18th. All the processed files were backed up overnight and the computer systems were packed up for transport back to Menlo Park.
James V. Gardner, Chief Scientist USGS
John E. Hughes Clarke OMG/UNB
Larry A. Mayer OMG/UNB
Tianhang Hou OMG/UNB
Sean Galway OMG/UNB
Jose Martinez OMG/UNB
Luciano Fonseca OMG/UNB
Tony Hewett OMG/UNB
Mark Paton IVS
James Chance C&C
Art A. Kleiner C&C
Ryan Larsen, Party Chief C&C
Scott McMay C&C
Page Malancon C&C
Ricky Bushinell, Captian C&C
Gardner, J.V., Field, M.E., Lee. H., Edwards, B.E., Masson, D.G., Kenyon, N., and Kidd, R.B., 1991 Ground truthing 6.5-kHz sidescan sonographs: What are we really imaging?. J. Geophys. Res., v.96, p. 5955-5974.
Gardner, J.V., Mayer, L.A., and Hughes Clarke, J.E., 1996, Cruise Report of cruise C1-96-SC, mapping the Santa Monica, CA continental margin, unpublished USGS report.
Gardner, J.V., Mayer, L.A., and Hughes Clarke, J.E., 1998, Cruise Report, RV Ocean Alert Cruise A2-98-SC, Mapping the southern Califonria continental margin. U.S. Geol. Survey Open-File Rept. 98-479, 25p.
Hughes-Clarke, J.E., Mayer, L.A., and Wells, D.E., 1996, Shallow-water imaging multibeam somars: A new tool for investigating seafloor processes in the coastal zone and on the continental shelf. Marine Geophysical Researches, 18: 607-629.
For more information, contact Pete Dartnell.