United States
Environmental Protection
Agency
Air And Radiation
(ANR-461)
EPA520/1-91-009
June 1991
f/EPA
Ocean Current Measurements
At The Farallon Islands
Low-Level Radioactive Waste
Disposal Site 1977 -1978
-------
EPA 520/1-91-009
Ocean Current Measurements
at the
Farallon Islands
Low-Level Radioactive Waste Disposal Site
1977 - 1978
Prepared under
Contract 68-01-0796
Project Officer
Robert S. Dyer
Office of Radiation Programs
U.S. Environmental Protection Agency
Washington, D.C.
Printed on Recycled Paper
-------
FOREWORD
Pursuant to Public Law 92-532 (the Marine Protection, Research and
Sanctuaries Act of 1972), the U.S. Environmental Protection Agency (EPA) has
developed criteria and regulations to govern ocean disposal of all forms of waste,
including low-level radioactive waste (LLW) materials.
In 1974, the EPA Office of Radiation Programs (ORP) initiated feasibility
studies to determine whether existing technologies could be applied toward assessing the
fate of radioactive wastes that had previously been disposed in the oceans. The ORP
developed a program for site characterization studies to determine the biological,
chemical, geological and physical characteristics of the marine environment in and near
sites that had been designated by the former Atomic Energy Commission (AEC) for
ocean disposal of LLW. These studies also included investigating the presence and
distribution of radionuclides within these sites.
A primary mechanism for physically dispersing and redistributing both
soluble and particulate radioactive materials from a disposal site is the action of ocean
bottom currents. Of particular interest is the magnitude and direction of these currents.
This report discusses the results of ocean bottom current measurements obtained from
the Farallon Islands LLW disposal site off the California coast, near San Francisco. The
report includes a discussion of the velocity of the currents over the time period and area
measured relative to large-scale currents off the California coast, and the possibility for
shoreward transport of LLW materials from the Farallon Islands site.
The Agency invites all readers of this report to send any comments or
suggestions to Mr. Martin P. Halper, Director, Analysis and Support Division, Office of
Radiation Programs (ANR-461), U.S. Environmental Protection Agency, Washington, DC
20460.
T. Oge, ActmgyDirector
of Radiatio£jPrograms
111
-------
CONTENTS
Section Pagt
1. INTRODUCTION 1
2. SUMMARY 3
2.1 DESCRIPTION OF STUDY 3
2.2 RESULTS AND CONCLUSIONS 3
3. MEASUREMENT PROGRAM 7
3.1 OPERATION AND OBJECTIVES 7
3.2 DESCRIPTION OF HARDWARE 8
3.3 CURRENT METER ARRAY LOCATION AND OPERATION 9
3.4 BATHYMETRY 15
3.5 SEDIMENT INFORMATION 15
4. DATA REDUCTION 18
4.1 CURRENT METER DATA 18
4.2 GENERATION OF BATHYMETRIC CHART 18
4.3 GRAIN SIZE FROM SEDIMENT SAMPLES 19
5. CURRENT METER DATA PROCESSING 21
5.1 DATA PRODUCTS 21
5.2 VACM AND AANDERAA COMPARISON 35
5.3 MEASUREMENT ACCURACY AND TIMING ERRORS 41
5.4 CONDUCTIVITY AND SALINITY 45
6. DATA INTERPRETATION 46
6.1 ANALYSIS OF CURRENTS 46
6.2 TRANSPORT POTENTIAL 62
7. RECOMMENDATIONS FOR FURTHER WORK 68
8. REFERENCES 71
APPENDICES [Contained In A Separate Report - EPA 520/1-91-009/A]
TIME HISTORY - SPEED AND DIRECTION A-l
TIME HISTORY PLOTS - NORTH, EAST CURRENTS B-l
HISTOGRAMS C-l
SCATTERGRAMS D-l
SPECTRA - NORTH, EAST COMPONENTS E-l
PROGRESSIVE VECTOR DIAGRAMS F-l
STICK PLOTS G-l
SUMMARY STATISTICS H-l
-------
LIST OF FIGURES
Number Page
1-1 Locator Chart 2
2-1 Current Meter Arrays for Radioactive Waste Disposal Site 4
3-1 Current Meter Locations 11
3-2 Data Coverage Plot 1977-1978 12
3-3 Profile Schematic 13
3-4 Current Meter Array Plan Schematic 14
3-5 Locations of Core Sample Sites 16
3-6 Depth Profile of Core Sample Sites 17
4-1 Bathymetry Chart 20
5-1 Time History Plot, Current Speed, and Direction (Meter 2920) 22
5-2 Time History Plot, North Current, East Current (Meter 2920) 24
5-3 Histogram Current Speed (Meter 2920) 26
5-4 Speed Versus Direction Scattergram (Meter 2920) 28
5-5 Power Spectrum (Meter 2920; February 1978) 29
5-6 Progressive Vector Diagram (Meter 2919) 31
5-7 Stick Plot (Meter 2918) 33
5-8 Summary Statistics (Meter 2919) 34
5-9 VACM Reconstruction 1 37
5-10 VACM Reconstruction 2 38
5-11 VACM Reconstruction 3 39
5-12 VACM Reconstruction 4 40
5-13 Speed Gain Transfer Function for Meter 2830 Versus VACM 42
5-14 Speed Phase Transfer Function for Meter 2830 Versus VACM 43
6-1 Power Spectrum for Meter 2918 50
6-2 Power Spectrum for Meter 2919 51
6-3 Power Spectrum for Meter 2830 52
6-4 Power Spectrum for VACM 53
6-5 Power Spectrum for Meter 2920 54
6-6 Tidal Ellipse Calculation 55
6-7 Tidal Ellipse for Meter 2918 56
6-8 Tidal Ellipse for Meter 2919 57
6-9 Tidal Ellipse for Meter 2830 58
6-10 Tidal Ellipse for VACM 59
6-11 Tidal Ellipse for Meter 2920 60
6-12 Exceedence Curves for Total Measurement Periods 66
6-13 Exceedence Curves for October through December 1977 67
vii
-------
LIST OF TABLES
Number Page
3-1 Farallon Islands Current Meter Locations 9
3-2 Data Record 10
3-3 Box Cores Collected for Sediment Studies During
R/V VELERO IV Cruise (August-September 1977) 15
5-1 Estimated Clock Drift 44
6-1 Major Current Energy Sources for Entire Operating
Period of Each Meter 49
6-2 Vector-averaged Current Velocities 61
6-3 Sediment Surface Grain Size Data from Farallon
Islands LLW Site Survey Cores 63
6-4 Percent of Total Time Measurement Speeds Exceed 20 cm/s 65
IX
-------
1. INTRODUCTION
This report describes the work performed by Interstate Electronics Corporation
(IEC) under contract to the U.S. Environmental Protection Agency's (EPA) Office of
Radiation Programs (ORP). It presents data from an ocean current measurement study,
conducted during 1977 and 1978, in the area of the U.S. low-level radioactive waste (LLW)
disposal site near the Farallon Islands (Figure 1-1) off the coast of San Francisco,
California. It also contains other oceanographic and environmental data acquired from the
same area.
The purpose of this study was to measure near-bottom and bottom currents in the
area, and utilize available historical data, to determine the potential for transport of LLW
from the disposal site toward populated areas in the vicinity of San Francisco. The
remainder of this report is sub-divided into sections, as follows:
Section 2 describes the study and summarizes results and
conclusions;
Section 3 discusses the measurement program that generated
the data analyzed in this report;
Section 4 describes the steps taken to improve the quality of
data from two of the current meters, and gives an assessment
of the overall quality and temporal coverage of the current
meter data;
Section 5 discusses data processing and verification;
Section 6 contains an interpretation of the current meter data
combined with other available data taken at the site during
previous studies conducted in 1974 and 1975, including an
analysis of the current, sediment, and bathymetric data and of
transport mechanisms related to the data;
Section 7 contains recommendations for future sampling;
Section 8 lists the references utilized during the study; and,
the Appendices contain computer generated graphical displays
of the current meters' output data.
-------
D DUMP SITES
O STATIONS SAMPLED FOR WATER, SEDIMENT, OR BIOTA
O CURRENT METER ARRAYS
Figure 1-1. Locator Chart.
Dashed line indicates area defined [11 as the Waste Disposal Area; squares
A, B, and C are disposal sites. Circled numbers 1 through 10 are the locations of
January 1977 sampling stations [2]. Hexagons B, C, and D indicate the locations
of current meter arrays during this (1977-1978) study.
-------
2. SUMMARY
This section briefly summarizes the major events of this current measurement
program, the subsequent analysis of data, and the conclusions obtained. The analysis of
data included the following five tasks: (1) assessing data quality, (2) comparing data from
the different types of current meters used, (3) determining relative clock drift and timing
errors, (4) evaluating the measured current components, and (5) estimating potential for
sediment transport of LLW from the disposal site. The conclusions derived from each task
are given in Sub-Section 2.2., below. Examples to support these conclusions are discussed
in the appropriate portions of Sections 3 through 7 of this report.
2.1 DESCRIPTION OF STUDY
In October 1977 four mooring line arrays containing seven current meters were
deployed in the vicinity of the LLW disposal site near the Farallon Islands (Figure 2-1).
The instruments on the arrays included six Aanderaa model 5 Recording Current Meters
(RCM), and one Sea-Link Vector-Averaging Current Meter (VACM). All were pre-set for
an operational period of approximately 90 days, but actual deployment lasted for about a
year.
During recovery, in October 1978, it was observed that many of the meters were
capable of continuous data recording for a longer period of time than was expected (Figure
3-2). This is especially true for the VACM, which continued to record data for the entire
year.
It was also observed during recovery that the transmitter on the acoustic release
mechanism for Array A did not transmit when interrogated. When the release mechanism
was activated the array could not be located. Thus, two of the RCM 5 meters (numbers
2916 and 2917) were lost.
2.2 RESULTS AND CONCLUSIONS
2.2.1 Data Quality
Data was recovered from five current meters-four Aanderaas and one VACM. The
VACM recorded data for 1 year, while the Aanderaas recorded for a period of 2 to 4Vi
months each. After recovery of the meters and subsequent initial inspection, it was
observed that the VACM was still recording data, while two of the Aanderaa meters
(numbers 2919 and 2920) had filled their output tapes. The other two Aanderaa meters
(numbers 2830 and 2918) had experienced battery failures during their operation. Details
regarding the operation and performance of the hardware can be found in Sections 3.1 and
3.3.
-------
TO SURFACE
ARRAY B
ARRAY D
ARRAY C
ARRAY A
- SUBSURFACE BUOYS
A ANCHOR/ACOUSTIC RELEASE
• AANDERM CURRENT METER
• VECTOR-AVERAGING CURRENT METER (VACM)
(NOT DRAWN TO SCALE)
Figure 2-1. Current Meter Arrays for Radioactive Waste Disposal Site
4
-------
All of the data recovered from the VACM and two of the Aanderaa meters (numbers
2918 and 2919) was of a quality suitable for analysis and interpretation. The data from the
other two Aanderaa meters (numbers 2830 and 2920) had to be recovered by a special
technique, as described in Section 4.1, but after this additional processing it was also found
to be of good quality. There were a few isolated data points in some of the records for each
of the meters which had obvious nonphysical values due to either electronic noise or
mechanical problems, but these were replaced by interpolated values yielding a continuous
record of high quality. The data was then translated to metric engineering units for
analyses and generation of data products. See Section 4.1 for further discussion.
The statistical properties of the recovered data were generally in accordance with
expected physical processes, although it became evident that the VACM data record
contained a number of transpositions in data sequence. In addition, Aanderaa # 2830
apparently contained gaps in its recording. Approximately 9 hours of data were missing
from the beginning of the record commencing October 26,1977. Further discussion appears
in Section 5.2.
2.2.2 Comparison of the VACM and Aanderaa Meters
After taking into account the observed inconsistencies in the VACM data sequence,
the corresponding VACM and Aanderaa meter # 2830 records were compared by examining
the time series and obtaining the speed transfer function. The data records were found to
be a relatively close match. The speed transfer function was relatively flat and fell within
a range of 0.8 to 1.05, with the VACM giving a generally larger magnitude than observed
on meter # 2830. The phase portion of the transfer function was linear, indicating no
substantial timing rate shifts between the two meters (see Section 5.2).
2.2.3 Relative Clock Drift
The semidiurnal tidal occurrences were examined to determine the degree of relative
clock drift for the 5 meters. Analyses in both the time and frequency domains indicated
that although clock drift appeared to be of a larger magnitude than advertised by the meter
manufacturers, relative timing errors were within a range suitable for cross-analysis of all
five current meter data records. Section 5.3 addresses the details of this analysis.
2.2.4 Evaluation of Measured Current Components
For each meter the short-term periodic and long-term average current components
were examined (see Section 6.1) with the following results:
a. In the deep western part of the dump site, most of the current energy is from
12-hour tides in the east-west direction;
-------
b. In the deep eastern part of the dump site, most of the current energy is from
12-hour tides with a slightly greater component of energy in the north-south
direction;
c. The total current energy is significantly greater in the deep western portion
of the site than in the deep eastern portion;
d. The VACM and Aanderaa meter # 2830 in the open ocean (midwater) show
a relatively equal distribution of tidal energy and also have a significant low
frequency component in the north-south direction; and,
e. Long-term currents average less than 2 cm/s and move in the northeast
direction for the bottom measurements (Aanderaa meters 2918, 2919, and
2920) and the southeast direction for the deep midwater (VACM and
Aanderaa # 2830) measurements.
2.2.5 Estimation of Sediment Transport Potential
Analysis of sediment distributions has indicated that fine-grain sediments composed
the majority of sediment volume ^. Since these sediments can occasionally be suspended
when water velocities exceed about 20 cm/s, an estimate of transport potential was obtained
by determining the extent to which these velocities occur. Average currents were then used
to determine the probable distance and direction of transport (see Section 6.2). The general
conclusions are as follows:
a. In the deep western part of the site, current measurements exceed 20 cm/s no
more than 3 percent of the time. This may be sufficient to suspend fine-grain
sediments (silt and clay), providing a potential for transport in the water
column.
b. Long-term average near-bottom currents move north and eastward throughout
the site. Average current speeds diminish from 1.7 cm/s at the deep western
end of the site to 0.17 cm/s at the eastern end, thus it appears that this vector
decreases with proximity to the shore.
c. Results, though inconclusive, indicate that there may be reason to further
investigate the extent and scale of transport as it relates to contaminated
sediment.
-------
3. MEASUREMENT PROGRAM
On October 25, 1977 four arrays containing six Aanderaa current meters and one
VACM were installed in an area of the radioactive waste disposal site near the Farallon
Islands (Figure 1-1). This section discusses the procedures and operations used in the
deployment of the meters and the subsequent acquisition of data.
3.1 OPERATION AND OBJECTIVES
A total of seven current meters were to be emplaced using four arrays, supported
by subsurface buoys (Figure 2-1). One array (A) was lost, consequently no data were
recovered from Aanderaa meters 2916 and 2917. The remaining four Aanderaa meters
and the VACM functioned and recorded current velocity, temperature, and conductivity
measurements for between 2 months and 1 year for each meter, giving a cumulative
volume of data totaling 800 days.
The selection of the current measurement hardware was based on a number of
factors including (1) cost, (2) anticipated performance at the depths and durations
required for this program, and (3) known handling and reliability characteristics. The
Aanderaa RCM, model 5, was deemed most suitable in terms of these considerations.
The VACM was deployed immediately above one of the Aanderaa meters in order to
study the operational characteristics of the two meter types relative to each other.
The study was designed to provide preliminary information about the speed and
direction of water moving through the region of the disposal site. This information,
along with the results of previous physical and chemical studies of the region, was
intended to aid in the development of an understanding of the natural physical processes
and any subsequent impact related to potential or existing transport of radioactive
materials from the disposal area.
Deployment of the arrays (Figure 2-1) was planned to establish two approximately
triangular planar surfaces of measurement; one parallel to and within a few meters of
the bottom, (arrays A, B, and D, and the other approximately parallel to the 3,000-foot
depth level (arrays A, B, and C). In addition, the positioning the VACM and Aanderaa
meter # 2830 close together on the same string would allow for a direct comparison of
these two meter types in a midwater location, where the assumption could be made that
both meters were experiencing the same water mass dynamics.
-------
3.2 DESCRIPTION OF HARDWARE
The Sea-Link VACM records current speed and direction in cartesian coordinates,
average water temperatures, and timing information sampled at an adjustable interval
for a period as long as 1 year. Current speed is obtained from a savonius rotor, and
direction is measured relative to magnetic north by means of a neutrally buoyant
magnetic-coupled vane mounted above the rotor.
The VACM records eight current speed and direction measurements with each
turn of the rotor. The readings are divided into components, combined, and referenced
to time and temperature on a cassette tape recorder at preselected intervals. This
averaging is believed to filter out high frequencies caused by mooring motion. A 15-
minute interval was selected for this measurement program. The meter is powered by a
self-contained alkaline battery pack with a life expectancy of 9 months at an average
current speed of 150 cm/s. The operational ranges and accuracies for measured
parameters are:
Current speed: 2.5 to 300 cm/s
Current direction: 0* to 360' ±2.8*
Temperature: -2*C to 35*C ±O.10'C
Timing accuracy: 6 minutes/year
The Aanderaa RCM-5 current meter also uses a savonius rotor to record current
speed. The meter is mounted in front of a large vane so that the rotor is always pivoted
into the oncoming current. In addition to current speed and direction and water
temperature, the instrument may contain optional devices for measuring conductivity
and water depth. One of the meters used in this study, # 2920, recorded conductivity by
means of an inductively coupled toroidal coil. The measurement interval was preselected
at 20 minutes to allow for approximately 4 months of recording. Battery life is 3
months. Operational ranges and accuracies for measured parameters are:
Current speed: 1.5 to 250 cm/s
Current direction': 0' to 360* ±0.1*
Conductivity: 0.0 to 60 millimoho/cm
Timing accuracy: 12 minutes/year
'Aanderaa RCM's have demonstrated compass errors due to magnetization of
nickel plating in pressure cases. The meters deployed for this survey used new epoxy-
coated pressure cases with nickel only in the upper 0-ring region.
8
-------
3.3 CURRENT METER ARRAY LOCATION AND OPERATION
Table 3-1 and Figure 3-1 give the locations, site depths, and meter depths of each
current meter in the water column. Figure 3-1 also shows the local bathymetry in
relation to the current meters. The meters were deployed along a gully running roughly
east-west. The VACM and Aanderaa meter # 2830 were attached on the same array
anchored at the bottom in a depression, and the other Aanderaa meters were anchored
on a ridge just north of the gully with meters 2918 and 2919 attached on the same array.
Bottom topography increased from east to west
Table 3-1. Farallon Islands Current Meter Locations
Meter
(Array)
Aanderaa
2920 (B)
Aanderaa
2830 (C)
VACM (C)
Aanderaa
2918 (D)
Aanderaa
2919 (D)
Start
10/25/77
10/25/77
10/25/77
10/25/77
10/25/77
End
3/15/78
2/06/78
10/24/78
12/21/77
3/10/78
North
Latitude:
37°36'36"
37°36'52"
37°36'52"
37°36'51"
37°36'51"
Wesl
Longitude
123°07'32"
123°14'46"
123°14'46"
123°17'27"
123°17'27"
Site
Bep*
914
1372
1372
1829
1829
Meter
Depth
911
912
911
1800
1826
Table 3-2 and Figure 3-2 show the beginning and end of the data records, and the
number of days of recorded data for each current meter. The four Aanderaa meters
recorded from 57 to 141 days of data at 20-minute intervals while the VACM recorded
one complete year of data at 15-minute intervals.
Figures 3-3 and 3-4 show the meter array schematically in profile and plan views,
respectively. The loss of the two meters on Array A (Figure 2-1) resulted in a loss of a
third reference point, which was necessary for the full definition of the two planes
discussed in Section 3-1; however, a sufficient amount of information was recorded by
the remaining meters from which to draw preliminary conclusions.
-------
Table 3-2. Data Record
Meter
Aanderaa 2920
Aanderaa 2830
VACM
Aanderaa 2918
Aanderaa 2919
Start Date
10/25/77
10/25/77
10/25/77
10/25/77
10/25/77
End Date
03/15/78
02/06/78
10/24/78
12/21/77
03/10/78
Number of
Days
141
104
365
57
136
Number of
JJata Records
10154
7547
35018
4016
9800
Although the array was designed to acquire water flow information for 3 months,
considerably more data were acquired, most notably from the VACM. The following
observations were made during the recovery operation conducted on October 24,1978,1
year after deployment:
Meter # 2830 collected data on approximately 75 percent of the output
tape before battery failure occurred. Timing problems with the tape drive
mechanism necessitated a special data recovery technique described in
Section 4.1.
Meter # 2918 collected data on about 50 percent of the tape before
experiencing a battery failure.
Meters 2919 and 2920 had both completely filled their respective output
tapes. Meter # 2920 had experienced tape drive problems similar to the
problems experienced with meter # 2830 (see Section 4.1).
The VACM meter was still functioning at the time of its recovery.
10
-------
37°36'60"
3736'55
37 36'50'
37°36'45"
37°36'40"
37°36'35"
ARRAY
LONGITUDE (W)
123°17'
123°15'
123°13'
123°9'
o
o
\
\
+ARRAY C
'ft
§
o
og
123°7'
123°5'
ARRAY B
(VERTICAL EXAGGERATION = 19:1)
Figure 3-1. Current Meter Locations
-------
AANDERAA .57 DAYS
NO. 2918
AANDERAA .
NO. 2919
136 DAYS
AANDERAA ,_
NO. 2920
141 DAYS
AANDERAA h
NO. 2830
104 DAYS
VACM I-
365 DAYS
OCT
NOV|DEC
1977
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
1978
Figure 3-2. Data Coverage Plot 1977-1978
12
-------
LONGITUDE (W)
500
SI 1000
UJ
Q
1500
2000
123°17' 123°15'
~i r
123°13'
S/N 2918
S/N 2919
o;
123U09'
1
123°07'
123°05'
"
VACM
S/N 2830
S/N 2920
0 km
3.9 km
HORIZONTAL DISPLACEMENT FROM ARRAY D
14.4 km
Figure 3-3. Profile Schematic
-------
N
>,
ARRAY D
ARRAY C
ARRAY D - AANDERAA 2918, 2919
ARRAY C - AANDERAA 2830, VACM
ARRAY B - AANDERAA 2920
TOP VIEW
CURRENT METER ARRAY
(HORIZONTAL SCALE = 10 TIMES VERTICAL SCALE)
Figure 3-4. Current Meter Array Plan Schematic
-------
3.4 BATHYMETRY
Bottom depth profiles were acquired during surveys of the Farallon Islands LLW
disposal site in 1974,1975, and 1977. The equipment used during the 1977 survey
consisted of two precision depth recorders, a GIFFT recorder serving as the primary
sounding device, and an EDO-Western recorder for backup. Sounding profiles were
made with both devices.
Navigation equipment for the 1977 survey included a Motorola Mini-Ranger III,
which received reference signals from Point Reyes and Montara Point This allowed
position readings with an estimated accuracy of approximately 200 meters.
3.5 SEDIMENT INFORMATION
During August and September 1977, six box cores were collected for sediment
analysis ^. The locations of those coring sites are shown in Figures 3-5 and 3-6 and
Table 3-3. The sample at Station 41A was not fully recovered because the jaws of the
box core did not fully close during retrieval.
Table 3-3. Box cores collected for sedimentological studies during R/V
tO IV cruise (August-September 1977)
Station
<
13A
2A
47
39
41A
48
Coring Bate
8/31/77
8/3 1/77
9/1/77
9/1/77
9/1/77
9/1/77
Latitude
37*38.1'N
37«39.8'N
37-383'N
37*38.0'N
37-38.0'N
37-36.6'N
Longitude
123*08.0'W
123-07.1^
123-14.0W
123-17.0W
123-20.7^
123-12.7W
Water
Depth (in)
1042
878
1335
1469
2350
1216
Total
Recovery
C«m)
33
60
34
37
surface
grab
45
Sediment subcore samples were obtained from each of the box cores (except
Station 41A) and were investigated for sediment properties, x-radiographic properties,
and radiochemical properties. See ^ for the procedures followed to obtain and analyze
the samples.
15
-------
I23°30'
123 00
37° 50' -
37°30 -
D NUCLEAR WASTE DUMP SUB5ITES
* BOX CORING STATIONS
AB IS AN EAST-WEST TRANSECT AT NORTH LATITUDE 37° 38'
Figure 3-5. Locations of Core Sample Sites ^.
16
-------
r SEA LEVEL
oc.
LU
UJ
I
I-
o.
500 -
1000-
1500-
2000 -
2500
3000-
• NUCLEAR WASTE DUMP SUBSITES
I BOX CORING STATIONS
10 Km
VERTICAL EXAGERATION 8:1
123°20.7' 123°17' 123°14'
123°08'
-OFFSHORE
AB IS EAST-WEST TRANSECT
Figure 3-6. Depth Profile of Core Sample Sites [31.
-------
4. DATA REDUCTION
In this section the data reduction for the current meters is described through the
initial stages of processing, and the quality of the resulting data is assessed. Steps taken
to improve the quality of the data are also described.
4.1 CURRENT METER DATA
The initial processing of the current data was accomplished in five stages: (1)
data tapes were translated into a computer-readable digital format, (2) data tapes were
checked for format errors such as missing bits in a word, missing words in a record, or
missing records (these errors were then flagged for later processing), (3) digital counts
from the translated data were converted to metric engineering units, (4) data in
engineering units were checked for physically unreasonable values, and (5) both format
errors and physically unreasonable data were removed and replaced by linearly
interpolated values.
When the resulting data were examined it was found that Aanderaa meters 2830
and 2920 contained a large amount of unexplained scatter and were deemed unsuitable
for analysis. In an attempt to recover the data the raw binary data tapes from these
meters were examined with an oscilloscope. It was found that the pulse widths for the
binary ones and zeros varied beyond manufacturer specifications during the record.
However, the ratio of the pulse widths to the time between the trailing edges of
consecutive pulses remained constant. The tapes for meters 2830 and 2920 were then
retranslated using this ratio to distinguish between ones and zeros in the record. In this
way, it was possible to recover the data for the entire period of record for both current
meters.
Verification of the current meter data in terms of relative calibration and timing
accuracy is discussed in Section 5.2.
4.2 GENERATION OF BATHYMETRIC CHART
The bathymetry obtained was annotated to the ship's position at the time of
deployment, and the resulting position and depth coordinates were compared with those
found on an oceanographic data chart, "LBL Publication 92," produced by the Lawrence
Berkeley Laboratory ^, and with bathymetric data from previous surveys in 1974 and
1975. The bathymetric measurements taken during the deployment of the meters
18
-------
generally corroborate the larger features depicted on published bathymetry. However,
smaller features were observed in the local bottom terrain which are not shown in the
standard publications. This situation was also observed with regard to the soundings
taken during the 1975 survey. The soundings performed in 1977 indicate the presence of
an east-west running depression as shown in Figure 2-1. A section of the LBL chart has
been reproduced and is shown in Figure 4-1.
43 GRAIN SIZE FROM SEDIMENT SAMPLES
The tube cores (subcores) described in Section 3.5 were separated by 2-millimeter
and 63-micrometer sieves to determine the percentages of sand, silt, and clay. Sand is
defined as being composed of particles between 2 millimeters and 63 micrometers in
diameter, while silt and clay are less than 63 micrometers in size.
Cumulative probability distributions of grain size were obtained for each of the
sampling stations (see ^). Statistical parameters, including mean, median, standard
deviation, skewness, and kurtosis were obtained for each distribution as a function of
depth beneath the bottom surface in order to determine the cross-sectional profile of the
sediments. This information is used in Section 6.2 for determination of transport
potential from the site.
19
-------
38°N
Kilometers
Nautical
Miles
WASTE
DISPOSAL
AREA
BATHYMETRY PLOT
100-METER INTERVALS
500-METER INTERVALS
30'
20'
10' 123°W 50'
10' 122°W
Figure 4-1. Bathymetry Chart
-------
5. CURRENT METER DATA PROCESSING
This section describes the processing and verification of current meter data that
were performed prior to interpretation. The data from each meter were separated into
data sets after preparation (e.g., translation, conversion, editing, and interpolation)
described in the previous section. The processing consisted of applying various
statistical algorithms to the data in both the time and frequency domain. The output
consisted of time history plots, histograms, scattergrams, progressive vector diagrams,
stick plots, power spectral density plots, and listings of summary statistics.
5.1 DATAPRODUCTS
Various statistical functions were performed on the data to examine different
aspects of current conditions at the site. Time history plots chart continuous
measurements of current flow and direction; histograms and scattergrams provide
individual and combined distributions of the measured quantities; power spectra show
periodic components in the data; and progressive vector diagrams and stick plots
illustrate the long-term average components in the data.
Definitions and samples of the data products prepared from the reduced data are
given below. A complete set of plots and tables are contained in the Appendices.
5.1.1 Time History Plots of Current Speed and Direction
Monthly plots are included for each current meter on a horizontal scale of 1 day
per division. Speed is plotted on a scale of 5 cm/s per division from 0 to 40 cm/s, and
current direction is plotted on a scale of 45 degrees per division. To aid in the
interpretation of current directions, the pen was lifted whenever two consecutive current
directions differed by more than 180 degrees, with the assumption that the current
vector went through 0 degrees (True) between these readings (e.g., for three successive
readings of 359', 0*, and 1', the pen would be up between the first and second point, but
down between the second and third). A reading of 0 degrees (True) for current direction
implies that the current is moving toward true north, while 45* corresponds to northeast,
etc. A sample plot for February 1978 is given for Aanderaa meter # 2920 in Figure 5-1.
The average velocity for this sample is approximately 4 cm/s, with a mildly predominate
northeast direction. The complete set is provided in Appendix A.
21
-------
RRNDERRR »2920
TIME HISTORY PLOT CURRENT SPEED
0
(_>
LU
1 I 2 13
r4
yij-
f 6 I 7 Is" I 9 °"1 "*i 0 "rtr"T 12 13.
li> ~l 16 In I" 18 T19 I 20 ~1 21 122 1 23 1 24 I 2'j> 26 ' 27 ' 28
FEDRUHHY 1978
M1
' \ ,\
o
LLJ
a
RRNUtRRR «2920
TIME HISTORY PEOT CURRENT DIRECTION
'3 ' i I :, I 6 T 7 Ta 1 j I 10 r n ( 12 I 13 f u I it. 1 i» I iv 1 is I i« 1 20 I 21 1 i.1" I"22i I 24
FEBRUHRl 1978
Figure 5-1. Time History Plot, Current Speed, and Direction
(Meter No. 2920)
-------
The time history plots provide a quick and complete visual summary of the
activity of the recorded data over the extent of the plot. Both long-term and short-term
trends are visible in the display and implications can be made concerning the
periodicity, stationarity, ergodicity, and overall quality of the data. Periodicity may be
defined as the degree to which the time history is composed of periodic events.
Stationarity and ergodocity are closely allied concepts. A data set may be considered
stationary if its statistical properties (i.e., mean, variance, etc.) do not exhibit significant
trends with time, and it is ergodic if the statistical properties of its data subsets do not
significantly differ for different subsets. Stationarity and ergodicity are necessary
preconditions for spectral analysis. Visual inspection can frequently be used to pick out
questionable data records (such as excessive values represented by spikes) and locate
places where there was a sensor failure (flat spots in the data or unexplained changes in
the character of the time series).
The plot of current speed in Figure 5-1 clearly shows the tidal periods in the data
and also shows that the data are reasonably stationary and ergodic. The data is
evidently of good quality, as there are no unexpected changes in the overall appearance
of the plot. Peak events are visible on the 9th, 13th, 23rd, 24th and 27th of the month.
The direction time series shows evidence of the tidal rotation in the corkscrew
appearance of the data (i.e., the way it appears to increase through a 360* range with the
passage of time).
5.1.2 Time History Plots of U and V Current Components
The current speeds and directions are also represented in U (eastward-flowing)
and V (northward-flowing) components. The plots are provided on a horizontal scale of
1 day per division and a vertical scale of 10 cm/s per division. The U and V components
are derived as follows:
U = S sin(a)
V = S cos(a)
where
U = East component (cm/s)
V = North component (cm/s)
S = Current speed (cm/s)
a = Current direction (degrees true)
A sample plot for February 1978 is given for meter # 2920 in Figure 5-2. The
complete set is contained in Appendix B.
23
-------
flRNDERflfl «2920
TIME HISTORY PLOT NORTH CURRENT
S 2-
2 ~T~3 f 4 ~TS 1 6 I 7 ~ T8 Tg I
jo \\ 12
n a! 9
T2o rVi~T22~1 23
I 26 '27 I za '
FEBRUflRY 1978
«2920
TIME HISTORY PLOT ERST CURRENT
n I |2 I |3 I u I |'j I IS 1|7 I 18 ' 19
J;LBRUl1RY 1 978
"22 I 23 I 24 I 2b "I 26 TJ7 "1 28~>
Figure 5-2. Time History Plot, Horth Current, East Current
(Meter No. 2920)
-------
These plots corroborate the speed and direction plots by showing the tidal periods
as before, and by revealing no abrupt changes in statistical character and no
nonstationary processes. The currents appear to be more or less symmetrical with
respect to direction, showing that the predominant factors affecting current are periodic
in nature. Tidal influences seem to prevail in these plots. These observations are
supported by inspection of the associated power spectral density plots (see Section 5.1.5).
5.13 Histograms
Histograms for each meter are included to provide the distributions of the
measured quantities over the entire period of record. The following histograms were
calculated:
Current speed using 2.O cm/s intervals
Eastward (U) current component using 4.O cm/s intervals
Northward (V) current component using 4.0 cm/s intervals
Conductivity (for meter # 2920 only) using O.02 millimoho intervals
A sample histogram of current speed for meter # 2920 is given in Figure 5-3.
The complete set is contained in Appendix C.
The frequency histogram displays the frequency of occurrence within specified
class intervals for the parameter shown in the headings. The class intervals are along
the horizontal axis, with the frequency of occurrence within each interval shown on the
vertical axis. The lower bound for each interval is greater than or equal to the value for
the class, and less than the value of the next higher class (i.e., the histogram for speed is
in 2 cm/s intervals and interval class "10" represents all the occurrences of speed greater
than or equal to 10.0 cm/s and less than 12.0). Note that this representation will make a
distribution appear to shift to the left by half a class interval.
Actual frequency for each interval is listed vertically in the heading above the
interval. The total number of events (in this case 10154) is found to the right of the
heading. The resolution of the frequency bars is variable depending on the data being
plotted. The frequency scale is adjusted to give maximum resolution on each plot. Data
flagged as erroneous are not counted in any interval or in the total number of events.
The histogram of Figure 5-3 assumes the general form expected for a non-
negative parameter such as speed. The histogram is skewed towards the origin, or zero
interval class (where 1500 events out of 10154 are observed), and displays a gradual
taper towards the higher amplitudes to the right. There are no isolated events in the
higher class intervals, suggesting that there are probably no unreasonable spikes in the
corresponding time series plot.
25
-------
INTERSTRTC CLBCTTONICS CORP.
OCFJRNIC ENGINEERING DIV.
gWHEIM. CR "IW 772-2811 SITE DEPTH - gu METERS
TI^TQI i SrT«M nSrF1^J?9^5 «, „ l"10^ S/N ~ 292° METER REF - 7H9 RNCHORED BUOT B METER DEPTH - 911 METERS
ipSl^LHTION DRTE 10/24/77 IQW HRS RNCHOR DN 35 HRS RNCHOR RELEflSE 620HRS RECOVERT S/Hi/78 620 HRS
POSITION 37 36 56. ON 125 7 32.1W MflG. VRRIRTION - -17 DECS TIME ZONE - POT CGMT H>8 HRS) SRMP INTV - 20 MIN
FREQUENCY HISTOGRflM
SPEED
1 2
5961
032311
22^082400000000000000000000000000000000000000000000
31
FREQ. 1 1 3
12051
88297520000000000000000000000000000000000000000000
3118
TOTflL
N3. OF
EVENTS
10154
ERRORS
0
1870
1247
0623
\
*••«•»••«« WH«W«««*«VB««BlHHM«MM_«K««««M_«»HH«>BMa>a*WW^**HW^W«**^MM«M(<|llB*» •JBW«.H^_a»«w^««^W^MH« ^
1 2
0 0
3
0
4
0
5
0
6
0
7
0
SPEED
8
0
1
0
0
1
1
0
1 1
3 4
0 0
1 1
5 6
0 0
1 1
7 8
0 0
2.00 CM/SEC INTERVftLS
Figure 5-3. Histogram Current Speed (Meter No. 2920)
-------
5.1.4 Scattergrams
Scattergrams for each meter are included to provide joint distributions of
measured quantities taken over the entire period of record. The following Scattergrams
were calculated:
Current speed versus current direction with speed binned at 2 cm/s
intervals and direction binned at 18 degree intervals.
Northward (V) versus eastward (U) current components with both
components binned at 4 cm/s intervals.
A sample plot of current speed versus direction for meter # 2920 is given in
Figure 5-4. The complete set is contained in Appendix D.
The scatter diagram is a two-dimensional histogram that tabulates the number of
occurrences of specified combinations of two variables. When the number of events in
each box is divided by the total number of observations, the result may be interpreted as
the joint probability density. The class intervals are indicated on the ordinate and the
abscissa. The size of the matrix is limited to 20 by 20. The range for each interval is
limited to integer values. Data flagged as erroneous in either of the variables causes the
occurrence to be dropped from the calculation. The total number of correct occurrences
represented in the diagram is printed at the top of each diagram.
Scattergrams can be used to assess the degree of statistical dependence of two
measured parameters. The scatter diagram of Figure 5-4 shows the relationship of
current speed to current direction for meter # 2920. It is apparent that the greatest
current speeds occurred when the current direction was north, and no currents larger
than 14 cm/s were observed with a southerly direction. The scattergram is smoothly
distributed over the 360 degree range, and there are no speed/direction events isolated by
more than one class interval from the distribution. There also are no conspicuous holes.
5.1.5 Spectral Density Plots of U and V Components
These plots were derived by the method of averaged periodograms from the U and
V components of the current time series. The scale of the monthly spectral plot was
determined by the peak spectral density in order to aid in the resolution of spectral
peaks. The horizontal scale is frequency in cycles/hour and the vertical scale is spectral
density in units of velocity squared times hours, where velocity is in cm/s. A sample plot
for February 1978 is given for meter # 2920 in Figure 5-5. The complete set is
contained in Appendix E.
27
-------
INTEftSTWE ELECTRONICS CORP.
GCERNIC ENGINEERING DIV.
flNFlHEIM. Cfl C7H4) 772-2811 SITE DEPTH - 91H METERS
^Irn, ,Tnr?nM S2W5fl^-5 METER S/N - 2920 METER REF - 7M RNCHORED BUOT B METER DEPTH - 911 METERS
inlTrHiS11^ S^US^iS7 19W HRS nNCM3R DN 35 HRS nNCHOR REVERSE 620HRS RECOVERY 3/U/78 620 MRS
POSITION 37 36 36. ON 123 7 32. IN MflG. VRRIRTION - -17 DECS TIME ZONE - PDT CGMT +08 HRS) SRMP INTV - 20 MIN
SPEELJ
CM/SEC
ua.oo
36.c
32. C
26.
2U.
22.
20.
18.
16.
14.
12.
1Q.
8.C
SCRTTER DIflGRRM
NUMBER OF EVENTS PER MflTRIX UNIT
SPEED VS 01R
TOTRL NUMBER OF EVENTSi 1015H
FLRGGED DRTfl POINTS! 0
.GO
.00
.00
.00
.00
.00
.00
.00
.00
.00
.ac
.00
.00
.00
.aa
.00
.00
.00
.00
nn
l
l
2
1
8
18
56
158
155
93
1
1
3
6
19
16
109
180
183
8«
1
2
5
1U
21
78
1145
165
71
1
2
2
9
17
ue
67
178
176
79
1
1
1
7
21
45
87
167
181
95
6
5
21
35
81
186
169
90
1
2
U
6
28
98
175
173
103
1
2
7
29
89
199
164
115
1
2
1
20
61
136
169
103
1
8
10
11
112
113
86
3
3
18
46
122
162
76
1
6
17
25
63
116
155
66
1
10
15
10
59
92
110
136
53
1
1
6
16
37
61
80
155
164
52
1
U
11
17
38
8>1
ma
161
57
1
1
2
11
20
32
U7
137
me
«6
1
1
1
7
10
22
51
117
121
49
3
1
2
8*
21
61
115
121
60
2
2
5
2
3
4
7
9
22
67
128
129
59
2
2
2
4
7
4
5
8
13
25
50
160
140
63
0.00 36.00 72.00 100.00 144.00 160.00 216.00 £52.00 269.00 321.00 360.00
16.00 54.00 90.00 126.00 162.00 193.00 234.00 270.00 306.00 342.00
DIR
DEC
Figure 5-4. Speed Versus Direction Scattergram (Meter No. 2920)
-------
NORTH CURRENT
320
280-
240
I 200
CM
160
00
-v.
5 120
80
40
0
0.
00 0.25 0.50 0.75 1.00 1.25 1.50
CPH
EAST CURRENT
C\J
00
o
320
280
240
200
160
120
80
40
0
0.
00 0.25 0.50 0.75 1.00 1.25 1.50
CPH
Figure 5-5. Power Spectrum (Meter No. 2920; February 1978)
29
-------
The spectral density function of a parameter displays the amount of variance
around the mean caused by components that occur with a regular frequency. Thus, it is
a method of describing the mean square value or relative energy of a time series as a
function of frequency. Periodic components of the data, such as diurnal and
semidiurnal tidal currents, are represented by peaks in the frequency spectrum. The
relative size of these peaks is an indication of the relative importance of the
corresponding periodic process. Higher peaks demonstrate that more energy is
contained in the measured parameter at the respective frequency.
The spectral density can be thought of as the Fourier transform of the
autocorrelation function, and is calculated using standard FFT algorithms. Figure 5-5
clearly shows the spectral peak for the semidiurnal tides in both the east and north
currents during February 1978. Some energy at higher frequencies is also visible,
especially for the east current, but nothing significant exists at frequencies higher than
0.5 cycles per hour.
5.1.6 Progressive Vector Diagrams
Progressive Vector Diagrams (PVDs) are provided for each of the 5 meters
showing the cumulative vector sums of the east and north current components over the
life of the meter. The PVDs were calculated from time series of low pass filtered data in
order to eliminate tidal variations from the plot, and give a smooth curve depicting long-
term trends. Each of the east and north component time series were filtered with a
cosine-tapered symmetric filter with a cutoff frequency of 0.020 cycles per hour
(corresponding to a 50-hour period). The PVDs were then plotted to a uniform scale of
40 kilometers per inch. A sample plot is given in Figure 5-6 and the remaining plots are
included in Appendix F.
PVDs provide a visual representation of the sequential current observations
recorded by a current meter. Current speeds and directions are broken into their U and
V components and are summed from consecutive data and plotted in vector form-tip to
tail.
An important fact to remember when using PVDs, is that the instrument is fixed
in space as it records current data. The vector sum of the components is not the path of
a water particle through space, but rather the sum of instantaneous current
measurements at one fixed point. The current velocity field must be determined from a
sufficiently defined array of meters before water particle trajectories can be derived.
However, a PVD can be used to support observations about current behavior in the
immediate vicinity of the meter, or to compare measurements from separate locations.
The PVD in Figure 5-6 for meter # 2919 shows the prevailing direction to be toward the
north with a slight tendency to the east. The current vector is relatively straight,
suggesting that a fairly steady residual (after removal of tides) current exists at this
point.
30
-------
FflRflLLON ISlflNO CURRENT METER DflTR
flflNOERflfl METER « 2919
CCT 25. 1977 - M3R 10, 1973
STSHT.iO/25/77 3 0
EMO . 3/10/73 7 0
S3WUHG INTEftVBLi ;.00 W5
PROGRESSIVE VECTOR OlflC-RflM
U CM/SEC V
-^. 220.
K. KN
_^ 202
__ MO. ,
__ IQQJ
_. 30.
__ SO.
_- 40.,
CM/SEC
-120. -100. -80. -SO, -40. -20. 1. 2C. 40. SO. 80, 100.
_L -20.
Figure 5-6. Progressive Vector Diagram (Meter No. 2919)
31
-------
5.1.7 Stick Plots
Three-hour averages of speed and direction are plotted on a polar coordinate
system that moves through time. Stick plots showing current speed and direction are
included in Appendix G. The length of the vector represents speed, and the angle of the
vector represents direction in the true geographic coordinate system (degrees of direction
increase clockwise from north, which is vertically oriented on the diagram). Time is
indicated along the bottom of the diagram. Plotted beneath each stick plot is a time
series plot of speed for quick cross-reference. Peaks in the time series plot can be
matched with their respective vectors on the stick plot.
The stick plot contains the same information as the PVD, except that the vectors
are plotted on a horizontal scale instead of end to end. Figure 5-7 shows a stick plot for
meter # 2918. Inspection reveals a wide range of scatter in both current magnitude and
direction, but there appears to be a strong tendency for the current to travel east-west,
as evidenced by the relatively shallow angle between the stick plot vectors and the
horizontal axis.
5.1.8 Summary Statistics
Daily, monthly, and whole study summary statistics are compiled for each current
meter. The statistics consist of mean, minimum, and maximum values, and also
standard deviation (SIG), skewness (SKW), and kurtosis (KUR). These statistics are
calculated for current speed and north (V) and east (U) current components. A sample
listing of daily statistics for meter # 2919 is given in Figure 5-8. Note that the time
given for each entry in the table is the time at which summary statistics were generated
for all preceding data since the last entry. The remaining tabulations are included in
Appendix H.
The summary statistics show the first four moments of the data distribution and
can be used to assess data quality and stationarity. For stationary data the statistical
properties remain constant over time. Consequently, if the statistical moments (mean,
variance, skewness, and kurtosis) remain within reasonable ranges, the time series can
be considered stationary.
The variance or the standard deviation provide measures of the dispersion of the
data about its mean. An excessive degree of dispersion can indicate discontinuities in
the time series, Or bimodality in the distribution, while a very small dispersion (i.e.,
standard deviation approaching zero) may indicate the absence of a meaningful signal.
This may occur in the event of a sensor failure.
The skewness represents the degree of asymmetry, or departure from symmetry,
of a statistical distribution. If the distribution has a longer tail to the right than to the
32
-------
.
no :u-2o/7i i?
HfEKOCIHC IfclfR.IL. 110. OH 1IH
AANDERAA METER #2918
SPEED
STICK PLOTS
CM/SEC 01R OEG
20 Jt ;2 23 ^* M 1$ 21 2t 23 30 . 2 3
Figure 5-7. Stick Plot (Meter No. 2918)
-------
INTERSTATE ELECTRONICS CORP.
OCEANIC ENGINEERING QEV
ANAHEIM.CA (71*) 772-2811
SUMMARY STATS FOR U
START TIME: 10/25/77 05:
METER TYPE - AANDERAA RCH-5
1NTH
10
10
10
10
«• 10
10
11
11
11
11
11
11
11
11
-: 11
11
11
11
1 1
11
11
11
»: 11
11
11
11
11
11
11
11
11
11
» 11
11
11
11
12
12
12
1 2
12
12
1?
12
=. 12
j 12
12
ii 12
12
OAY
26
27
28
29
30
31
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
23
29
30
1
2
3
4
5
6
7
8
9
10
11
12
13
TR
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
77
TIME
8
3
3
3
3
a
8
3
,3
3
3
3
3
3
3
8
3
3
3
3
3
3
3
8
8
3
3
3
3
3
3
3
3
3
3
3
3
3
3
;)
MEAN
.92
.23
1.00
.14
1.70
1.27
2.30
-.63
.47
-.01
-.32
.70
-.04
-1.09
.31
-.52
-.57
.50
1.00
-.36
1.59
2.15
.76
1.17
1.12
1.21
1.15
2.43
-.30
.71
.61
.97
-.82
-1.07
.71
.59
1.28
.30
1.33
.59
.97
-.15
-.54
-.48
-1.64
.57
-.98
-.93
.00
MIN
-7
-10
-3
-11
-7
-7
-6
-11
-7
-1C
-10
-a
-11
-22
-20
-20
-20
-5
-10
-11
-7
-10
-9
-12
-14
-15
-9
-22
-8
-7
-14
-23
-22
-14
-16
-9
-10
-15
-ft
-8
-18
-2ft
-22
-26
-25
-25
55 - END TIH£:
METER » - 2919
MAX
10
11
12
13
12
7
14
7
7
17
12
12
18
12
14
12
3
9
a
9
11
11
14
19
14
14
13
9
3
11
13
14
23
8
19
11
13
12
11
10
12
16
14
14
14
15
SIG
5.06
3.76
3.55
5.60
4.39
3.66
5.12
4.56
3.96
5.86
4.92
5.01
6.53
8.68
3.12
7.94
7.29
4.24
4.41
4.40
3.97
4.25
4.72
5.19
7.29
7.47
4.89
5.51
3.06
4.03
3.96
5.90
3.76
a. 33
8.55
6.31
5.43
5.30
7.33
6.52
5.19
4.42
7.25
8.57
J.77
7.23
3.57
9.45
9.68
: 3/13/78 07:48
SAMP INTV - 20 HIM
SKU
.13
-.03
.16
.29
-.15
-.50
.55
-.34
-.20
.96
.27
.13
.24
-.08
-.56
-.12
-.76
-.28
-.09
.01
-.22
-.82
-.27
.04
-.29
.16
.13
-.31
-.93
-.04
-.06
-.63
-.34
-.57
.35
-1.25
.73
-.07
-.64
-.69
.07
.43
-.oa
-.01
.26
-.73
-.90
-.71
-.93
KUR
1.49
3.37
1.91
2.52
2.51
2.17
2.30
2.13
1.90
3.26
2.22
2.03
1.59
2.54
2.29
2.63
3.25
1.97
2.20
2.20
2.26
2.80
2.09
2.94
2.19
2.51
2.39
2.19
3.36
2.32
2.19
2.57
3.00
2.93
1.87
3.66
3. SB
1.75
2.ao
2.91
1.74
2.45
2.08
2.09
3.52
3.35
3.39
2.90
3.48
Figure 5-8. Summary Statistics (Meter No. 2919)
34
-------
left of its mean, it has a positive skewness. Likewise, it has a negative skewness if the
distribution is slanted to the left. Quantities which are expected to follow a Gaussian
distribution, such as the east and north current components, tend to exhibit little
skewness. Current speed, on the other hand, should tend towards a positive skewness
since the distribution is bounded on the left by zero.
The kurtosis is a measure of the degree of peakedness of a distribution. The
normal distribution has a kurtosis of 3. A distribution with a steeper peak has a higher
kurtosis and is called leptokurtic. A low kurtosis, corresponding to a platykurtic
condition, is indicative of a relatively flat-topped distribution. A failed sensor can
frequently be identified by the presence of a leptokurtic condition in the measured data
since, in this case, the measured signal will frequently assume a constant or near
constant value. Platykurtic data may indicate bimodality in the distribution.
The minima and maxima are useful in isolating extreme events for closer study,
and may also be used to corroborate the amount of dispersion indicated by the standard
deviation.
Figure 5-8 lists daily summary statistics for the east current from meter # 2919.
Based on the physical nature of the data, the east current would be expected to follow a
more or less Gaussian distribution, and a glance at the tabulated values of kurtosis,
which range between 1.49 and 3.68, suggest a mild degree of platykurtosis. There is also
a degree of negative skewness, indicating the presence of a slant towards the right. This
observation is verified by the frequency histogram (Appendix C). Consequently the PVD
for meter # 2919 might be expected to occupy the first or fourth quadrants, and it does
in fact does occupy the first quadrant.
5.2 VACM AND AANDERAA COMPARISON
The VACM and Aanderra meter # 2830 were located less than 1 meter apart on
the same string and were almost 0.5 kilometer above the bottom. These two meters were
compared by looking at the gain and phase transfer functions between their respective
speed measurements. Since the meters had different sample rates, the VACM output
was interpolated to give equivalent sample rates for comparison. In the ideal case when
the meters measure exactly the same phenomena, the transfer function magnitude would
be equal to one at all frequencies. The phase angle would show any lead or lag between
the meters for different frequencies, and would be linear.
Prior to calculating the transfer function between the two meters, the lower order
statistical functions were reviewed to determine how closely the statistical parameters
were in agreement. It was discovered that the PVDs were not in agreement, and that the
35
-------
VACM data was displaced in time relative to the Aanderaa meter by more than 2
months. Close visual inspection of the two sets of time series indicated that the two
meters were measuring the same events over the following intervals:
Aanderaa # 2830 VACM
10/25/77 4:22 1/09/78 16:41
2/06/78 23:42 4/24/78 21:06
It is possible to visually match these two data records on a peak-by-peak (and
indeed almost a point-by-point) basis.
Thus, it is apparent that at least some of the VACM data was transposed in time.
Evidently, at some point during the initial handling and transcription of the raw VACM
data tape, portions of the data record were processed out of sequence, and the altered
data sequence was retained during all subsequent processing until its discovery during
the analysis described in this report. Further scrutiny of the VACM's PVD revealed the
presence of many unaccounted for cusps in the progressive vector path, suggesting that
these might be break points where transpositions occurred. Efforts were made to
reconstruct the correct sequence of data by examining the data around the cusps and
matching trends with data beginning or ending at other cusps. Each of the
reconstructions began with the portion that matches the PVD of meter # 2830. In
Figure 5-9, a new PVD was constructed by removing the known out-of-sequence data
segment at the beginning of the record to the end of the record while leaving the
remainder of the sequence intact. In Figure 5-10, an effort was made to match the
directions of the low-pass filtered data on either side of the cusp. Figure 5-11 was
constructed in the same manner, but using unfiltered PVD segments so that short
duration direction trends could be matched. Figure 5-12 was made by observing the
visual record of spring and neap tides in the unfiltered PVD sequence, and attempting to
construct a logical sequence from these observations. No one single reconstruction
presented itself as being indisputably correct, and several possible alternatives are shown
in Figures 5-9 through 5-12. Note that all of the reconstructions, as well as the original
PVD, result in a markedly eastward vector average.
In accordance with the above findings, the comparison between meter # 2830 and
the VACM used the 104-day stretch of VACM data starting on January 9, 1978. It was
possible to visually identify corresponding events in the two sets of series, so a timing
comparison was made between the two meters by matching peaks visually and
comparing elapsed times since the beginning of the record. This leads to the observation
that, although the relative drift was approximately 17 minutes for the last 80 percent of
both records, the VACM appeared to lag behind meter # 2830 by around 9 hours for the
first 20 percent of the compared intervals. Visual inspection of the speed time series for
meter # 2830 suggests that there are gaps in the beginning of its data.
36
-------
VACM RECONSTRUCTION 1
A JAN
B FEB
C MAR
D APR
JUN
MAY
6 JUL
N SEP
OCT
AUG
Figure 5-9. VACM Reconstruction 1
37
-------
VACM RECONSTRUCTION 2
NOV
AUG
K OCT
L DEC
Figure 5-10. VACM Reconstruction 2
38
-------
VACM RECONSTRUCTION 3
A JAN
B FEB
MAR
JUN
APR
MAY
JUL
H SEP
I OCT
J AUG
K NOV
L DEC
Figure 5-11. VACM Reconstruction 3
39
-------
VACM RECONSTRUCTION 4
JAN
FEB
MAR
D APR
E
F
G
MAY
JUL
JUN
H SEP
Figure 5-12. VACM Reconstruction 4
OCT
AUG
NOV
DEC
40
-------
The speed transfer function was obtained between meter # 2830 and the VACM
by # 2830/VACM. The unfiltered time series were used so that high-frequency events
would be retained in the comparison. The gain and phase functions are shown in
Figures 5-13 and 5-14, respectively. It is seen that the speed transfer function is fairly
flat and within a range of 0.8 to 1.05, with the VACM generally giving a larger
magnitude than # 2830. The phase portion of the transfer function is linear, indicating
no substantial timing rate shifts between the two meters. This is the type of gain and
phase relationship that should exist between similar sensors measuring the same
parameter under equivalent conditions.
5.3 MEASUREMENT ACCURACY AND TIMING ERRORS
The main known phenomenon in the current measurement field is the occurrence
of tides. All of the current measurements contained a significant amount of tidal energy
at approximately 12-hour intervals. This information was used to determine timing
errors. Once it had been determined with an initial inspection of the time series that
there were no gross timing errors other than the one described for the VACM in the
previous section, meter timing was evaluated by locating tidal peaks and comparing
elapsed time since meter startup with the time obtained by multiplying the tidal period
by the number of elapsed tidal periods.
Semidiurnal tides contain two major frequency components, one due to the
gravitational influence of the moon, the other due to the pull of the sun. The lunar
component has a period of approximately 12 hours, 25 minutes and is the major
component. The solar contribution has a 12-hour period and accounts for around 46.6
percent as much influence as that of the lunar tidal contribution. In accordance with
these facts, the tidal peaks were assumed to be 12 hours 25 minutes apart and heavily
contaminated by other frequencies, mainly the solar 12-hour tide.
Several tidal peaks were selected from the time series for each meter to establish
clock drift as a function of time. Peaks were selected for incorporation into the timing
calculation based on the following criteria:
only peaks which had the greatest speed magnitude in a given
24-hour period were selected; and,
these speed peaks were used only if they occurred
simultaneously with the daily maximum for the east-west
current component.
41
-------
1.60
1.401
1.20
1.00
0.80
5 0.60
5
0.40
0.20-
0.00
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24
FREQUENCY CPH
Figure 5-13. Speed Gain Transfer Function for Meter No. 2830 Versus VACM
-------
200
-200.00
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24
FREQ CPH
Figure 5-14. Speed Phase Transfer Function for Meter No. 2830 Versus VACM
-------
A reference peak was chosen from the beginning of the time series, and the
elapsed number of lunar semidiurnal periods (12 hours, 25 minutes) was calculated for
each candidate peak with reference to the reference peak.
The calculated elapsed tidal intervals consisted of an integral number of periods
plus some residual fraction of a period. Since the relative clock drift for each meter was
found to be within one tidal period for each meter (see bottom of page 41) it was
expected that the calculated residuals would be close to each other so long as the
candidate peaks were all coincident with lunar semidiurnal tides.
In order to eliminate peaks that were not coincident with the lunar tides, the
residuals were histogrammed to tenths of a tidal period to determine which bins
contained the greatest numbers of peaks. In this manner those peaks belonging in the
most frequented bins were readily identified as lunar semidiurnal tidal peaks, while the
remainder were discarded.
A least-squares regression analysis was performed on the final set of residuals to
obtain a linear relationship between the residuals and the elapsed number of periods.
The slope from this relationship is the estimated rate of clock drift for the meter under
consideration. Table 5-1 contains the estimated clock drift for each of the five meters.
Table 5-1. Estimated clock drift
Meter
2830
2918
2919
2920
VACM
Days of Operation
104
57
136
141
104*
Estimated % JMft
Rate
0.024
0.049
-0.035
-0.004
0.011
Estimated
Maximum Time
Shift (minutes)
35
40
-68
-8
17
* Only 104 days were used since this was all that could be
matched with Aanderaa No. 2830.
These results show the meters to be within 108 minutes of each other during the
4 month measurement period. This is less than the advertised accuracies (6 minutes per
year for the Aanderaa meters and 12 minutes per year for the VACM), but sufficient to
perform the type of analyses done for this study.
44
-------
5.4 CONDUCTIVITY AND SALINITY
The conductivity measurements taken by meter # 2920 were reduced to salinity
for the measurement period October 25, 1977 to March 15, 1978. Salinity was calculated
by normalizing the conductivity measurements to 15*C and one atmosphere pressure,
and then applying the following empirical formula (which is given in the operating
manual for the Aanderaa current meters):
S = -0.08996 + 28.2972R + 12.8O832R2 - 1O.67869R3
+ 5.98624R4 - 132311R5
where S = Salinity in parts per thousand (ppt)
R = The ratio of conductivity of a sample (with a temperature of 15*C
and pressure equal to atmosphere) to the conductivity of sample
(with a temperature of 15*C, pressure equal to 1 atmosphere, and
having a salinity of 35 ppt)
The calculated mean value of salinity for the measurement period was found to be
34.49 ppt. This value is slightly below the historical range (34.6 to 35 ppt) for deep
bottom water ^. The calculated monthly means and ranges are listed below (units =
ppt).
Month Minimum Mean Maximum
Octl977 34.49 34.59 34.71
Novl977 34.23 34.57 34.71
Dec 1977 33.79 34.52 34.64
Jan 1978 33.61 34.49 34.64
Febl978 34.29 34.42 34.55
Mar 1978 34.25 3433 34.53
Full Period 33.61 34.49 34.71
Note: Appendix C contains a histogram of the conductivity measurements
from which these salinity statistics were derived.
Appendix H contains summary statistics of the conductivity measurements.
45
-------
6. DATA INTERPRETATION
Three major pieces of information concerning the Farallon Islands LLW Disposal
Site are combined in this section to evaluate the type of currents observed at the site and
to derive an estimate for transport potential of waste material from the site. The
information used for this evaluation includes the current measurements processed as
described in Section 5, the bathymetry describing the depths and slopes in the
measurement area, and (see ^) an evaluation of sediment grain size distributions.
6.1 ANALYSIS OF CURRENTS
Most of the energy stored in deep ocean currents can be characterized by five
components listed below.
Current Component Period Frequency
Large scale > 6 months > 0.0002 cph
Meteorological 3 to 7 days 0.006 to 0.014 cph
Tidal (major) 12 and 24 hrs 0.083 and 0.042 cph
Inertial 12 hr/sin (lat) Sin (lat)/12 hr
Internal waves 10 min to 8 hr 0.125 to 6.0 cph
Large-scale currents balance pressure gradients resulting from density variations
and the Coriolis Force. Since the total measurement time for most current meters is
less than 4 months, this current component will appear as part of the mean, or average
of the current measurement time series. Meteorological currents result from stress at
the sea surface caused by the average wind speed and strength (i.e., length of blow and
fetch area) over the ocean. This component will appear as a low frequency part of the
current measurement spectrum. Tidal currents result from sea surface displacements
caused by astronomical forces. There are approximately 400 different harmonic
components of tides that can be classified into three principal components; semidiurnal
(12 hours), diurnal (24 hours), and long period (0.5 month, 1 month, 6 months). The
semidiurnal tides appear to make up a large part (close to 50 percent) of the total
current energy at the Farallon site. Inertial currents may be excited by other currents
and propagate because of the coriolis force until they are damped by friction. The
period of this component is approximately 20 hours at the latitude of the measurement
field. Internal waves and their associated currents are also excited by other currents and
propagate along density gradients due to stratification of the ocean. Some energy at
internal wave frequencies in the deepwater measurements was observed in the data from
meter # 2919 for the months of December and January, as can be seen from
examination of the spectra in Appendix E.
46
-------
The interpretation of the current data from the present study and from previous
studies ( '*! and ^ ) is described below. The data are resolved into periodic
components, such as tidal currents, and long-term drift currents. The periodic
components are assessed through an examination of the spectral frequency components
and the drift currents are assessed by taking long-term vector averages of the current
time series. These data products are described in Section 5.1. The relationship between
currents measured at the site and large-scale current patterns is also discussed.
6.1.1 Periodic Current Components
The periodic components in the current measurements can be evaluated using the
spectrum generated from the measurement time series. The monthly spectra for each
meter are given in Appendix E. It can be seen that the measurements are dominated by
12-hour periods corresponding to semidiurnal tides. In general, there is also significant
energy in the 20 to 24-hour region, as well as very low frequencies. Occasionally there is
also a peak in the spectrum with a 6-hour period in the north-south direction. This
situation was most prominent during December 1977 and January 1978 for Aanderaa
current meter # 2919. It appears to be the first harmonic of the dominant 12-hour tidal
current, and its exact cause is unknown. One possibility is that this peak is due to an
internal wave generated by strong tidal currents, which reach their maxima during this
period as they flow into and interact with the sloped bathymetry in the area of the
trench south of the deployment site.
To examine the major current energy components in more detail, the data was
decimated (i.e., resampled with larger sampling intervals) using a finite impulse
response digital filter with the following characteristics:
Cutoff frequency: 0.1 cph
Symmetrical, low pass
Number of weights: 121
Cosine taper
Decimation interval: 3 hours
The spectra for the decimated time series are shown in Figures 6-1 through 6-5.
Each spectrum corresponds to the entire measurement record for the respective current
meter. The total relative energy, or statistical variance, in the current measurement
record can be calculated from:
N N
Z S (fi) A fi = 1/N Z (Xi2 - X 2)
,=1 ,=1
47
-------
Where Xi = measurement time series
X = mean of series
N = total number of samples; i = 1, 2,...N
S(fi) = spectral value at frequency fi
Afi = frequency resolution (reciprocal of signal length)
The relative energy near various frequencies of interest can be calculated from:
j+k
E = Z S (fi) Afi
>=j-k
where 'k' is some number of frequency bins about the frequency bin 'j' of interest; the
results are given in Table 6-1. This table shows relative energy (variance) values for low
frequencies (less than or equal to 0.01 cycles per hour), 24-hour tides with 20-hour
inertial currents, 12-hour tides, and the total mean square energy in the record. It is
seen that for the deep near-bottom current meters (# 2918 and # 2919), most of the
energy is related to 12-hour tides and is along the east-west axis. However, for the
bottom meter # 2920 in the east most section of the measurement area, the current
energy (including semidiurnal tides) is slightly greater in the north-south direction.
Aanderaa meter # 2830 and the VACM have a more equal energy distribution with a
significant low-frequency component dominated by the north-south direction. In
conclusion, the periodic tidal energy is dominant in the east-west direction for the deep-
site location, and is more multidirectional at the other measurement points. The
difference between the total energy and the sum of the current components in Table 6-1
is general noise that is relatively flat through the remainder of the spectrum.
To give an indication of the orientation of the rotary tidal current in the area,
tidal ellipses were calculated for the 12-hour energy components listed in Table 6-1. The
calculation was performed as described in Figure 6-6, using the cross-spectrum between
the north-south and east-west current measurements to compute the phase angle. The
results are shown in Figures 6-7 through 6-11. As would be expected, the near-bottom
ellipses have a larger ratio of major to minor axes than those for meters in the midwater
column (Aanderra # 2830 and the VACM).
The measured current speeds generally range between 0 and 20 cm/s, as can be
seen from the time series plots, histograms, and scattergrams given in Appendices A, C,
and D, respectively. The speeds exceed 20 cm/s less than 3 percent of the time, and this
is discussed in the next section as being relevant to sediment transport.
48
-------
Table 6-1. Major Current Energy Sources for Entire Operating Period of Each Meter
Mster
2918 N-S
2918 E-W
2919 N-S
2919 E-W
2830 N-S
2830 E-W
VACM N-S
VACM E-W
2920 N-S
2920 E-W
Lew Frequency
to 0.01 cph
0.4
0.2
6.6
2.5
9.4
3.1
-
Current Components
Tide (24 hour)
Inertial $0 hour)
4.4
0.5
2.9
2.7
2.1
2.4
2.8
0.2
0.4
Tide
(12 how)
0.083 cph
0.5
14.6
0.7
26.1
43
5.2
6.0
6.9
1.4
0.8
Total
Energy
em/s*
2.2
24.2
4.6
35.0
17.6
13.0
22.1
16.8
43
3.8
49
-------
NORTH CURRENT
o
3C
CVJ
o
SEMIDIURNAL TIDES
0.000 0.028 0.056 0.083
CPH
0.111
0.139 0.167
EAST CURRENT
SEMIDIURNAL TIDES
0.000 0.028 0.056 0.083
CPH
0.111
0.139
0.167
Figure 6-1. Power Spectrum for Meter No. 2918
50
-------
NORTH CURRENT
0
0.000
DIURNAL TIDES
SEMIDIURNAL TIDES
0.028 0.056 0.083 0.111 0.139
CPH
0.167
EAST CURRENT
o
o
X
CXL
O
^
CVI
oo
\
O
160
140
120
100
80
60
40 j
20
0.000
DIURNAL TIDES
SEMIDIURNAL TIDES
0.028
0.056
0.083
CPH
0.111
0.139
0.167
Figure 6-2. Power Spectrum for Meter No. 2919
51
-------
NORTH CURRENT
0.000
SEMIDIURNAL TIDES
20-HOUR PERIOD
0.028
0.056
0.083
CPH
0.111
0.139
0.167
320
280
240
200
160
oo
80
40
EAST CURRENT
20-HOUR PERIOD
SEMIDIURNAL TIDES
\
0.000 0.028 0.056
0.083 0.111
CPH
0.139 0.167
Figure 6-3. Power Spectrum for Meter No. 2830
52
-------
NORTH CURRENT
800
700
2 600
X
LUNAR
SEMIDIURNAL
1 TIDES
/
g 500 I!
1 P
z 400
CM
\ 300
~ 200
100
0
DIURNAL 20-HOUR
TIDES, PERIOD
k - \ji.__ J
r
SOLAR
SEMIDIURNAL
TIDES
/
s
\
0.000 0.028 0.056 0.083 0.111 0.139 0.167
CPH
EAST CURRENT
640
560
480
o
*" 400
X
= 320
o
in
CM_ 240
§ 16°
80
LUNAR
ll rmiDMAi 20-HOUR
| D URNAL pERIOD
M TIDESv /
L x ./
Q|" W A^WA^^s^y
^_^ SEMIDIURNAL
TIDES
SOLAR
SEMIDIURNAL
TIDES
V-..I
0.000 0.028 0.056 0.083 0.111 0.139 0.167
CPH
Figure 6-4. Power Spectrum for VACM
53
-------
NORTH CURRENT
160 1
140
o
- 120
X
§5 100
o
xT 80
oo
s 60
o
40
20
0;
^/SEMIDIURNAL
-"V^yVV^N-^-N/A_/Ay'~~v^~ ^^^^\J
^ TIDES
\^ ^__
!i iii
0.000 0.028 0.056 0.083 0.111 0.139 0.167
CPH
EAST CURRENT
320
280
240
1 200
DT
^ 160
o 120
*" **
80
40
n
u
0.
DIURNAL
TIDES^
^^ 1
1
y\ /V A XI » A HA^J, ^ A
/ \J Vw^WvUvJ V^^V^X/^
^^SEMIDIURNAL
TIDES
V
v^ _
000 0.028 0.056 0.083 0.111 0.139 0.167
CPH
Figure 6-5. Power Spectrum for Meter No. 2920
54
-------
ROTARY TIDAL CURRENT:
WHERE
Cx = Ax sin (wt + 0X) (east/west)
Cy = Ay sin (wt + 0y) (north/south)
THE VECTOR (T(t) WILL PARAMETERIZE AN ELLIPSE
BOUNDED BY THE RECTANGLE
xl fi Ax and ICyl £ Ay
C
-*x
/
/^
y
Ay
/
•/
-Ay
AX
Figure 6-6. Tidal Ellipse Calculation
55
-------
8.00
6.00-
4.00-
2.00-
-2.00
-4.00
-6.QO--
-8.00
-12.
00 -10.00 -8.00 -6.00 -4.00 -2.00 0.00 2.00 4.00 6.00 8.00 10.00 12.00
2918 E/W
Figure 6-7. Tidal Ellipse for Meter No. 2918
-------
8.00
6.00
4.00
2.00--
CT.0.00
i-H
CM
-2.00-
-4.00
-6.00-
-8.00
-12.00 -10.00 -8.00
-6.00 -4.00 -2.00 0.00 2.00
2919 E/W
4.00 6.00 8.00 10.00 12.00
Figure 6-8. Tidal Ellipse for Meter No. 2919
-------
8,00
6.00- - - —
4.00
2.00
c/o
o 0.00
co
OJ 00
OO CM
-2.00
-4.00-
-6.00-
-8.00
-12.
00 -10.00 -8.00 -6.00 -4.00 -2.00 0.00 2100 4.00 6.00 8.00 10.00 12.00
2830 E/W
Figure 6-9. Tidal Ellipse for Meter No. 2830
-------
S.OOr
6.00
4.00-
2.00
0.00
o
-------
8.00
6.00--
4.00-
2.00-
CO
§ o.oo-
CM
-2.00-
-4.00-
-6.00-
-8.00
-12
00 -10.00 -8.00 -6.00 -4.00
-2.00 0.00 2.00
2920 E/W
4.00 6.00 8.00 10.00 12.00
Figure 6-11. Tidal Ellipse for Meter No. 2920
-------
6.1.2 Average Current Components
Long-term average currents were calculated by taking vector averages of the
current velocities for each meter. The results are shown in Table 6-2 and depicted
graphically by the PVDs given in Appendix F. The average currents are much smaller
than the periodic currents (less than 2 cm/s) and move in the northeast quadrant for the
bottom measurements and the southeast quadrant for the midwater column
measurements.
Table 6-2. Vector-averaged Current Velocities
Meter
2918
2919
2830
VACM
2920
Speed
(cm/s)
1.40
1.70
0.67
0.90
0.17
Direction
(degrees true)
69
2
137
152
57
In an attempt to determine characteristics of near-bottom circulation patterns
within the area, an earlier measurement program (^) deployed two meters above the
seafloor near the 1700-meter site in 1975. The current meters, emplaced 21 August 1975
with recordings over a 27-day period, were located 16 kilometers apart and occupied a
square centered at 37*38'N and 123*18'W. After removal of the tidal components, bottom
currents were found to be clearly north with a mean direction of 004* azimuth and mean
velocity of 133 cm/s in general agreement with the present (1977-1978) study results.
Although the detailed current structure of the ocean is complex and is not totally
understood, there are certain large-scale patterns caused by global forces such as the
prevailing winds, tides, and rotation of the earth. These forces produce large-scale
patterns in both the surface and deepwater currents that can be seen in the data.
The major surface current off California is the California Current. It moves
southward and is part of the north subtropical gyre in the Pacific Ocean, caused by the
prevailing winds in the Northern Hemisphere, the continental boundary of North
America, and the Coriolis Force. Like all wind-driven currents, it does not penetrate
beyond a depth of approximately 100 meters. There is a large-scale deepwater current
off California which flows northward, and tends to offset the southward California
Current, but it is much slower.
61
-------
The Coriolis Force (caused by the earth's rotation) tends to give the southward
surface current an offshore (westward) component, and the northward deepwater current
an onshore (eastward) component. In the coastal region the colder bottom water tends
to rise to take the place of the surface water that has been transported offshore.
The general trend of the survey data agrees with this global phenomena, but
shows significant variation due to local conditions such as bathymetry. This can be seen
from examining the periodic current component energy in Table 6-1 and the long-term
average currents in Table 6-2.
6.2 TRANSPORT POTENTIAL
In this section the current data, bathymetry, and sediment grain size are
evaluated to determine an estimate of the potential for sediment transport from the
waste disposal site. It should be noted that this inquiry makes no claim to present any
quantitative estimates of sediment transport paths. However, the data do justify a
qualitative assessment of the likelihood that sediments will become suspended, and
should this occur, in which general directions transport is most likely. The method
followed includes a review of the determined grain size distribution (see *• ')$ a
determination of the potential for suspension of these grain sizes relative to observed
current speeds; and, an estimation of transport speed and direction from long-term
average currents at each measurement location.
Transport may take place in essentially two modes: (1) transport within the
water mass which can be inferred from measurements of two-dimensional observations
such as those presented in this report, and (2) bedload transport. To realize transport
within the water mass, material must be suspended in that water mass. One important
assumption for transport analysis is that suspended material follows water mass motion
as a particle of water and without divergence.
Transport at the interface of the water and sediment is called bedload transport.
This is a mobilization of the deposited material which moves it horizontally across the
bottom, possibly resuspending it for short intervals in a siltation process.
This section considers only the transport of fine-grained sediments (silts and
clays) in the water mass for the following reasons:
Grain size analysis (see ^) shows these relatively fine
sediments are present throughout the site region and
adjacent areas. In combination they comprise more than
half, and usually nearer to 65 percent of the sediment
volume (Table 6-3).
62
-------
The fine-grain sediments offer large surface area-to-volume
ratios that enhances the ability to act as a carrier for
adsorbed radioactive material.
The site contains packaged waste material. Only when the
implosion of the container causes exposure of the surrounding
sediments and water to radioactivity is transport of concern.
Coarse material is restricted to transport by major, impulsive
events such as slumps or turbidity flows which move short
distances with long time periods of no activity in between.
Fine sediment, coupled with currents that exceed 20 cm/s,
offer the major possibility for transport over long distances,
and at high enough speeds to be seriously considered for impact
on regions outside site perimeters. This speed range is
approximately the threshold above which sediment resuspension
r^i
can occur (l J).
Table 6-3. Sediment Surface Grain Size Data
from Farallon Islands LLW Site Survey Cores
Station
Humber
13A
48
47
39
Sample Depth
0 to 1 cm
0 to 1 cm
0 to 1 cm
0 to 1 cm
Weight Percent
Sand
28.8
13.8
15.9
38.0
Silt
44.8
65.7
61.0
36.5
Clay
26.8
20.5
23.1
25.5
Proportional
Sediments
(% Fines)
71.6%
86.2%
84.1%
62.0%
The data available for this study consist of the observations of currents described
in Section 6.1, and grain size distributions (see I3J) from four locations in the area of
the dump site extending from water depths of 878 meters to 1350 meters. See [3] for
complete descriptions of sampling locations and the analytical techniques applied. It is
important to note that a wet-sieve technique was used to determine the distributions at
the sample locations.
63
-------
The arrangement of current meter arrays for the 1977-1978 study was established
to describe movement of water at positions near bottom (within 1.5 meters) in an effort
to infer where particulate matter would travel while moving near the waste material
lying on the bottom. Other meters were suspended at locations in the water column as
follows:
A current meter about 30 meters above the bottom at the
deepest location to estimate the depth of the boundary
layer dynamics and the shift of speed and direction due
to boundary layer influences.
Current meters at 911 meters below the surface some 10
kilometers seaward of the near-bottom meter location in
914 meters of water. This approximates a plane permitting
examination of the horizontal large-scale motions onshore
or offshore from the shallower site.
Because of the nature of the study the conservative assumption is that fine
sediments in the form of silts and clays can become suspended in the water mass at
current velocities exceeding approximately 20 cm/s. The percent of time that the
measured speeds exceeded various velocities was determined for each current meter.
The results for 20 cm/s are listed in Table 6-4 and two exceedence diagrams are given in
Figures 6-12 and 6-13. Figure 6-12 shows the percent exceedences for all five meters
with all of the measurement period considered for each meter. Figure 6-13 shows
exceedences with the measurement periods truncated to cover the same period (October
through December 1977). The suspension threshold velocity will vary with grain size and
cohesivity, but the effect of choosing a different threshold than 20 cm/s can be estimated
by referencing these figures.
The following discussion uses a 20 cm/s threshold as its basis. At each near-
bottom measurement location, the speeds do become great enough to cause some
suspension of the fine-grain sediments. Of the two bottom meters, # 2919 had the
greatest percentage of large current speeds, approximately ten times that of meter #
2920. Both meters operated for more than 4.5 months. Since the grain size distribution
is similar at both meter locations, it is concluded that a greater potential exists for fine
sediment suspension at the western (down slope) end of the site than at the eastern end.
The fine-grain sediment will drop out of suspension only at very low current speeds. The
speed histograms in Appendix C imply that there will be current velocities sufficient to
keep the sediment in suspension for periods of days to weeks once it becomes suspended.
64
-------
Table 6-4. Percent of Total Time Measurement Speeds Exceed 20 cm/s
Aitay
B
C
C
D
D
Meter
2920
2830
VACM
2918
2919
Percent
of Time
0.23
0.01
0.08
0.77
2.74
Itotal
Measurement
(Days)
141
104
365
57
136
For the purposes of this inquiry, transport of fine-grain sediments in suspension
can be estimated by examining the long-term vector-averaged velocities given in Table
6-2. At the deep western part of the site, meter # 2919 measured an average speed of
1.7 cm/s (1.46 tan/day) due north. Meter # 2918 is 100 meters above # 2919 and gives
similar average speeds, predominately in the eastward direction. For sediment in the
water column, the general direction of transport is to the northeast (up slope). This can
be seen in greater detail from the PVDs in Appendix F. However the average speed
measured by # 2920 at the eastern (up slope) end of the site is 0.17 cm/day to the east,
which is significantly less than that measured at the western end. Also, # 2830 and the
VACM show a southeast movement of 0.67 and 0.90 cm/s, respectively. Since these
meters effectively define a horizontal plane with meter # 2920, the onshore transport
diminishes significantly at the east end of the site. It therefore appears that there is
little chance for any significant transport to continue toward shore; however, there are
no explicit measurements to verify this.
In conclusion, current measurements suggest that fine-grain sediment may
become suspended with the greatest likelihood of suspension being in deep waters at the
western end of the site. Long-term average currents move eastward within the site, but
diminish at the shoaling eastern part. The potential for sediment suspension and
transport exists, but is small. If necessary, it can be further defined with more detailed
analysis, as suggested in Section 7.
65
-------
32.00
28.00
24.00
20.00
£16.00
212.00
8.00
4.00
0.
,00 12.00
14.00 16,00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34,
SPEED (CM/S)
00
Figure 6-12. Exceedence Curves for Total Measurement Periods
-------
32.00
28.00
0.00-
10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.00
SPEED (CM/S)
Figure 6-13. Exceedence Curves for October through December 1977
-------
7. RECOMMENDATIONS FOR FURTHER STUDY
A major task for this study was to prepare recommendations for future sampling
requirements necessary to develop reliable predictive models for the potential transport
of soluble and particulate radioactive materials from formerly used U.S. Pacific Ocean
LLW disposal sites. Section 6 described the mechanisms for potential transport based
on data accumulated to date. Detailed knowledge of current speeds and sediment
composition at the disposal site will yield qualitative information as to whether
sediments that may contain radioactive particles from the waste materials could be
transported from the site. Transport by ocean currents is believed to be an important
mechanism for the possible redistribution of the radioactive materials. The measured
speeds from the 1977 survey were great enough so that at least some transport could
occur. However, without a more detailed picture of the physical mechanisms involved,
the amount and distribution (i.e., direction) of the radioactive sediments cannot be
determined.
To develop predictive models which will indicate the probable quantity and
direction of the transport of radioactive waste from a disposal site, it is necessary to first
develop a conceptual model of the important physical processes involved and their
relationships. Presently, there are some gaps in the knowledge of the physics of
sediment transport and dispersion of dissolved and suspended particles. However, this
should not be a deterrent to the model formulation for various reasons. First, the
monitoring and assessment of the potential environmental effects from waste disposal
sites is of great concern, and a good conceptual model can be of great use in the design
and planning of future measurement programs. The results of the surveys can be used
to enhance the model, which will then result in a more effective monitoring program.
Second, it may be sufficient to use the model to derive an upper limit for the transport
of waste materials. The upper limit can take the extreme case of additive adverse
assumptions, and if the resulting exposure is still within acceptable levels, further
research of the physical processes becomes less important. If, on the other hand, the
results are not within acceptable limits, the model may be further elaborated to see if a
more realistic (but more complicated) picture of the physical situation will produce
more acceptable results. Thus, in the case of the Farallon Islands study, one may
consider the curve for the threshold of sediment transport and suspension as a function
of sediment grain size and current speed, to be a very simple conceptual model. If the
measured current velocities were not great enough to transport sediment, there would be
a strong indication that this is not a problem at the dump site unless the measurements
were taken during an unusually calm period. Since this was not the case during this
study, it is recommended that further development of a conceptual model be initiated for
the distribution and dispersal of radioactive materials from waste disposal sites. This
model may take various forms, but is generally a specification of the important physical
quantities and boundary conditions with a set of mathematical equations expressing
their relationships.
68
-------
One possible approach would be to consider the current speed and direction data
as inputs of a stochastic process, which would generate a set of transport probability
distributions based on a well-defined sediment transport formulation such as the
Hjulstrom curves (m). This formulation would assign probabilities of initial and
sustained suspended transport in terms of current speed and mean grain diameter.
Implementing this approach would involve the use of standard Monte Carlo
computational procedures. The probabilistic approach could take into account
unmeasured or inherently stochastic aspects of the processes being evaluated (i.e., the
variability of suspension threshold velocities with the cohesivity of fine grained
sediments). It would also offer some direct insight into the degree of uncertainty
involved in the predicted transport modes, and statistical confidence units could be
ascribed.
To be of the greatest possible utility, the conceptual model (verified and modified
by measured data) should be supported by a numerical model. The numerical model is
a realization of the conceptual model, which will provide quantitative estimates of the
process under investigation for conditions relevant to the area under study. The
numerical model can be as simple as a mathematical formula or statistical relationship
between the conditions in the area and the resulting effects, or as complex as an
evaluation of the coupled equations of motion of all of the important quantities. It is
important to also estimate the accuracy of the model to assure that the quality of its
predictions are within acceptable limits. It is also important not to make the model
more detailed than is justified by the knowledge of the conditions in the area and the
physical processes involved. The model can be realized in modular form so that as more
information about the area and mechanisms are gathered, it can be easily enhanced to
give more accurate predictions. It can then be run under a variety of conditions and
assumptions and easily applied to other areas that contain related processes.
A relatively simple procedure for estimating the potential for transport is to
implement suspension and bedload transport criteria for a distribution of grain sizes on
each of the raw current velocity/direction time series to produce output time series with
the following features:
Whenever it is assessed that there is potential for transport
(i.e., particle is suspended) the output velocity time series
will be set equal to the input time series multiplied by the
estimated probability of suspension.
When transport is deemed unlikely (i.e., particle is out of
suspension) the output velocity series will be zeroed.
The direction time series is kept identical with the input
direction series.
69
-------
The resulting velocity/direction output series will be a representation of the
motion of sediment at the given current meter for each grain size. This time series can
then be used to generate PVDs and scattergrams to visually represent the potential for
transport as a function of direction. The sensitivity of these estimates (and hence the
amount of uncertainty) will be largely reflected in the amount of variation of the PVDs
as a function of grain size.
The most direct and accurate determinations of the physical conditions in an area
of the ocean are determined from direct measurements. Measurement programs are,
however, expensive, and therefore limited in both spatial extent and temporal coverage.
It is usually essential that any oceanographic investigation be supported by some
measured data, as was provided by the present Farallon Islands LLW disposal site study.
It is desirable that the program be designed with a conceptual model of the important
processes taken into account, and that the results be used to test, verify, and calibrate
the model.
It is believed that additional current measurements would yield a greater
understanding of the potential for transport of radioactive materials in the case of the
Farallon Islands LLW disposal site. Measurements of deepwater currents are rare and
difficult to obtain, and the success of the survey analyzed in this report is encouraging.
However, only one of the meters was capable of recording data for an entire year. The
others recorded data for the late fall and winter seasons only. In addition, it is not
known whether the data was taken during a typical year, or a year with unusually low or
high activity. Additional data would help to answer this question and provide a greatly
increased statistical data base from which assessments of conditions in the area can be
made. Statistical data for normal and extreme conditions are of importance. Once both
of these data bases have been collected and are large enough to inspire confidence, the
area can be characterized in a statistical sense, and projections of current related
processes can be derived for the future.
If more current measurements are to be obtained, it is recommended that a
careful calibration program on all instruments be carried out. The present study
provided an excellent opportunity to compare the responses of two commonly used
current meters: the Aanderaa and VACM. These gave the same general statistical
results, but differed somewhat in their detailed time series measurements under the
same conditions.
70
-------
8. REFERENCES
111 Joseph, A. 1957. United States' Sea Disposal Operations, A Summary to
December 1956. Technical Information Service Report No. WASH-73. United
States Atomic Energy Commission, Washington, DC.
121 Noshkin, V. E. 1978. Radionuclides in the Marine Environment near the
Farallon Islands. Lawrence Livermore Laboratory, Univ. of California,
Livermore.
[31 Dayal, R., Duedall, I.W., Fuhrmann, M. and Heaton, M.G. 1979. Sediment and
Water Column Properties at the Farallon Islands Radioactive Waste Dumpsites.
Final report to the Office of Radiation Programs. U. S. Environmental Protection
Agency, Washington, DC.
141 Wilde, P. 1976. Oceanographic Data Off Central California, 37* to 40* North
Including the Delgada Deep Sea Fan. LBL Pub. 92. Lawrence Berkeley
Laboratory, Univ. of California, Berkeley.
151 Sverdrup, H., Johnson, M.U. and Fleming, R.H. 1955. The Oceans - Their
Physics, Chemistry and General Biology. 6th ed. Prentice-Hall, Inc. Englewood
Cliffs, NJ.
U.S. Environmental Protection Agency. 1983. Analysis of Ocean Current Meter
Records Obtained from a 1975 Deployment off the Farallon Islands, California.
Office of Radiation Programs, Washington, DC. EPA 520/1-83-019.
Dyer, R. 1976. Environmental Surveys of Two Deep Sea Radioactive Waste
Disposal Sites Using Submersibles. Proceedings of an International Symposium
on Management of Radioactive Wastes From the Nuclear Fuel Cycle, Vol. I.
International Atomic Energy Agency, Vienna, Austria. Pages 317-338. IAEA-SM-
207/65.
181 Hjulstrom, R. 1935. Studies of the Morphological Activity of Rivers as
Illustrated by the River Fyris. Bull, of the Geol. Inst, Uppsala, Sweden. Vol. 25,
Pages 221-527.
71
-------
9. BIBLIOGRAPHY
Aanderaa Instruments. 1976. Operating Manual for Recording Current Meter
Model 5. Aanderaa Instruments, Bergen, Norway.
AMF Sea-Link Systems. 1976. Vector Averaging Current Meter. AMF Sea-Link
Systems, Herndon, VA.
Bendat, J.S. and Piersol, A.G. 1971. Random Data: Analysis and Measurement
Procedures. Wiley-Interscience, NY.
Bouma, A.H. 1962. Sedimentology of Some Flyash Deposits, Elsevier,
Amsterdam, Netherlands.
Carney, R. S. 1979. A Report on the Invertebrate Megafauna Collected by
Otter Trawl at the Farallon Islands Radioactive Waste Disposal Site during the
August, September and October, 1977 Cruises of the R/V Velero. Prepared for
the Office of Radiation Programs. U.S. Environmental Protection Agency,
Washington, DC.
Conomos, T.J., McCulloch, D.S, Peterson, D.H., and Carlson P.R. 1971. Drift of
Surface and Near Bottom Waters of the San Francisco Bay System, March 1970-
April 1971. Miscellaneous Field Studies Map, M.F. 333. U.S. Geological Survey,
Reston, Virginia.
Dayal R., Oakley, S. and Duedall, I.W. 1976. Sediment-Geothermal Studies of
the 2800 m Nuclear Waste Disposal Site. Final report to the Office of Radiation
Programs, U.S. Environmental Protection Agency, Washington, DC.
Defant, A. 1961. Physical Oceanography Vols. I & II. MacMillan Co., NY.
Friedman, G.M. and Sanders, J.E. 1978. Principles of Sedimentology. John
Wiley & Sons, NY.
Goldberg, E.D. et al., eds. 1977. The Sea: Ideas and Observations on Progress in
the Study of the Seas. Vol. 6: Mar. Modeling. John Wiley and Sons, NY.
Gotshall, D.W. and Dyer, R.S. 1987. Deepwater Demersal Fishes Observed From
The Submersible AVALON (DSRV-2) Off The Farallon Islands, 24 June 1985.
Marine Resources Technical Report # 55. California Department of Fish and
Game, Long Beach, California.
72
-------
BIBLIOGRAPHY (continued)
Hill, M.N., ed. 1962. The Seas: Ideas and Observations on Progress in the
Study of the Seas. Vol. 1: Phys. Oceanog. John Wiley and Sons, NY.
Keenan, Pat. 1980. Aanderaa RCM Compass Errors. Exposure; a Newsletter for
Ocean Technologists. Vol. 8, No. 2.
Kuenen, P.H. 1953. Significant Features of Graded Bedding. Bull, of American
Assoc. of Petrol. Geol. Vol. 37., Pages 1044-1066.
Reish, D.J. 1978. Study of the Benthic Invertebrates Collected from the United
States Radioactive Waste Disposal Site off the Farallon Islands, California.
Prepared for the Office of Radiation Programs. U. S. Environmental Protection
Agency, Washington, DC.
Schell, W.R. and Sugai, S. 1978. Radionuclides in Water, Sediment and
Biological Samples Collected in August-October, 1977 at the Radioactive Waste
Disposal Site Near the Farallon Islands. Final report to the Office of Radiation
Programs, U.S. Environmental Protection Agency, Washington, DC.
Shepard, F. P. 1973. Submarine Geology. 3rd ed. Harper & Row Pub., NY.
Swift, D.J. et al., eds. 1973. Shelf Sediment Transport: Process & Pattern.
Academic Press, Inc., NY.
U.S. Environmental Protection Agency. 1975. A Survey of the Farallon Islands
500 Fathom Radioactive Waste Disposal Site. Office of Radiation Programs,
Washington, DC. Tech Note ORP 75-1.
U.S. Environmental Protection Agency. 1982. Analysis of Current Meter Records
at the Northwest Atlantic 2800 meter Radioactive Waste Dumpsite. Office of
Radiation Programs, Washington, DC. EPA 520/1-82-002.
U.S. Environmental Protection Agency. 1983. Survey of the Marine Benthic
Infauna Collected from the United States Radioactive Waste Disposal Sites off the
Farallon Islands, California. Office of Radiation Programs, Washington, DC.
EPA 520/1-83-006.
U.S. Environmental Protection Agency. 1988. A Study of Deep-Ocean Currents
Near the 3800 m Low-Level Radioactive Waste Disposal Site, May 1984 - May
1986. Office of Radiation Programs, Washington, DC. EPA 520/1-88-007.
73
-------
BIBLIOGRAPHY (continued)
U.S. Environmental Protection Agency. 1988. Waste Package Performance
Criteria for Deepsea Disposal of Low-Level Radioactive Waste. Office of
Radiation Programs, Washington, DC. EPA 520/1-88-009.
U.S. Environmental Protection Agency. 1988. A Monitoring Program for
Radionuclides in Marketplace Seafoods. Office of Radiation Programs,
Washington, DC. EPA 520/1-88-010.
U.S. Environmental Protection Agency. 1988. Sediment Monitoring Parameters
and Rationale for Characterizing Deep-Ocean Low-Level Radioactive Waste
Disposal Sites. Office of Radiation Programs, Washington, DC. EPA 520/1-87-
011.
U.S. Environmental Protection Agency. 1989. A Review and Evaluation of
Principles Used in the Estimation of Radiation Doses Associated with Deep-sea
Disposal of Low-Level Radioactive Waste. Office of Radiation Programs,
Washington, DC. EPA 520/1-89-019.
U.S. Environmental Protection Agency. 1990. Analysis and Evaluation of a
Radioactive Waste Package Retrieved from the Farallon Islands 900-meter
Disposal Site. Office of Radiation Programs, Washington, DC. EPA 520/1-90-
014.
U.S. Environmental Protection Agency. 1990. Recovery of Low-Level Radioactive
Waste Packages from Deep-Ocean Disposal Sites. Office of Radiation Programs,
Washington, DC. EPA 520/1-90-027.
74
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA 520/1-91-009
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Ocean Current Measurements at the Farallon
Islands Low-Level Radioactive. Waste Disposal
Site, 1977-1-978
5. REPORT DATE
April 1991
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
D. Crabbs, R. Crane, and D. Friedlander
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Interstate Electronics Corporation
1001 East Ball Road
Anaheim, CA 92803
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
Contract 68-01-0796
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Radiation Programs (ANR-461)
401 M Street, SW
Washington, DC 20460
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
ANR-461
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report presents data from an ocean current measurement
study, conducted during 1977 and 1978, in the area of the Farallon
Islands low-level radioactive waste (LLW) disposal site off the
coast of San Francisco, California. The purpose of this study was to
measure near-bottom and bottom currents in the area, and utilize
available historical data, to determine the potential for transport
of LLW from the disposal site toward populated areas in the vicinity
of San Francisco. Interpretation of the current meter data combined
with other available data taken at the site during previous studies
in 1974 and 1975 is also presented. The appendices to the report
contain computer generated graphical displays of the output data
from all the current meters.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
:. COSATi Field/Group
1. ocean disposal
2. low-level radioactive waste
disposal
3. radionuclide transport
4. ocean bottom currents
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (Tins Report!
Unclassified
84
20. SECURITY CLASS (This page;
Unclassified
22. PRICE
EPA Form 2220-1 (R»v. 4-77) PREVIOUS EDITION is OBSOLETE
------- |