United States        Office of
           Environmental Protection     Radiation Programs      June 1988
           Agency          Washington, DC 20460
           Radiation
&EPA     A Study of
           Deep-Ocean Currents
           Near the 3800 m Low-Level
           Radioactive Waste
           Disposal Site
           May 1984 - May 1986

-------
                                     EPA 520/1-88-007
                                   SAIC - 87/7503/133
  A STUDY OF DEEP OCEAN CURRENTS NEAR THE 3800M

   LOW-LEVEL RADIOACTIVE WASTE DISPOSAL SITE


             May 1984  -  May 1986
                       By
   Christopher Casagrande and Peter Hamilton
Science Applications  International Corporation
             Newport, Rhode Island
Prepared as an account of work sponsored by the
 United States Environmental Protection Agency
under Interagency Agreement No. DW14930778-01-0
                Project Officer

               William R. Curtis

         Analysis and Support Division
    Office of Radiation  Programs  (ANR-461)
     U.S. Environmental  Protection Agency
           Washington, D.C.    20460

-------
                            FOREWORD
     Between 1946 and 1970, the United States disposed of
low-level radioactive wastes (LLW) at sites in the Atlantic and
Pacific oceans that had been designated for disposal by the
former U.S. Atomic Energy Commission.

     In 1972 the Congress enacted the Marine Protection,
Research, and Sanctuaries Act (Public Law 92-532).  This Act
charged the Environmental Protection Agency (EPA)  with the
responsibility to develop criteria and regulations for any ocean
disposal of waste materials, including LLW.

     In 1974 the EPA Office of Radiation Programs (ORP)
initiated studies at previously used LLW disposal sites to
determine whether present ocean survey technologies could
determine the fate of radioactive wastes disposed in the past.
Early studies focused on collecting sediments and marine
organisms for radionuclide analyses and evaluating the condition
of LLW packages previously disposed on the seafloor.  Subsequent
studies were conducted to identify pathways and determine
mechanisms for potential transport of radionuclides from
disposal sites.  Sediment resuspension and transport, due to the
effects of near-bottom ocean currents, is a potential mechanism
for dispersal of radionuclides in the deep-sea.  Thus, ORP has
studied near-bottom circulation in and near disposal sites to
determine the potential for shoreward transport of LLW.

     This report presents the results of a study of near-bottom
currents at the Atlantic ocean 3800 meter LLW disposal site.
Continuous measurements were obtained between 1984 and 1986 at
approximately 5, 100, 250, 400 and 1,000 meters above the
bottom.  The report describes deployments and recoveries of the
Aanderra current meters, data return and processing, data
analysis, and the potential for sediment transport of
radionuclides.

     The Agency invites all readers of this report to send any
comments or suggestions to Mr. David E. Janes, Director,
Analysis and Support Division, Office of Radiation Programs
(ANR-461), Environmental Protection Agency, Washington DC, 20460,
                                Richard J. Gpimond, Director
                                Office of Radiation Programs
                                   ill

-------
                            PREFACE
          This report presents the results of a two-year study
of deep ocean currents from May 1984 through May 1986 at the
U.S. 3800-m low-level radioactive waste disposal site near the
mouth of the Hudson Canyon.  The program objectives were to
describe the currents, including their source and variability,
and deduce from the data the potential for, and direction of,
transport of contaminants from the disposal area.  The study
was accomplished under an interagency agreement between the EPA
and the Minerals Management Service (MMS) which combined the
disposal area study with the on-going MMS study of physical
processes on the Mid-Atlantic slope and rise in order to gain
the maximum benefit from a larger data base.

          The results show that the currents in the disposal
area range in strength from a few to 62 cm sec~l and are
principally due to the presence of low frequency topographic
Rossby waves having periods of approximately two to four weeks.
The currents generally flow towards the southwest, in line with
the general topography of the mid-Atlantic region.

          The canyon acts to distort the southwest flow,
resulting in currents below the canyon rim which are aligned
with the canyon onshore-offshore axis.  The direction of
currents along the canyon axis appears to be determined by the
proximity of both the Gulf Stream and the Western Boundary
Undercurrent.  The currents also exhibit the Rossby wave
characteristic of bottom intensification and penetrate to the
floor of the Hudson Canyon.

          During the study period, several relatively strong
current events occurred in up- and down-channel directions of
the canyon, with the strongest current (62 cm/sec) directed
up-channel (westward).  The currents are sufficiently strong to
transport dissolved contaminants toward shallower water where
they could potentially be entrained by the Western Boundary
Undercurrent and transported toward Cape Hatteras.  On the other
hand, particulate matter that is fine enough to be transported
by the relatively strong currents at the disposal site would
probably be deposited further up-channel, as the influence of
the Rossby wave current amplitudes decrease toward the
continental rise.  Whether the particles could leave- the canyon
and reach the middle continental rise would require further study,

          This report is organized into five sections, as
follows: (I)  Introduction; (II) Current Measurement Program;
(III) Literature Review; (IV) Data Analysis; and (V)  Summary.
References are contained in Section VI.

-------
                        TABLE OF CONTENTS

                                                        Page

 FOREWORD                                                iii

 PREFACE                                                   v

 LIST OF FIGURES                                        viii

 LIST OF TABLES                                           xi

 I.    INTRODUCTION

      1.1.  Overview                                        1

 II.   CURRENT MEASUREMENT PROGRAM

      2.1  Mooring Design and Instrumentation              6

      2.2  Deployment,  Rotation and
                Retrieval Cruises                          8

      2.3  Data Return                                      8

      2.4  Data Processing                                14

 III.  LITERATURE REVIEW

      3.1  Introduction                                   16

      3.2  The Gulf Stream                                17

      3.3  Near Bottom  Currents                           18

 IV.   DATA  ANALYSIS

      4.1  Introduction                                   24

      4.2  Mean Currents                                   24

      4.3  Topographic  Rossby Waves  (TRWs)                 25

      4.4  Relationship to the  Gulf  Stream                 37

      4.5  Higher  Frequency Motions                        41

V.    SUMMARY                                              45

VI.   REFERENCES                                           48

VII. ACKNOWLEDGMENTS                                      51
                         vii

-------
                         LIST OF FIGURES
Figure 1-1
Figure 1-2



Figure 1-3


Figure 2-1


Figure 2-2



Figure 3-1



Figure 3-2



Figure 3-3



Figure 4-1


Figure 4-2a




Figure 4-2b
Locations  (circles) of the mooring deployments
and the 3800-m low-level radioactive waste
disposal site superimposed on the bathymetry
of the Hudson Canyon (Hanselman and Ryan,      2
1983) .

Location of the U.S. 3800-m low-level radio"
active waste disposal site in relation to the
MASAR program moorings.                         3

Location of the SEEP and Gulf Stream Vari-
ability programs affiliated with MASAR/EPA.     4

Design of Mooring I deployed at the 3800-m
low-level radioactive waste disposal site.      7

Time line of the deployment period and the
usable data from each instrument at the
designated nominal depth.                       9

Gulf Stream north wall separation distance
from the 3 MASAR mooring lines derived from
NOAA AAVHRR satellite infrared imagery.        19

Map of MASAR study area showing near-bottom
mean current vectors from historical and
MASAR data.                                    21

A comparison of mean velocities and variance
ellipses at 4000 m for the western North
Atlantic.                                      22

7-DLP temperatures (top) and currents (bottom)
by level on Mooring I.             -             26

Five month data record beginning May 29, 1984
of 40-HLP of temperatures (top) and currents
(bottom) by level on Mooring I at the 3800-m
low-level radioactive waste disposal site.     28

'Five month data record beginning September 2 6
1984 of 40-HLP of temperatures (top) and
currents (bottom) by level on Mooring I at
the 3800-m low-level radioactive waste
disposal site.                                 29
                            viii

-------
                     LIST OF FIGURES (CONT.)
Figure 4-2c
Figure 4-2d
Figure 4-2e
Figure 4-2f
Figure 4-3
Figure 4-4
Figure 4-5
Figure 4-6
Figure 4-7
Five month data record beginning January 1,
1985 of 40-HLP of temperatures (top) and
currents (bottom) by level on Mooring I at
the SSOO^m low-level radioactive waste
disposal site.
                                                              30
Five month data record beginning April 30, 1985
of 40-HLP of temperatures (top) and currents
(bottom) by level on Mooring I at the 3800-m
low-level radioactive waste disposal site.     31

Five month data record beginning August 28,
1985 of 40-HLP of temperatures (top) and
currents (bottom) by. level on Mooring I at the
3800-m low-level radioactive waste disposal
site.                                          32

Five month data record beginning January  1,
1986 of 40-HLP of temperatures (top) and
currentss (bottom) by level on Mooring I  at
the 3800-m low-level radioactive waste
disposal site.                                 33

Spectra of (a) kinetic energy for levels  I 2
(dashed) and I 4  (solid), and  (b) temperature
for instrument levels I 2 (solid) and I 5
(dashed).                                      35

Spectra of the (a) U component and  (b) V
component for I 2 (1 - solid) and I 4 or  I 5
(2 - dashed).                                  36

Coherence and phase differences between (a)
the U components  (1 - solid) and the V
components (2 - dashed), with axes rotated
30°, of I 2 and I 4, and (b) the temperature
at I 2'and I 4 or I 5.                         38

Coherence and phase differences between the
U (1 - solid) and V (2 - dashed)  components,
with axes rotated 30°, and the temperature at
level (a) I 2 and (b)  15.                      39

Rotary spectra for (a) II clockwise component
(1 - solid)  and anticlockwise component
(2 - dashed)  and  (b) I 4 clockwise component
(1 - solid)  and anticlockwise component
(2 - dashed).                                  42
                              ix

-------
                     LIST OF FIGURES  fCONT.l
Figure 4-8     (a) Kinetic energy spectra plotted in variance
               preserving form for I 1 for 171 days beginning
               November 2, 1985.                               43

-------
                          LIST  OF TABLES
                                                             Page


Table 2-1      Aanderaa Current Meter Specifications,
               Model RCM-5                                    10


Table 2-2      EPA Cruises To Deploy, Service and Recovery
               Mooring I Between February 1984 and May 1986.  11


Table 2-3      Location of EPA Mooring I from May 1984 To
               May 1986.                                      12


Table 2-4     ^Overall 2-year EPA Mooring I Instrument
              -Average Data Return By Parameter Measured
               At Each Level.                                 13


Table 4-1      40-HLP Statistics for EPA Mooring I
                                                              47

-------
                         I.   INTRODUCTION
1.1       Overview

          In  1976,  the  U.S.  Environmental  Protection  Agency
(EPA), Office  of Radiation  Programs,  initiated a survey  of the
Atlantic  3800-m  radioactive dumpsite, the  approximate  center of
which  is  indicated  by the  triangle  shown  in  Figure 1-1.   The
objective of this survey was to determine  the  potential for any
possible migration  of  radioactive  materials, dumped between 1957
and  1959, toward shore and/or productive fishing areas.   However
the  determination  of  currents  in the  area was  not completed.
Therefore  in  March,  1984,  the  EPA under  interagency  agreement
#DW14930778-01-0   with  the Minerals  Management Service    (MMS)
contracted  with  Science  Applications International  Corporation
(SAIC) to study the currents in and  around the 3800-m dumpsite
area,  as  SAIC was  already contracted to  MMS to  study  the Mid-
Atlantic  Slope and  Rise (MASAR)  dynamics west of the dumpsite to
a depth of 3000 m.

          This report  presents  the final results of the two year
current measurements effort,  from  May 1984  to May 1986,  in which
a  single EPA  mooring  was deployed  near  the  Hudson Canyon  to
determine the  circulation effects  on the  3800-m  dumpsite.   The
array  was  re-deployed  four times - in May  and  October  1984, and
April  and November 1985.  Recovery occurred  in May 1986.

          The measurement  of near-bottom currents  at the 3800-m
dumpsite  was  part  of  the  MASAR program.   The location  of the
3800-m dumpsite  current measurement  mooring in relation  to the
MASAR  moorings is shown in Figure 1-2.

          The MASAR program, as part of the MMS Outer Continental
Shelf  Environmental   Studies  (OCS)   Program,   focused  on  the
following:

     •   Eddies,  rings,  streamers,  and other Gulf  Stream (GS)
          related events

     •   The Western Boundary Undercurrent  (WBUC)

     •   Circulation   in   the   surface   layer  above  the  main
          thermocline  (less than 200 m)

     •   The shelf/slope front

          To study these dynamical processes, the MASAR principal
investigators utilized  hydrography, satellite imagery,  data from
affiliated  programs   in  the   area,  and  Eulerian   current
measurements.

          Several  of the  MASAR affiliated programs (Fig.  1-3)
also supported the dumpsite interpretative  effort.   These were:

-------
       70°40-    39'     38'
                                                              I   I
Figure 1-1.   Locations  (circles)  of  the mooring deployments  and
the 3800-m low-level radioactive waste disposal  site  superimposed
on  the bathymetry of  the  Hudson Canyon  (Hanselman  and  Ryan,
1983).   Positions 1,  2,  3, and  4 refer to  the May and  October
1984  and  April and  November 1985  mooring  deployment  positions,
respectively.   The. triangle represents  the center of  the 3800-m
disposal site.

-------
       78 W
76 W
74W
72 W
70 W
68
     42 N
     40 N
     38 N
     36 N
      34 N
                         200m
                         lOOOm
                         2000m
                                      - 38 N
                                      - 36 N
        78 W
76W
74W
 72 W
70W
                                        34 N
68 W
Figure  1-2.    Location of  the  U.S.  3800-m  low-level  radioactive
waste disposal  site in  relation  to  the  MASAR program  moorings.
The triangle  represents  mooring I,  at the 3800-m  site,  initially
deployed in May  1984;  the circles denote MASAR moorings  deployed
in February 1984 and the  stars  those  deployed  in  September  1985.

-------
                                                       - 36N
                                                         34N
Figure  1-3.   Location  of  the  SEEP and  Gulf Stream  Variability
programs  affiliated with  MASAR/EPA.    Circles  represent  current
meter  moorings,  boxes  represent  locations  of   Inverted   Echo
Sounders (IES).

-------
The  Warm-Core Rings Experiment  (WCRE),  funded  by NSF
and  conducted by the  Rosenstiel School of  Marine and
Atmospheric  Science (RSMAS).   The  satellite  infrared
imagery  part of  the program,provided much  of  the  10
year statistical  information  on  Warm-Core  Rings (WCRs)
used in this report.

The Gulf  Stream Meander Dynamics Program,  sponsored by
the National Science Foundation  (NSF) and the Office of
Naval  Research  (ONR).   This  program was designed  to
study the meandering processes of the Gulf Stream  (GS).

The  Shelf-Edge  Exchange Processes  (SEEP),  a  program
supported  by  the  Department  of  Energy  (DOE).    The
program's objectives were  to describe and quantify the
cross-shelf  transport  and  subsequent deposition on the
slope of organic carbon.

-------
                 II.  CURRENT MEASUREMENT PROGRAM


2.1       Mooring Design and Instrumentation

2.1.1     Mooring Design

          One  of  the major requirements  of  a successful mooring
design  is that  the instruments  attached to the  mooring remain
within   a   predetermined  depth  range  when  subjected  to
environmental  conditions.   The first  step  in the design process
is  to  estimate the range  of current   speeds  expected at the
deployment sites.

          Three velocity profiles (speed vs.  depth)  representing
low, mean,  and high speed, were  generated  from historical data.
The mean  speed profile  was used to determine the initial mooring
design  and the high and low profiles were applied to that design
as  a  check  on  potential  vertical  instrument excursions  and
mooring tilt at the  instrument attachment points.

          Next,  the  mooring  design  was  configured  with  the
instruments at their required depth and the mooring assumed to be
completely  vertical.   Sufficient  flotation  was  calculated and
configured  to  maintain a positive  tension  in the  supporting
cable;  a  simple summing of submerged weights provided the amount
of tension.   A safety  factor of two, between  working and yield
strengths, was used  for this mooring design.

          The  preliminary  design was then  subjected  to SAIC's
Static  Buoy  Analysis  Computer  Program  (SBAP).    The  program
computes  the  shape  of the  mooring cable  when  subjected  to
expected  ocean current forces.    Confirmation  of the computer
design  was  later analyzed by  comparing  the  initial  design with
the actual mooring performance.

          The  final  design used  at the  3800-m  dumpsite is shown
in Figure 2-1.  Attached to  the mooring  were five Aanderaa RCM-5
current  meters.     The meters  were nominally  spaced  to  be
approximately 5, 100, 250,  400, and 1000 m above the ocean floor.
The spacing was designed to allow  comparison of currents above,
at, and below  the  Hudson canyon  rim.  The  lower two instruments
were in the  canyon, the third  approximately  level with the rim,
and the upper  two situated approximately 100  and 700  m above the
rim.

          The required  flotation  to maintain the mooring upright j
was distributed along  the entire  mooring  length.   In  case  'of'
severance at some point along the mooring array, the distribution
of flotation would  insure  that the  remaining portion functioned
normally and would rise to the surface when released by comma'nd^

-------
               2898
               2930'
         OT
         erf
         w
         H
         W
         33-
         H
         CM
         W

         °
               3530
               3672
               3826
               3920
               3925

I4
                                            80
LEGEND

STROBE AND
RADIO BEACON

16 INCH
GLASS FLOAT

AANDERAA
  (CM)

PAIRED
RELEASES

ANCHOR
WHEEL
Figure 2-1.  Design of Mooring I  deployed at  the 3800-m
level  radioactive waste  disposal  site.

-------
2.1.2       Instrumentation

            The five Aanderaa RCM-5 current meters used on
 the mooring were supplied by EPA as government furnished
 equipment  (GFE).  After twelve months of field use they
 were replaced by newly calibrated RCM-5 current meters
 (also GPE).  All current meters were calibrated before
 and after  deployments in order to minimize data error.

            These meters are intended for use in deep ocean
 waters below the wave zone, are self-contained and record
 current speed, direction, temperature, and other parameters
 as configured by the user.

            Each current meter was attached to a swivel and
 shaft assembly mounted in-line with other mooring components
 The current speed was measured by a rotor.  Revolutions
 were counted and averaged, and then recorded on magnetic
 tape along with current direction from a magnetic compass
 and water  tempetature from a thermistor.  Specifications
 for the Aanderaa current meters are given in Table 2-1.


2.2        Deployment, Rotation, and Retrieval Cruises

           During the two-year study there were five cruises
 to deploy, service, and recover the EPA Mooring designated
 as "I".  The cruise dates, ships used and ports are shown
 in Table 2-2.

           Figure 1-1 shows the position of mooring "I"
 near the northeast side of the canyon after each deployment.
 The exact  locations, determined from Loran-C navigation,
 are given in table 2-3.  The first three deployments were
 within 500 - 600 m of each other.   The fourth deployment
 was situated along the southwest canyon wall to better
 define in-canyon circulation.   Although each mooring was
 identical in design, the differences in total water depth
 for each location resulted in different instrument depths.
2.3        Data Return

           Total usable data, from each instrument level,
 are shown in Table 2-4.  A time line of deployment periods
 and data obtained at each instrument level is shown in
 Figure 2-2.  The lowest semiannual data return occurred
 during the first 6-month deployment when older-designed     '
 current meters were used, resulting in a variety of failures',
 One meter flooded because of transducer failure.  Another
 meter experienced compass failure associated with in-situ
 low temperatures.  The compass functioned normally before
 and after deployment.  A third meter's battery failed.
 During the second deployment period, the top meter

-------
        INSTRUMENT

       LEVEL/NOM  DEPTH <«)
DEPLOYMENT PERIOD
I 1
I 2
I 3
I 4
I 5
2930
3530
3672
3826
3920

1 	 ; 	 j
1

L

.

L_
1 1 1 1 III
MJJflSOND
1984

1 	


~ J L j
r — 	 J
J L_


1 1 1 1 1 1 .1 1 ] 1 1
JFMflMJJflSOND
1985

i 	 1
j

i_ |

_[
1 1 1 1 1
J F M R 11
1986
                            KEY:
                                    I1  CURRENT  SPEED/DIRECTION


                                    1	J TEMPERflTURE
Figure 2-2.   Time line of the deployment period   and the  usable data from each instrument
at the designated nominal depth.

-------
Table 2-1.     Aanderaa  Current   Meter  Specifications,   Model
               RCM-5.
Current Speed       Sensor:  Rotor with magnetic coupling
                    Output: Revolutions counted electronically
                    Threshold:  2.5 cm/s
                    Maximum:  250 cm/s

Current Direction   Compass:  Magnetic; coupled to external vane
                              assembly
                    Resolution:  0.35 degree
                    Accuracy:  7.5 degrees

Temperature         Sensor:  Thermistor
                    Resolution:  0.1% of range
                    Accuracy:  0.05 degree C

Time Base           Source:  Quartz
                    Accuracy:  2 sec/day

Data Storage        600 feet of 1/4 inch reel to reel magnetic
                    tape
                    Capacity: 10,000 readings

Mechanical          Housing:  Cu Ni Si alloy, epoxy coated
                    Depth:   6000 meters
                    Weight:  26.6 kg in air
                             17.3 kg in water

Sampling            Interval:  30 sec - 180 min
                    Modes: one
                            10

-------
Table 2-2,
EPA  Cruises  To   Deploy,   Service  and  Recovery
Mooring I Between February 1984 And May 1986.
     SHIP
               DATES
   PORT(S)
U. of Del. vessel
  Cape  Henlopen

NOAA vessel
  Mt. Mitchell

NOAA vessel
  Atlantis IV
       27 - 31 May 1984


       26 - 28 October 1984


       29 April - 1 May 1985
SAIC contract vessel
  Seaward Explorer    29 Oct. - 11 Nov. 1985

EPA vessel
  Peter W. Anderson    5-10 May 1986
Lewes, DE

Sandy Hook, NJ
to Shinnecock, NY
WHOI

Morehead City, NC
to Norfolk, VA

Little Creek, VA
to Balitimore, MD
                              11

-------
Table 2-3.
Location Of EPA Mooring I From May 1984 To
May 1986.
     Deployment:
   No.       Date
           Location
          Coordinates
 Loran
  TD
Water
Depth
Moor
Top
    1   5/31-10/28/84
           37°49.2'N   25723.6  3925 m  2898 m
           70°36.2'W   42292.6
        10/28/84-5/1/85
           37'49.3'N
           70°35.7'W
25721.4  3905 m  2879 m
42293.8
    3   5/1-11/11/85
           37°49.2'N
           70'35.9'W
25722.7  3922 m  2895 m
42294.2
    4   11/11/85-5/10/86  37C49.5'N   25723.5  3922 m  2895 m
                          70°37.1'W   42294.0
    *NOTE:   Slight  variation  in total  water  depth following  each
            deployment  results  in instrument   depths.
                              12

-------
Table 2-4.     Overall  2-year EPA Mooring  I Instrument  Average
               Data Return By Parameter Measured At Each Level.
                   Nominal         Current
                    Depth          Speed &
Instrument          Level          Direction       Temperature
    II             2930 m            51%               51%

    12             3530 m            75%               75%

    13             3672 m            69%               69%

    14             3826 m            75%               93%

    15             3920 m            76%               75%
                             13

-------
flooded because  the input/output port design  (this  port is used
during the  pre-development check of the  current  meter to verify
proper operation)  did  not use an o-ring;  later designs corrected
this  problem.    This  first group of  meters was  replaced  with a
second group  of newer meters during  the 12th  month turnaround.
Subsequently,  the  overall data  return for all meters increased
from about 58% to better than 80%.


2.4       Data Processing

          The  data from the  four deployments were  generally of
good   quality   with   the  exceptions  noted   below.     After
transcription  to  nine-track  computer tape,  the  velocity  and
temperature time series  were processed and checked  for bad data
points using established  techniques  discussed  in  SAIC  (1987).
The time series were merged together by interpolating through the
gaps caused by the rotation  of the  mooring  at 6 month intervals.
The data  were then filtered  by a three-hour  low  pass  (3-HLP)
Lanzcos kernel  decimating to  intervals of  1 hour,  a forty-hour
low pass (40-HLP) Lanzcos kernel decimating to 6-hourly intervals
and a 7-day low pass (7-DLP)  Lanczos squared kernel decimating to
1 day intervals.   The  latter  filter was used  to remove energy at
periods less than 7 days so that low frequency Topographic Rossby
Wave  (TRW)  motions could be  presented  on two-year axes.   The
majority of  Rossby wave energy is  at periods longer than 7 days
(Thompson,  1977; Hamilton,  1984).   This  7-DLP  filter  was also
used extensively in the analysis of the MASAR data.

          The current  vector time series (see  Figs.  4-2a to -2f
in Section 4) were plotted with the axes rotated such that the V-
component was directed along  030'T and the orthogonal U-component
along  120°T.   Rotated axes  are designated  in the  data  record
identification label as  "R"  for  rotation  along with the adjacent
number "30" designating the degree of rotation, for all direction
data  in  this  report.   Axes rotation has resulted  in  the V-
component being  approximately parallel to  the general  trend of
the isobaths on the  continental rise NE  and  SW of  the Hudson
canyon and the U-component being in-line with the canyon axis.

          The  data recovered  from  instrument  level  I  4  in the
first and  second deployments  was  only partially complete._  The
low temperatures of the deep ocean caused the compass to seize in
this  instrument  and only  a  speed record  was obtained.   For the
first and second deployment  periods the record was reconstructed
by combining  the direction from the  nearest  instrument level, I
3, and I  5  when that at level I  3  malfunctioned, with the speec}
record from level I 4.  This analysis procedure has been shown to
produce a good approximation  of  direction in  the presence of low
frequency planetary wave  motions which are essentially columnar.
The  results  are  shown  in   Figures  4-2a  to  -2c   (Section  4).
Directions  so  constructed for level  I 4  should  not  be used for
relatively high  frequency inertial  or tidal  motions since there
may be  large changes  with depth.   Thus  only the  low frequency
                              14

-------
data for  the  first two deployments of  I  4 are used  in  the data
analysis section (Section 4) of this report.   The columnar motion
at low  frequencies was  checked  using  data  from deployments 3 and
4 with  the result that the phase differences  through the depth
array were not significantly different from zero.

          A  second   problem   with  the  data  concerned  the
calibration of the temperature data for deployments 1 and 2.  The
temperature sensors  for deployments  3  and 4 were  calibrated  by
Lament Doherty Geological Observatory and SAIC and are considered
accurate to the precision of the instrument.   Deployments 1 and 2
used  old  GFE  instruments  which  evidenced  calibration  drifts.
There were differences  in mean temperatures between the first and
second year deployments, but  variances  were  similar.   Therefore,
temperature records from deployments 1 and 2  were post calibrated
by  comparing  the  temperatures  recorded  by  the   older  GFE
instruments at the end of deployment 2 to those  recorded by the
newer replacement  instruments at  the beginning of deployment 3
since the  recovery and redeployment time difference  was  only   a
few hours.  After this correction, the mean temperature profile
for the first year was  very similar to  that  for the second year,
differing by about 0.02°C at maximum.
                            15

-------
                     III.  LITERATURE REVIEW
 3.1       Introduction

          A  review  of  the  oceanography  of  the  Mid  Atlantic
 continental  rise  was conducted as part  of  a previous study into
 the  deep  currents  at  the  3800-m  low-level  radioactive  waste
 dumpsite  (Hamilton,  1982).  Major experimental  studies up to 1981
 were  reviewed  and  discussed.    This  review  is  basically  an
 updating  of  the information  in Hamilton (1982) and the reader is
 referred  to  that document for background material.   The 3800-m
 dumpsite  is  about 50 km north of the mean historical position of
 the north wall  of the Gulf Stream at 71°W and therefore the Gulf
 Stream  proper  generally  dominates the  upper part of  the  water
 column  above the main thermocline.   The Gulf Stream and related
 warm-  and cold-core rings have  been the subject of  a number of
 studies in the  last  five years.

          Earlier current measurements on the continental rise of
 the Mid-Atlantic  Bight  have  shown that low frequency motions are
 dominated by  topographic Rossby  waves  (TRW)  (Thompson,  1977;
 Hogg,  1981;  Hamilton,  1984).  These  are planetary wave motions,
 with periods ranging from about a week  to  several months,  which
 are  characterized   by  bottom  intensification  (i.e.  current
 amplitudes  increase in magnitude  with  depth);  columnar motions
 (i.e.  zero phase  difference  between currents at different depths
 at   any  particular  wave   frequency)   and  the  direction  of
 propagation  is  such that shallow water  is  on the right (i.e. to
 the  west  at  the 3800-m  site) .    Horizontal wavelengths are of
 order  100-200  km   and  TRWs  in  the  Mid-Atlantic  Bight  show
 direction of energy  propagation consistent with generation by the
 Gulf Stream  (GS)  (Hogg, 1981).  The dynamics of TRW generation by
 the meandering  GS are not yet well understood.

          Mean  near-bottom flows on the continental  rise show a
 consistent westward  and southwestward drift following the isobath
 towards Cape Hatteras.    This is 'known  as  the Western Boundary
 Undercurrent  (WBUC)  and recent  measurements from the  MASAR and
 POLYMODE  programs are discussed below.

          The Gulf Stream leaves the continental margin and flows
 into deep water at Cape Hatteras, where it becomes a free flowing
 meandering jet.   Time-dependent wavelike,  lateral displacements
 or  meanders  of  the GS  are  observed  along  its  entire  path.
 Northeast  of  Gape  Hatteras,  these  meanders  show  considerable
 range  of  periodicities  (4-100 days  are dominant),  wavelengths
 (200-600  km)  and phase  speeds  (5-40  km  d"1) .   The  domin/ant
meanders  also  grow  downstream of  Hatteras  with a  tendency to
 exhibit a lengthening of period  and a decrease  in  phase speed.
 By convention,  a  meander  crest  (.trough)  is  defined to be  the
extreme   shoreward   or  northwards  (seaward  or  southwards)
displacement of a downstream propagating meander.
                              16

-------
3.2       The Gulf Stream

          Recent  studies  of  the  Gulf  Stream  have  involved  the
extensive use  of  SOFAR floats  (Richardson  et  al.f  1981;  Schmitz
et al., 1981;  Owens,  1984;  Shaw and Rossby,  1984).   SOFAR floats
follow  motions on  surfaces  of equal  pressure  (i.e., they  are
isobaric  floats)   and  thus  are   not  true followers  of  water
particles.   New  developments  in  float technology  have produced
floats  that more readily  follow  water particles on  constant
density  surfaces  (Rossby  et  al.,  1985a).   These  are known  as
"Rafos"  floats and some  preliminary results  from  these  devices
placed  in  the  Gulf Stream  off Cape  Hatteras  have  been reported
(Rossby et al., 1985b).   A number of Rafos  float tracks released
at a  nominal depth of 500 m  (corresponding to about 27 sigma-t)
in the  GS  off  Cape Hatteras have  remained  within  the stream for
distances  of over 2000 km.   Only a  few  transferences of floats
into  slope water to  the north of the GS  front occurred.   The
implication is that at thermocline depths, the GS is an effective
barrier to transfer of Sargasso  Sea water  into slope  water.   A
similar  conclusion was  arrived  from  an  isopycnal water  mass
analysis of Gulf  Stream 60 hydrographic sections by Bower et al.,
(1985).

          Shaw and Rossby (1984),  using SOFAR floats, discuss GS
trajectories   above  and  below the  thermocline.    Using  only
portions of the  float tracks  that correspond to GS water masses,
they  showed that the shallow  700-m  level  trajectories remained
within  the GS  and between Cape Hatteras  and the New England Sea
Mounts  (approximately 59°W).  Thus  fluid particles  above the main
thermocline  are   advected  along with the current  over distances
comparable  to  the  downstream  scale of  the  current  (i.e.,  the
distance between  meander  crests).   A  comparison of  the deep SOFAR
float trajectories  (1300  and  2000 m)  with the  shallow ones showed
that  a  continuous  current  does   not  exist  below  the  stream.
However the   deep  floats  do  show  intermittent   eastward  flow
related to the  surface  current  but it  is apparent  that water
parcels below  the  thermocline can  escape  from the stream both
northward  and   southward   anywhere  along  its  path.    The
intermittent deep GS  may be  formed  by the  acceleration of the
shallow stream east  of Cape  Hatteras.   Shaw and Rossby  (1984)
also  provide   direct  evidence  that  large  meanders  and  the
formation  of warm- and cold-core rings can be a direct influence
on  the 1300-2000-m currents.   This  adds substance to the ideas
that the meandering GS  is the source  of energy for  the ubiquitous
Topographic  Rossby  Wave motions   found   in  the   slope  region
(Luyten, 1977; Hamilton,  1984).

           Gulf Stream  meanders  have  been  studied  from AVHRR
thermal imagery  (Halliwell  and Mooers,  1983; Cornillon, 1986) and
from  data provided by an  array  of  bottom mounted inverse echo
sounders  (IBS) to sense the position of the main thermocline and
thus  infer the position  of the GS front (Watts and Johns, 1982;
Tracey  and Watts, 1986).   The  standard deviation of the envelope
of  GS  frontal  positions increases linearly from Cape Hatteras to
                               17

-------
71°W.   There  is a  relative minimum at  about 69°W  followed by
about  a  two-fold  increase  in  the  width  envelope  which  then
remains relatively  constant (standard deviation approximately 80
km) between 68° and 58°w  (Cornillon, 1986).

          Meander  propagation  speeds   and  wavelengths  vary
considerably.   Halliwell  and Mooers  (1983)  found,  from  an EOF
analyses of GS frontal maps, that there were two dominant meander
paths downstream  of Cape  Hatteras.   At periods greater than four
months the  first meander mode was a  standing  wave with nodes at
Cape Hatteras  and  about  at 38 °N,  68CW.   This accounts  for the
relative  minimum in  the  GS  path  statistics  found by Cornillon
(1986)   and others.   The second mode  resulted  from downstream
propagating meanders with periods of a few weeks to a few months.
The dominant wavelength and period averaged about 330 km and 45
days, respectively.   Phase speed was about  8  cm  sec"1 with some
amplitudes  exceeding 200  km.   These  meanders exhibited  a  wide
range  of  wavelengths,   periods  and, propagation  speeds,  while
undergoing exponential  growth in the first  900 km downstream of
Cape Hatteras.

          Halliwell  and Mooers  (1983)  showed  that there was  a
weak seasonal  dependency  in the path of the  stream.   Tracey and
Watts (1986), using  IES data showed that  the stream tended to be
displaced southward in winter which is also the period of highest
transport.   Recently,  Fu  et  al.   (1987)   showed the  seasonal
variation in sea-level difference across the GS front in the Mid-
Atlantic  from  satellite   altimeter  data,  which   showed  a  good
correlation  with   the  seasonal   variation  in   large-scale
wind/stress curl  (with  a  three  month lag).   The  intense winter
cooling over the northwestern Atlantic also plays a  role in the
intensification of the transport.  However the MASAR study showed
there is a great  deal of  inter-annual variability in  the path of
the stream.  For most of  the 1984-1985 KASAR  period the GS was
displaced north  of  its mean position.   This  is  illustrated by
Figure 3-1  (SAIG,  1987)  which shows the  distance  from the shelf
break  along the MASAR mooring  transect  (north, south and  a
transect  midway   between)  of  the mean  monthly  position of  GS
front.   Large amplitude fluctuations are shown  in 1982-1984 which
were not present  in  1985  when the stream  was closer,  on average,
to the slope.

3.3       Near Bottom Currents

          Near bottom  flows have  been  generally studied  with
current meters.   Hendry   (1982)  discusses  records from  a zonal
array of  4 moorings along  40°30'N between 56°  and  55 °W.   IJhe i
instruments were  1000 m  off the bottom  in 5000-m water depth.,'
The mean flows were  westward along  the  isobaths and were similar
to the westward mean flows measured along 70CW  in water depths o,f
less than 4000 m  (Luyten,  1977) .  The westward deep mean flpw at
55 °W is probably  part of  the Western  Boundary Undercurrent which
is  observed on   the upper  continental   rise as far  as  Cape
Hatteras.    It   was  more   difficult  to   relate the  current
                             18

-------
                  qAiJ.UUAWAAAiiAUAAWAAAiU4AAAWAAA,i:.Ai6N6j^
                   1982    I     1983     I     1984     I      1985
                                    Month/Year
Figure  3-1.  Gulf  Stream north wall separation distance from the 3 MASAR mooring lines
derived from NOAA  AVHRR satellite infrared imagery.  A, B,  C refer to separations from
the south, middle  and north lines, respectively.   (from.SAIC,  1987)

-------
fluctuations to  GS  meanders as has been done intermittently with
deep  float  trajectories (Owens, 1984).   Some of the temperature
perturbations   at   these   sites   could   be  from  Gulf  Stream
displacements or from the formation of cold- and warm-core rings.


          Observations  of near bottom currents west of 70 "W have
been considerably increased by the MASAR experiment  (SAIC, 1987).
Thus,  mean  flows  upstream  and  downstream  of  the  low-level
radioactive  waste  sites  have  some  degree  of  documentation.
Figure  3-2   shows  the  mean  currents derived  from near bottom
current meters  (generally 100 - 500 m off the seafloor) from the
MASAR  experiment (SAIC,  1987); the  SEEP experiment  (Aikman et
al.7  1987);  the  Rise  Array experiment  along 70°W and 69°30'W
(Luyten,  1977);  near  bottom measurements  along  36 °N  on  four
moorings  in  water depths from 1600 to 3700 m  off Cape Hatteras
(Casagrande, 1983) ;  records  500 and  1000 m off  the bottom from
five  current meters moored  under the GS off Cape Hatteras (Johns
and Watts, 1986); five  12-month long current meter records at 500
m and  1000 m off the  bottom in the same deep GS region  (Bane and
Watts,  1985) ;  the EPA  2800-m dumpsite  (the mean  of  four near-
bottom four month current records is shown, Hamilton (1982)); and
finally one  additional  record from Richardson's  (1977)  study of
the Western  Boundary  Undercurrent  southeast of  Cape Hatteras is
included.  Although this latter record is only two months long,
it was 100 m off the  bottom in 2600-m water depth, apparently in
the core of the WBUC.

          The  currents   are  sufficiently  consistent to  suggest
that  the  estimates of  the  mean are statistically significant.
The resulting pattern suggests that the strongest mean flows are
at the base  of  the continental slope.  However,  there  is also a
region between  the 3000 and 4000-m isobaths in  the vicinity of
700-72°W  where  the  westward along-isobath  flow  is  relatively
stronger  than  in the middle of the  rise.   This  region  is  just
north of the mean position of the GS front.  Mean flows under the
GS  (MASAR site  H and  the Rise Array  deeper  than  4000  m) tend to
be negligible or much more variable  in direction with perhaps a
northward trend.

           Figure 3-3  (Fofonoff and Hendry, 1985). shows the low
frequency variance ellipses of deep currents, near bottom or 4000
m whichever  is  shallower,   superimposed  on Richardson's (1983)
surface drifter  trajectories.  The  primary data  used  came  from
the  Rise Array  (Luyten,   1977),   the  Polymode   Local  Dynamics
Experiment (Owens et al., 1982), and  the  Polymode Arrays 1 and 2
(Schmitz,  1976;  1980).   At  55"W the kinetic energy  (proportional
to the  area of  the variance ellipses)  is clearly greatest under
the GS.   There is a marked decrease in variance  to the east at1
approximately 45°W  and  south to 32°N, and to  a lesser degree, to
the west  at  70CW.  There  is also a  tendency  towards  increasing
ellipticity of  the variance  ellipses to  the north  of  the  GS as
stronger bottom  slopes  are  encountered on  the  continental rise.
                             20

-------
                                                           NTS 100-600m oil bottom
                                                           I • • •  thin 6 month!

                                                             months or I o n a • r
Figure  3-2.    Map  of  MASAR  study  area  showing  near-bottom mean  current  vectors  from
historical and MASAR data.   (from SAIC, 1987)

-------
NJ
K)
                   50'
                        80-
                               75-
                                      TO
                                                                                           50-
                                                                                           4CT
                                                     VS?     0     0.5 mfs  °s
                                                    •/V;/ •;    	i    .  ' cr
                                                    •:- fiNsi •  s-e             •.".•"<»
                                                                                           JO-
                                                                                          35-
         Figure 3-3.   A  comparison of mean  velocities and variance  ellipses at 4000 m  for  the
         western Nortli Atlantic.  The approximate position of  the Gulf Stream is shown  by  surface
         drifter trajectories (Richardson,  1983).  The ellipse axes are proportional to rms speeds
         and are drawn to the same scale  as the mean velocity  vectors (from Fofonoff and Hendry,
         1985) .

-------
The major  axis  of these ellipses appear to  align  with the local
isobaths.

          In a number of these experiments, it has been suggested
that there  may  be a relationship,  possibly  indirect,  between GS
meander  motions and deep  currents  in  the  vicinity  of  the  GS.
There is some evidence that  the formation  and  passage of warm-
core  rings in  the  slope  sea  can  affect the  energy  levels of
deeper currents.   Louis et al.,  (1982)  detected strong shoreward
propagating topographic waves over  the Scotian slope and rise.
The current oscillations  occurred in bursts  of  4-5 cycles which
indicated  isolated  deterministic  forcing rather  than  a steady
stochastic  energy  flux.    Louis et  al.,  (1982)   also observed
Rossby waves  having baroclinic depth structures  on the slope and
upper  rise   and   predominantly  barotropic   at  the  slope-rise
junction.   One  of  the  events in the Scotian rise current meter
array was  related to the  formation of  a  small  warm-core eddy at
the crest  of  a  large GS meander due south  of the moored array.
The  wave  field  was successfully , modeled  as  being  due to an
isolated vortex by Louis and  Smith  (1982).  A description of the
kinematics and  dynamics of a mature warm-core ring  is provided by
Joyce  (1984).   He  shows that a large  warm-core  ring  can induce
perceptible warming as  deep as 4000 m but the velocity structure
is  confined to  the  upper  1000 m.   Similar  kinds  of  changes in
depth  structure  are also observed  in the  deep  MASAR current
records  from the  southern  Mid-Atlantic  Bight  (SAIC, 1987).
                             23

-------
                        IV.   DATA ANALYSIS
4.1       Introduction

          Motions  on  the lower continental rise are dominated by
Topographic  Rossby Waves  (TRWs),  which  are  ubiquitous  on  the
continental  rise  and  slope of  the Mid-Atlantic Bight  and have
been  reviewed   above.     The  Hudson  Canyon  3800-m  low-level
radioactive  waste disposal  site  mooring  is  north of  the mean
historical position of the cold wall or surface temperature front
of the GS at 70°W.  However, for about two thirds of the two year
mooring I deployment period the monthly mean position of the cold
wall was north of  mooring I (SAIC, 1987).   The mooring was about
30'  west  of  the position of  the  4000-m  Rise  Array  mooring
(Luyten, 1977; Thompson, 1977) deployed for  nine months in 1973
and  about  15' west  of   the  northern end of the  Inverted Echo
Sounder  (IBS)  line G  from 1983  though  1984  (Tracey  and Watts,
1986).

          A  sketch of the  Hudson Canyon  bathymetry  (Fig. 1-1),
taken  from  Hanselman  and  Ryan  (1983) ,  with  the four  mooring
deployment positions indicated, shows that the channel is about 3
km  wide,  steep  sided  with   walls  about 250   m  high, . and  a
relatively  flat  bottom.    About 10  km southeast of  the mooring
site, the  channel broadens and begins to  merge  into  the abyssal
plain at depths beyond 4000 m.

          The first three deployments were on the eastern side of
the  channel.   The fourth deployment was  deliberately  placed on
the western  side of  the  canyon  (about 2 km west  of the previous
locations)   so  as to determine if there were  any differences in
flow characteristics, particularly mean currents, between the two
sides of the channel.   Thus an attempt was made  to determine if
there was  a  possibility of  a  residual  circulation within  the
canyon consisting  of northwestward  flow on  one  side  and return
flow on the  other side.   This turns out not  to  be the case,  and
as discussed below,  the four  deployments may be  treated as  a
single mooring  at a  site  approximately  in  the  middle  of  the
channel.
4.2       Mean Currents

          Mean currents, of  approximately  3-5  cm sec"1,  at water
depths shallower than 4000 m along 70°W have been associated with
the edge of the WBUC (Thompson, 1977) ,  but mean currents in wafter
deeper  than 4000  m measured  by  the  rise  array showed  little
consistent westward drift.   In the historical compilation of liieah
currents, Figure 3-2, relatively  uniform westward drift  of a few
cm sec"1 was found on the upper part of the rise.

          Mooring  I   had  relatively  strong  along-isobath
(southwestward) flow above the canyon of about  4  cm sec"1 which
                               24

-------
 indicated that the site was within the WBUC.  Table 4-1 shows
 the statistics for currents measured during the complete two
 year deployment.  Mean velocity, in cm sec~l, and temperature,
 in °C, are listed for each instrument level, along with the
 variances (U1 denotes the deviation of the U-component of
 velocity from , mean temperature
 fluxes  and , and eddy temperature fluxes 
 and , where angle brackets <> denote that the mean of the
 quantity between them is calculated.  The data from levels
 I 1, I 3 and I 4 are broken into time segments due to instrument
 failures.  A comparison of the mean currents at I 2 and I 5 shows
 that the I 2 mean flow is directed across the canyon, while the
 I 5 mean flow, near the bottom, is directed up the canyon as
 might be expected.  Mean current speeds increase slightly with
 depth between I 2 and I 5, reaching a maximum at 100 m above
 the bottom at I 4.  However, a comparison between the first and
 fourth I.I deployments show that six- month means can vary
 considerably between periods; a reflection of long time scales
 present in TRW motions and indirect influences of the GS.  The
 more westward position of deployment 4 seems to make no
 significant difference in the direction of mean flows compared
 to earlier deployments.  It appears, therefore, that the mean
 flow is relatively uniform across the channel.  No residual
 internal horizontal circulation, due to canyon topography,
 is apparent in these measurements.

           The statistics in Table 4-1 also show that the
 lower 400 m of the water column, effectively the depth of the
 canyon, is virtually homogeneous.  Only very small (differences
 of 0.02°C)  vertical temperature gradients  T| exist between I 2
 and I 5.  This substantial, bottom-mixed layer, with a larger
 temperature variance at I 1 than at I 2 and below, is apparently
 caused by localized mixing and strong currents interacting with
 the canyon topography.  The maximum current speed observed was
 62 cm sec'1 and westward in direction at I 4.  It occurred
 during the first deployment (Fig. 4-1) on approximately Julian
 Day (JO) 195.  This maximum speed is similar to data from a
 rise array at 4000m depth (Luyten, 1977).


4.3        Topographic Rossby Waves (TRWs)

           Deep water flows on the continental rise are
 dominated by TRWs.  Hogg (1981) showed that nearly barotropic
 Rossby waves tend to dominate in water depths greater than
 4000 m and bottom-trapped TRWs are more prevalent on the middle
 and upper rises.  Therefore, it is expected that TRWs will
 dominate low frequency motions at periods longer than the Rossby
 wave cut-off period given by T =(Na)-1;  where N is the
 Brunt-Vaisala frequency of the lower water column and a is the
 bottom slope.  For the general topography of the lower rise,
 T is about 8-10 days.  Thus, low frequency current fluctuations
 have characteristic time scales longer than 10 days.
                          25

-------
                  o
                  u
                  Q
                  a
                  Q
                  U
                  M
                  Q
                  U
                  Q
                  O
                  n
                  K
                  0
                  U
                  in
 11
 11
 I •
 11
 11
 11
  n
  .11
                      •0
                      «
                      it
                     t.ei J
                                                                K
40 -
20 -
 0 -
-20 -
-40
               A
13
                                                                               3826m
                                                                          15   3920m
                                                               r^S*
                                    "y^rV^tV^t
                                                      ii

                                                      12  3530m
                               ;p  - ;
-------
Important characteristics of TRWs are that the wave velocities
are bottom intensified and columnar (i.e., no phase differences
with depth).  TRWs are discussed in more detail by
Thompson (1977); Hogg (1981) and Hamilton (1984).  The original
theory is presented by Rhines  (1970).

          An important question addressed by the Hudson Canyon
current data is the degree to which TRWs propagating westward,
along the rise, influence the currents in a narrow channel cut
into the seafloor; the channel having small dimensions compared
with TRWs which have horizontal wavelengths on the order of
100-200 km and vertical scales of 2-3 km.  The 7-DLP current data
presented in Figure 4-1 and the 40-HLP currents and temperature
data (Fig. 4-2 a-f) clearly show that the low-frequency motions
penetrate all the way to the floor of the canyon, increasing in
magnitude with depth.  Occasional exceptions to this pattern
are discussed below.

          The main influence of the canyon on TRW motions seems
to be an uncharacteristic relative rotation of current vectors
with depth.  This is shown more clearly by the principal axes
calculations given in Table 4-1.  The major and minor principal
axes effectively give the direction in which the variance of the
fluctuations is maximized and minimized, respectively.  There
is a counter clockwise rotation of the major axis of about 20°
between I 1 and I 2 and a clockwise rotation of about 85°
between I 2 and I 5.  U-component variances dominate within
the canyon, compared to almost equal U and V variances at I 1.

          The time series current data plots (Fig. 4-2 a-f)
show the high degree of coherence of both the currents and the
temperature fluctuations through the water column.  However,
there is considerable variability in the energy levels of the
fluctuations over the two-year period.  The period from June
to August, 1984 (JD 160-240) showed energetic, generally
westward directed currents.  It was also a period of large GS
meanders and the formation of a warm-core ring (WCR 84-E)
downstream (i.e. east) of the site.  The month of June had a
large southward displacement of the GS at about 69-68°W which
formed a cold-core ring.  This meander was present for most of
June.  A large warm-'core ring began forming near 67°W with the
GS displaced southward throughout the Mid-Atlantic Bight.   The
GS then interacts with newly formed WCR 84-E towards the end of
July.  This causes a reduction in size of the warm-core eddy
which then propagates westward, reaching 70°W in the first part
of September (JD 250) .  This is a period of relatively weak
currents (approximately 10 cm sec"1)  at mooring I.  At the
end of September (JD 263) the GS shifted northward due to
interaction with a cold-core ring (SAIC, 1987) east of Cape
Hatteras. (JD 263) .  No strong westward currents occurred in
the months of March-April, June and August 1985 when rings 84-G,
85-B and 85-F were near 70.5° W.  Direct relationships between
                             27

-------
oo
                           o
                           UJ
                           o
                           o
                           b)
                           Q
                           U
                           U4
                           Q
                           15
                           Ul
                           Q
                           O
                           U
                           Q
                             -40
                                                 JULIAN DAYS 1984
                                                 DAY 150 IS 5/29/1984
II  2930m



12  3530m



13  3672m



14  3826m



15  3920m



II  2930m

12  3530m

13  3672m

14  3826m

15  3920m
         Figure 4-2a.   Five month data record beginning May  29,  1984  of 40-HLP of temperatures (top)
         and currents  (bottom)   by  level  on Mooring  I at  the  3800-m low-level  radioactive waste
         disposal  site.  Note Current sticks are  referenced to 030° tiue  (vertical up position) in
         order  to coincide  with the  orientation  of isobaths  Northeast and Southwest  of the Hudson
         Canyon and  are designated by R30 on the plots.

-------
KJ
                           O
                           8
                           O
                           U
                           Q
                           O
                           M
                           Q
                           o
                           w
                           a
                           U
                           w
                           U)
                           X
          -x-/ V^-^'

                                                    II  2930m
                                                                                  12   3530m
                                                    13   3672m
                                                                                      3826m
                                                     15   3920m


 40
 20-
 0-
-20-J
-40
II

12

13

14

15
                                                 JULIAN DAYS 198«
                                                 DAY 270 IS 9/26/1981
2930m

3530m

3672m

3826m

3920m
         Figure 4-2b.   Five month data record beginning September  26,  1984  of  40-HLP of temperatures
         (top)  and currents (bottom)  by level on Mooring I at the  3800-m  low-level  radioactive waste
         disposal site.  Note  Current sticks are referenced  to 030° true  (vertical  up position)  in
         order  to coincide  with  the  orientation  of isobaths Northeast  and Southwest  of  the  Hudson
         Canyon and are designated by  R30  on the plots.   There  is  approximately  a  one month overlap
         from Figure 4-2a.

-------
OJ
o
                             o
                             u
                             Q
O
Ul
Q
                             8
                             O
                             o
                             g
                                                      II
                                              ^^7-
                                                                       A,
                                                 ~\ __ A
                                                                                       3530m

                                                      13
                                                                                   14   3826m
                                                      15  3920m
                                                                                   II
                                     TT

                                                                                —  12
                                40
                                20-
                                 0
                               -40
                                                                                   14
                                                          2930m

                                                          3530m

                                                          3672m

                                                          3826m

                                                          3920m
                                                    JULIAN DAYS 1984
                                                    DAY 1 IS 1/1/1985
         Figure 4-2c.   Five month data  record beginning  January 1, 1985 of  40-HLP of temperatures
         (top) and currents (bottom)  by level  on Mooring  I at the 3800-m low-level radioactive waste
         disposal site.   Note  Current sticks  are  referenced to  030° true  (vertical up position) in
         order to coincide  with the  orientation  of isobaths Northeast and Southwest  of the Hudson
         Canyon and are designated by R30 on the plots.   There  is approximately a one month overlap
         from Figure 4-2b.

-------
u>
                          u

                          a
                           o
                           o
o
o
(3


(J
o
Q


O
o
Q
                          O
                          o
                          (X
II  2930m



12  3530m



13  3672m



14  3826m



15  3920m



II  2930m

12  3530m

13  3672m

14  3826m

15  3920m
                                                JULIAN DAYS 1984
                                                DAY 120 IS 4/30/1985
         Figure  4-2d.   Five  month data  record beginning April  30,  1985 of  40-HLP of temperatures
         (top) and  currents  (bottom)  by level on Mooring I at the 3800-m low-level  radioactive waste
         disposal  site.   Note Current  sticks are referenced to  030°  true (vertical up position) in
         order  to coincide with  the  orientation of  isobaths Northeast and  Southwest of the Hudson
         Canyon  and are  designated by R30 on  the plots.   There is approximately a  one month overlap
         from Figure 4-2c.

-------
uo
K)
                          O
                          I.I
                          0
                          O
                          Ul
                          O
                          111
                          O
                          UJ
                          Q
                          a
                          Q
                          U
                          10
                          U)
                              0-
                            -20-
                            -40
                                              — ^
II  2930m
12  3530m
                                                                                 13  3672m
14  3826m
15  3920m
                                                                                 II  2930m

                                                                                 12  3530m

                                                                                 13  3672m

                                                                                 14  3826m
                                                       no   jn  uo   >M
                                                JULIAN DAYS 1984
                                                DAY 240 IS 8/28/1985
        Figure  4-2e.   Five month  data record  beginning August  28,  1985 of 40-HLP  of  temperatures
        (top) and currents  (bottom)  by level on Mooring I at the 3800-m  low-level  radioactive waste
        disposal site.   Note Current  sticks are referenced  to  030°  true  (vertical  up  position)  in
        order to  coincide with the  orientation of  isobaths Northeast  and Southwest of  the  Hudson
        Canyon  and are designated  by R30 on  the plots.   There is approximately a  one month overlap
        from Figure 4-2d.

-------
00
                           O
                           M
                           Q
0

O

Q
                           O
                           LI
                           a
                           o
                           U)
                                                                                  II   2930m
                                                      12  3530m
                                                                                  13   3672m
                                                      14  3826m
                                                                                  15   3920m
                           u
                           t-1
                           T.
                           O
 40-
 20-

 0-
-20-
-40
II

12

13

14

15
                                                  JULIAN DAYS 1984
                                                  DAY 1 IS 1/1/1986
                                                          2930m

                                                          3530m

                                                          3672m

                                                          3826m

                                                          3920m
         Figure 4-2f.   Five month data  record beginning January  1,  1986 of 40-HLP  of temperatures
         (top)  and currents (bottom)  by level on Mooring I at the  3800-m  low-level  radioactive waste
         disposal site.  Note  Current sticks are  referenced  to 030°  true  (vertical  up position)  in
         order to coincide  with the  orientation  of isobaths Northeast  and Southwest  of  the Hudson
         Canyon and are designated by  R30  on the  plots.  There is approximately a  one month overlap
         from Figure 4-2e.

-------
the position of  the  GS  cold wall,  the presence of rings, and the
character  of the  flow on  the lower  rise  are not  obvious from
these  data,  even   though  subtle  connections  involving  the
generation and propagations of TRWs and the GS are thought to be
dynamically important.

          Compared  to  the  June  to  August,  1984 high current
period,  JDs  235-335,  1985  (days  600  -   700,   Fig.   4-1)  had
relatively  short  period,   weak   fluctuations  (<10  cm  sec'1)
exhibiting northward  flow along the axis of the canyon.  However,
in contrast  to  these relatively low-energy  periods,  the last 5-
month deployment showed a strong  eastward event (JD 10-30, 1986;
Fig.  4-2f)  at  I 1  which exceeds  the velocities of  the deeper
instruments.   A similar strong  eastward event,  accompanied  by
anticyclonic rotation of the current vectors and a sharp increase
in  temperature,  occurred  at  the   beginning  of  the  second
deployment  at  about  JD 310,  1985  (Fig. 4-2b).   Unfortunately,
current  measurements at  I  1 were  not obtained  for  this event.
The unusual nature of these two strong eastward events along with
the presence of  warmer water may indicate a more direct  influence
of  the  deep  GS,   though  this interpretation is  speculative.
Examination  of   the  satellite  imagery  for  these  two  dates show
that  on  both  occasions a GS meander  crest  is  over  the  site and
thus  a  direct  GS  effect  on the  flow may be a  reasonable
explanation of these events.

          The kinetic energy spectra, in variance preserving form
for I  2  and 14,  are shown in Figure  4-3.   Variance preserving
spectra  plots   give  a  better  feel  for  the  distribution  of
fluctuation energy  or variance with  frequency than conventional
log-log  spectra plots,   since  equal  areas  under  the  curve
represent  equal  amounts  of variance.   The  fluctuating kinetic
energy is  greater  at I 4  than at I  2,  at  all frequencies along
with a sharp decrease in energy at I  2  at  about 0.15 cycles per
day (cpd) which  corresponds to the Rossby wave cut-off frequency.
This  increase  in   energy  with  depth  is   consistent  with  TRW
dynamics despite the  existence of  the canyon.   The 18-month time
series  used  for  these  calculations  resolves  the  major  low
frequency peaks  at 0.03,  0.06  and 0.075 cpd,  (33,  16  and 13 day
periods, respectively).   These frequencies  have  been associated
with  Rossby  waves,  and GS  meanders  in previous  studies (Hogg,
1981).   A feature  of Figure  4-3  is  the occurrence  of  spectral
peaks at frequencies greater than the cut-off for currents within
the canyon (I 4) .    These peaks  occur at  discrete  frequencies
which  seem to be  harmonics of major  low-frequency  peaks.   An
examination of the U and V component  spectra  (>  0.15  cpd)  (Fig.
4-4)   shows  that  the  higher   frequency  harmonics  are  more
associated with the  U than V component at I  4.   The U component1
also shows  increased variance with  respect  to the  V component.
Thus, the currents are being channelized and distorted within the
canyon so that energy is transferred to  the  higher harmonics as
the Rossby waves propagate over this feature.   There is no visual
indication from 40-HLP data (Fig. 4-2) that distinct wave motions
associated with the  channel (i.e., channel modes)  are being set
                           34

-------
Ul
                          KINETIC ENERGY SPECTRA
                                CYCLES/DAY
SPECTRA
                                                                              CYCLES/DAY
                                 (a)
(b)
          Figure 4-3.  Spectra  of (a)  kinetic energy  for levels I 2 (2-dashed) and I 4 (1-solid)   and (b)
          temperature for  instrument levels  I  2 U-solid) and  I  5 (2-dashed).   The time series length is
          548 days beginning  November 2, 1984.

-------
LO

CD
                                SPECTRA
                  u
                  o
                  <

                  >
                               CYCLES/DAY
 SPECTRA
CYCLES/DAY
                                  (a)
   (b)
          Figure 4r4_-  Spectra  of the (a)  U component and  (b)  V component for I 2 (1 - solid) and  I  4

          or I 5 (2 - dashed).   The time series length  is  548  days beginning November 2, 1984.

-------
up  at  frequencies  greater than  0.1  cpd.    This  artifact  of
spectral  analysis  is similar to  the harmonics generated  by the
distortion  of  a  semi-diurnal   tidal  wave propagating  into  a
shoaling estuary.

          Unlike  the current velocity  spectra,  the  temperature
spectra  show  decreasing variance  with  increasing  depth   (Fig.
4-3) .     The  current  and  temperature  fluctuations  at  low
frequencies are  also highly coherent and in phase throughout the
depth of the mooring  (Fig. 4-5).  Temperature and the U component
of current  are also correlated  with velocity leading temperature
by  about  60°   at  all   frequencies  below  0.1 cpd  (Fig.  4-6).
Usually  for TRWs,  temperature  and  currents  are in  quadrature
(i.e. 90e  phase difference) since  waves do not  transport heat.
However,  in the case of waves  in the canyon,  there is a small,
down-canyon flux of heat most clearly  evident from  the velocity
and temperature  records  at  I 3,  I 4, and I 5.   This implies that
the  distortion  of the  TRW wave  motions  by  the canyon  causes
horizontal  mixing of  temperature ,or pollutants  which  causes  a
diffusive flux against the mean up-canyon currents.

          Table  4-1  contains  the  40-HLP  statistics  for  the
longest  combined  records  at  each  instrument.   Of  particular
interest  are  the  Reynolds Stresses    and the mean and
fluctuating components  of the heat flux.   The Reynolds Stress is
quite variable with  the four,  six-month  deployments,  but above
the  canyon  [U'V] is positive,  which is  consistent with Rossby
waves propagating  westward  and onshore.   Within  the canyon, the
momentum flux can change sign.   This seems to be the case for the
relatively  quiescent 1985 period  (compare I 2  and I  5) .  However
since the motions  are essentially in phase between I  2 and I 5,
the  change  in  momentum  flux must be related to the distortion of
the  wave  motions by the canyon.  Again the large variability in
all  the calculated statistics  is noted  even though the averaging
period is six  months to one year.   The eddy fluxes  are an order
of magnitude  smaller than  the  mean advective  heat  fluxes.   The
eddy  fluxes oppose the  advective fluxes.  The T1 variances and
eddy  fluxes at  the  upper   level  (I  1)   exceed  that  of  2800-m
dumpsite  (Hamilton,  1984)   and  are comparable  within  the canyon
indicating that the upslope and along slope, temperature gradients
are  relatively  substantial  for  the 4000-m  depth,  despite the
300-400 m deep bottom mixed layer.


4.4       Relationship to the Gulf Stream

          The  relationship   between  the  Gulf Stream and  deep
current   motions   under  and  north  of  the  stream  is  not
straightforward.  Further  theoretical  developments  are  needed
before the  mechanisms of GS meandering  and birth  of  rings, which
are thought to be the energy sources of deep Rossby wave motions,
are better  understood.   The relationship  of deep  currents to the
GS is quite subtle and only indirect connections can be suggested
in this repoirfe.
                            37

-------
00
00
                    LJ
                    I
                    OL
                    -I8B-
                                  107
                              CYCLE'S/PAY
                                                                   1M
                                                                             CYCLES/DAY
                                 (a)
(b)
         Figure  4-5_.   Coherence  and phase differences  between (a) the  U components  (1 -  solid)  and
         the  V  components  (2  -  dashed),  with  axes rotated  30",  of  I  2  and I  4, and  (b)  the
         temperatui=e  at I 2 and  I 4 or I 5.   The time series length is 171 days beginning November
         11, 1985.

-------
OJ
VO
                    tu
                    o
                    2
                    LU
                    tt
                    LJ
                    I
                    O
                    o
                    IM
                               CYCLES/DAY
LJ
O

LJ
o:
LJ

O
           CYCLES/DAY
                                (a)
        Figure 4-6,  Coherence and phase differences between the U  (1  -  solid)  and V  (2 -  dashed)
        components, with axes  rotated  30°,  and temperature at  level  (a) I 2 and  (b) 15.  The time
        series length is 548 days beginning November 2,  1984.

-------
          There is some evidence that the path of the GS
influenced the TRW energy levels at mooring I and at MASAR
sites D, E, G and H (SAIC, 1987).  During the first four months
of this study, the GS followed  its usual path, leaving the slope
at Cape Hatteras and remaining  south of the 4000-m isobath at
70°W.  This was also a period of intense meander activity and
formation of a cold-core ring and a large warm-core ring,
downstream of the site, between 70° and 65°W.  Energy levels
at mooring I were larger than in subsequent periods.  Since
TRWs propagate westward along the bathymetry (Hogg, 1981), it
likely that the formation of a  large warm-core ring, which
underwent an intense interaction with a GS meander in July 1984,
was partly responsible for the  increased wave energy.  Between
October 1984 and December 1985  (JD 270-720; Fig. 4-1) the GS
was further north of its historical position (Figure 3-1).  In
fact, the GS front was often north of the 3800-m dumpsite during
this period.  TRW fluctuations were weaker and of higher in
frequency than the earlier period.  Fewer large warm-core rings
were generated in 1985 wes.t of  65°W.  The propagation of three
rings into the Hudson Canyon region in the spring and summer of
1985 (84-G, 85-B and 85-F) appeared to have minimal effect, if
any, on the currents at mooring I.  No bursts of energetic,
low frequency current fluctuations were observed.  Meanders
of the GS were also less frequent, between 70°and 65°W, during
most of 1985.

          From December 1985 to the end of this study in
May 1986, TRW energy levels increased.  The increase seems to
correspond to the GS shifting back to a more normal path at the
beginning of December.  However, a very large southward meander
occurred near 68°W during December with the possible formation
of cold-core rings, similar to  the meander and cold-core ring
formation in June 1984.  Thus it is not clear which event, the
GS shift or the large meander, was responsible for the
increased TRW activity.

          Apart from TRWs, there appeared to be two events
that could be related to the GS proper.  Between JD 312 and
322 (1984) and between JD 14 and 30 (1986), see Figs. 4-2b
and 4-2f, substantial eastward  flows occurred.  The down-canyon
components were accompanied by relatively large increases in
temperature.  Imagery -shows large meander crests moving over
the site at both times. It seems reasonable to attribute both
events to intermittent deep GS  flow causing an acceleration in
lower layer currents, as discussed by Shaw and Rossby (1984).
This is substantiated in the second event since maximum
velocities occurred at I 1 rather than near bottom, as is more
usual for TRWs.  The I 1 record was not available for the first
event.

          GS thermocline maps, derived from an IES array
(Tracey and Watts, 1986) , clearly show a large meander
propagating eastward through the array between the middle

-------
of  September and  the end  of  November 1984  (Tracey,  Cronin and
Watts,  1985).   The crest arrives  at the  site 26-30 November, JD
270-275  about 20  days  prior to the warm  eastward flow event at
mooring  I.   The GS  remains in a  northern  position close to the
slope  for most of December.   Thus,  even  though the correlation
with GS meander crest events is not  exact,  the assumption that GS
caused the eastward  flow events at 4000 m seems possible.


4.5       Higher Frequency  Motions

          Near-inertial  internal   wave  motions  are  generated
primarily by changing wind  forcing  at the ocean surface such as
due  to the  passage  of storms  and fronts.   The waves propagate
down  below  the   wind-mixed layer  in  the  form  of  clockwise
horizontal  rotating  currents  (looking   down  above)  with  a
periodicity  close  to f,  the Coriolis parameter.   These motions
are  intermittent and thus the  spectra has  broad  peak  near f.  U
and V  component amplitudes  are  almost equal,  with V leading U by
90°  consistent with a  constant  current   vector  rotating  at  a
frequency  near  f.    Inert ial  currents  were  shown  to be the
dominant higher  frequency motions at  the  2800-m dumpsite on the
upper  rise  (Hamilton,  1984).    Maximum  speeds due to inertial
oscillations  approached  10 cm sec"1  on   occasion with average
amplitudes  3-4  cm sec"1.    The lower Hudson Canyon  dumpsite is
1000 m deeper than the 2800-m  site  and since surface layer wind
forcing  is  the  source  of inertial currents,  it is expected that
near-inertial  internal waves   would  be more  attenuated  at the
greater depth.  Figures 4-7  a & b  show the  rotary spectra for I 1
and  I  4  for the  last  six-month  deployment  period.    Linear
frequency  axes have been  used so  as to  emphasize  the  higher
frequency  inertial and tidal  motions.  At I 1,  a fairly  sharp
inertial peak is  observed in the clockwise rotating component at
about 1.3 cpd.  A  semi-diurnal  (M2)  spectral peak is also present
at 1.95 cpd.   At  I  4 both inertial and semi-diurnal tidal peaks
have become broader.  The M2 tide  has similar energy to that at I
1 though  the anticlockwise  rotary component  is attenuated.   The
inertial peak within the canyon is  not as energetic as  at  I l.
The  local  inertial  frequency,   f,  is  1.23  cpd  for  37°50'N
latitude.  Thus, the  inertial peak at I 1 is about 6% higher than
f as is expected for propagating near-inertial internal waves.

          Within the canyon however,  the  peak splits  into two
with the lower frequency peak being less than f (about 1.20 cpd).
The  reason  for this  are  not clear,  but the  implication is  that
the  mean  flow  field  contains   negative  (cyclonic)   relative
vorticity which  can  support sub-inertial  internal waves.   The
implication  is  that  the  mean  currents  have horizontal  shear,
possibly  due  to   the  frictional  effects  of  the canyon  walls
(Mooers, 1975).

          Figure 4-8a shows the kinetic  energy spectra  for the
last deployment of I 1 plotted  in variance preserving form, where
equal areas under  the curve represent  equal amounts of variance.
                            41

-------
N)
                         ROTARY SPECTRA
            Q
            CL
            O
            X
            LU
            O
<
>
                           I          2

                           CYCLES/DAY
                            (a)
                                                           ROTARY  SPECTRA
                                                              I          2

                                                             CYCLES/DAY
                                                                 (b)
         Figure  4-7.   Rotary • spectra for (a)  I 1 clockwise component (1 - solid)  and anticlockwise
         component  (2 - dashed)  and (b) I  4  clockwise  component  (1 - solid)  and anticlockwise
         component  (2. - dashed).  The time series  length is 171 days beginning November 11,  1985.

-------
          KINETIC  ENERGY  SPECTRA
                  CYCLES/DAY
                    (a)
                                                   180
                                                  -180
I          2

CYCLES/DAY
    (b)
Figure 4-8.   (a)  Kinetic energy spectra plotted in variance preserving form for I 1 for 171
days beginning November 2, 1985.  (b) Coherence of inertial and tidal motions for the clock
wise rotary  components at I 1 and I 4 for 171 days beginning November 2, 1985.

-------
It  can  be seen  that inertial  and semi-diurnal tidal  peaks are
quite  prominent  at  frequencies   greater  than  1   cpd  but  they
account for only  a  small  proportion of the total energy which is
dominated by  the low frequency TRW wave motions.   The amplitude
of  the  M2 tidal currents  is approximately  1  cm   sec~^  and the
inertial  currents  2-3  cm  see"*.    This  downward  propagating
internal wave  kinetic energy is expected to  be dispersed by the
canyon boundary layers and thus may contribute to mixing.

          The coherence of inertial and tidal motions is shown in
Figure 4-8b for  the clockwise rotary components at I  1 and I 4.
The tidal records show a high  degree  of  coherence, the inertial
currents less so.  The phase propagation is vertically upwards at
both frequencies as is expected for downward propagating internal
wave packets.  The  60°  to 90" phase differences imply a vertical
phase velocity of  about  3 cm  sec"-1-  which is  comparable  to the
values calculated at the 2800-m dumpsite (Hamilton, 1984).
                            44

-------
                           V.  SUMMARY


          The near-bottom  current and temperature  fields  at the
3800-m   low-level   radioactive   waste  dump   site  have  been
investigated using  data  from a two-year-long mooring  (May 1984-
May 1986) deployed  in  the  lower Hudson Canyon.   Aanderaa current
meters were nominally  placed  at about  5,  100,  250,  400,  and 1000
m off the bottom.   The upper two  instruments were  above the lip
of the canyon,  the  middle  instrument approximately  at the canyon
lip and the lower two  instruments within the canyon.

          The  dominant motions were  low  frequency topographic
Rossby waves with characteristic  periods  of 33,  16, and 13 days.
Two-year-long records were needed to resolve the energetic longer
period   motions.    The   wave  motions   were   generally  bottom
intensified, columnar, and penetrated to the floor of the canyon.
The effect of the canyon primarily was to distort the essentially
linear waves below  the lip,  introducing some non-linearity which
introduced spectral  peaks  at frequencies higher than  the Rossby
wave cut-off  (about 8 days).   The average  kinetic  energy of the
Rossby wave  motions  varied  by more than  a  factor of  two with
deployment period and  seemed to have a subtle relationship to the
position  of  the  Gulf Stream.    When  the  GS  was  north  of its
historical mean  position during most of 1985,  energy levels were
approximately 40-50%  lower than in  either  1984  or  1986  when the
GS  followed  its more normal path after leaving Cape Hatteras.
However,   the higher  energy  levels  were  also  associated  with
higher degrees  of  GS meander  activity  downstream   of the site.
There were two  strong eastward  flow events  accompanied by warmer
water that may be associated with GS meander crests overlying the
site and could be examples of the intermittent deep GS flows that
have been observed with SOFAR floats (Shaw and Rossby, 1984).

          Mean currents above the canyon for the two-year period
were substantial  (approximately 3.5  cm sec"1 at 230° True (T)  at
400 m  from the  bottom)  and  directed  to the  west.   Within the
canyon, mean  flows   (approximately 5 cm sec"1 at 100  m  from the
bottom) tended to be directed more upslope  (approximately 240° T)
and there was no evidence of any residualrcirculation internal to
the channel.  The westward mean flows suggest that the site is on
the southern edge of Western Boundary Undercurrent   (WBUC).  Mean
temperatures were essentially uniform  throughout the lower 400 m
of the water column,  implying that the interaction  of the Rossby
waves  with  the  canyon topography generate  enough  mixing  to
homogenize the density structure.

          Higher frequency motions are weak and consist of near-
inertial and semi-diurnal  internal waves with  amplitudes  of 2-3
and 1 cm sec"1, respectively.

          The  implications  for  disposal  of  pollutants  is  as
follows.  Pollutants dissolved in the water should be subject to a
mixing through the boundary layer which is greater than the depth
                             45

-------
of the canyon.  Maximum  current  speeds measured  at 5 m and 100 m
from the  bottom are 45  cm sec"1 and 62 cm  sec"1, respectively.
Mean  flows tend  to be  up  the  canyon towards  the  upper  slope
within  the canyon  and  westward above the  canyon.   Therefore,
whether a  constituent  is  mixed  above the canyon  lip or remains
within the  canyon,  it  will eventually end up in the WBUC and be
transported westward towards Cape Hatteras,  similar  to  the fate
of dissolved constituents at the 2800-m site on the upper rise.

          During  the  two-year  study,  there  were  two  strong
(approximately  50  cm  sec~i)  eastward  events  which have  been
attributed  to  the Gulf  Stream.    Such  flows  would  transport
pollutants eastward (i.e. down-canyon) towards deeper water.  The
deep  Gulf  Stream  is  an  intermittent  feature  and thus  water
particle trajectories, once  out  of  the Gulf  Stream are likely to
return northwards to the WBUC.

          Once  dissolved,  pollutants reaching the Cape  Hatteras
region via the WBUC, could be entrained into the deep GS which is
more prevalent  there  (Watts  and Johns,  1982)  and then  returned
back to the WBUC downstream or further east.   Alternatively, they
could remain within the  WBUC and continue  southwards towards the
South  Atlantic.    Water  mass   analyses   implemented with  some
current meter  data have traced  the WBUC,  from its  formation by
Denmark Strait overflow,  to as far south as the continental slope
off the Bahamas (Fine and Molinari,  1987).

          Particulate matter, if sufficiently fine grained,  has a
greater propensity  for  sediment transport,  due to  the  stronger
maximum currents observed, in the lower Hudson Canyon than at the
2800-m  site (60  cm sec"1  versus  40 cm  sec"1) .    Maximum near
bottom flows were  directed both  up-and down-channel but with the
strongest  event  (JD  180-200, 1984)  being  directed up-channel.
Therefore,  the long  period  fluctuations of  the Rossby  waves
combined with the mean currents near the canyon bottom would move
fine-grain sediments towards  shallower water.   Since Rossby wave
current amplitudes  decrease  towards the  slope  region,   the  net
result   is  probably   deposition   of  any  sediment  possibly
contaminated  with  radionuclides   in shallower   water  to  the
northwest  of the  site.   Whether sediment particles  would  leave
the  canyon  and  make   it  to the middle   continental   rise
(approximately  3000-m  depth)   would  require  some   model
calculations to provide  guidance.    The occasional deep  GS  event
also has enough energy to transport sediment eastward into deeper
water.
                               46

-------
                                                                                     Table 4-1
                                                                       40-III.P Statistics for  EPA Mooring I
                                                                           Coordinate Axis Rotated 030°T
*>
Hctcr
11
U
12
13
13
14
J4
15
Period
6/14/ to
10/23/84
11/15/85
to 5/4/86
11/2/84
to 5/4/86
6/14/86
to 3/25/85
5/6/ to
H/5/85
6/14/84
to 3/01/85
5/6/85
to 5/4/86
11/2/84
to 5/4/86
cm sec
-8.12
-1.67
-1.57
-4.24
-1.91
-5.31
-3.64
-3.44
cm sue °C
-8.86 2.57
-1.23 2.58
-3.14 2.17
-4.57 2.17
-2.35 2.15
-4.82 2.17
-1.67 2.14
-2.05 2.15

56.81
46.92
23.54
73.22
17.70
104.60
40.84
47.33

65.37
53.87
37.60
44.99
30.08
59.59
31.36
39.33

29.58
35.42
4.10
23.79
-2.46
38.03
-0.24
-4.21
 
-20.
- 4.
- 3.
- 9.
- 4.
87 -22.77
31 - 3.17
41 - 6.«1
18 - 9.92
*
11 -.5-05
-11.52 -.10.46
- 7.
- 7.
79 - 3.57
40 - 4;4l

0.0139
0.0276
0.00293
0.00198
O.OOOH8
0.0020
0.00062
0.00049
 <
0.451 0
0.505 0
0.0975 0
0.188 0
0.031 0
0.257 0
0.055 0
0.047 0
V'T'>
.400
.616
.0955
.090
.048
.151
.035
.036

-------
                         VI.  REFERENCES
Aikman, F., H-W, Ou and R.W. Houghton, 1987.  Current Variability
     Across  the  New   England  Continental  Shelf  and  Slope.
     Submitted to Continental Shelf Research.

Bower, A.S.,  H.T.  Rossby, and  J.L.  Lillibridge,  1985.   The Gulf
     Stream—Barrier  or  blender?   J. Phys. Oceanogr., 15, 24-32.

Cornillon, P., 1986.  The Effect  of the New England Seamounts on
     Gulf  Stream  Meandering   as  Observed  from  Satellite  IR
     Imagery. J. Phys. Oceanogr.  16, 386-389.

Fofonoff,  N.P.,  and  R.M.  Hendry,  1985.   Current Variability near
     the Southeast Newfoundland Ridge.   J.  Phys.  Oceanogr., 15,
     963-984.
               •2  -   -c
Fu, L-L.,  J.  Vazquez  and M.E.  Parke, 1987.  Seasonal Variability
     of the  Gulf Stream  from  Satellite Altimetry.   J.  Geophys.
     Res., 92(Cl), 749-754.

Hamilton,  P.,  1984.    Topographic  and  Inertial  Waves  on the
     Continental  Rise of  the  Mid-Atlantic  Bight.    J.  Geophys.
     Res., 89(Cl), 695-710.

Hamilton,  P.,  1982.   Analysis of  current meter records  at the
     northwest  Atlantic  2800   meter radioactive   dumpsite,  EPA
     Tech. Rep. 521/1-82-002.

Halliwell, G.R.,  Jr., and C.N.K.  Mooers,  1983.   Meanders of the
     Gulf  Stream downstream from Cape  Hatteras  1975^1978.   J.
     Phys. Oceanogr., 13, 1275-1292.

Hendry, R.M.,  1982.   On  the structure of the  deep Gulf Stream.
     J. Mar. Res., 40, 119-142.      ^       y.

Hogg,  N.G.,   1981.     Topographic  waves  along   70°W  on  the
     continental rise.  J. Mar. Res. 39: 627-649.

Joyce, T.M., 1984.  Velocity and Hydrographic Structure of a Gulf
     Stream Warm-Core Ring.  J. Phys. Oceanogr., 14, 936-947.

Louis, J.P.,  B.D.  Petrie and P.C. Smith,  1982.   Observations of
     Topographiq Rossby Waves  on  the Continental Margin off Nova
     Scotia. J. Phys. Oceanogr. 12, 47-55.

Louis,  J.P.,   and P.C.   Smith  1982.    The  Development  of the'
     Barotropic  Radiation Field  of  an  Eddy Over  a Slope.   J.
     Phys. Oceanogr., 12, 56-73.
                              48

-------
 Luyten,  J.R.,  1977.   Scales of motion in the deep Gulf Stream and
      across  the continental rise.   J.  Mar.  Res.,  35,  49-74.

 Mooers,  C.N.K.,  1975.   Several effects of a baroclinic current on
      the cross-stream  propagation  of  inertial-internal waves.
      Geophys.  Fluid  Dyn.,  6,  245-275,  1975.

 Owens,  W.B., J.R. Luyten and H.L. Bryden,  1982.   Moored  velocity
      measurements on the edge  of  the Gulf  Stream recirculation.
      J.  Mar. Res., 40(Suppl.),  509-524.

 Owens,.W.B., 1984.   A synoptic and statistical  description of the
      Gulf  Stream  and subtropical  gyre  using  SOFAR  floats.    J.
      Phys. Oceanogr., 14,  104-113.

 Rhines,  P.B.,  1970.   Edge-,  bottom-,  and Rossby  waves  in  a
      rotating  stratified fluid.   Geophysical Fluid Dynamics,  l:
      273-302.

 Richardson,  P.S.,  1977.  On the crossover between the  Gulf Stream
      and the  Western  Boundary   Undercurrent.     Deep-Sea   Res.
      24:139-159.

 Richardson,  P.L.,   1983.    Eddy  kinetic  energy  in  the  North
      Atlantic  from  surface  drifters.   J.  Geophys.   Res.,   88,
      4355-4367.

 Richardson,  P.L., J.F.  Price,  W.B.  Owens, W.J.  Schmitz,   H.T.
      Rossby,  A.M.  Bradley,  J.R.  Valdes and  B.C. Webb,  1981.
      North Atlantic  subtropical gyre:   SOFAR  floats  tracked  by
      moored  listening stations.  Science 213, 435-437.

 Richardson,  P.L.,   1983.    Eddy  kinetic  energy  in   the  North
      Atlantic  from  surface  drifters.   J.  Geophys.   Res.,   88,
      4355-4367.

 Rossby,  T.,  E.R. Levine, and  D.N.  Conners,  1985.  The isopycnal
      Swallow float—A simple device  for  tracking  water parcels  in
      the  ocean.    Progress  in  Oceanography, Vol.  14, Pergamon,
      511-525.          ,

Rossby,  T.,  A.S. Bower, and P-T Shaw, 1985a.  Particle Pathways
      in the  Gulf  Stream.  Bull. Amer.  Met. Soc. 66(9), 1106-1110.

 Science  Applications International  Corporation,  1987.   Study  of
      Physical  Processes  on  the   U.S.  Mid-Atlantic  Continental
      Slope  and  Rise.    Final  Report, U.S.  Dept. of Interior,
     Minerals Management Service,  SAIC Report No.  86/7539.

Schmitz,  W.J.,  Jr.,  1976.    Eddy  kinetic energy  in the  deep
     western North Atlantic.  J. Geophys. Res. 81, 4981-4982.
                              49

-------
Schmitz, W.J., Jr., 1980.  Weakly depth-dependent segments of the
     North Atlantic circulation.  J. Mar. Res., 38, 111-133.

Schmitz, W.J., Jr., J.F. Price, P.L. Richardson, W.B. Owens, B.C.
     Webb, R.E.    Cheney  and H.T.  Rossby,  1981.   A preliminary
     exploration of the Gulf Stream system with SOFAR floats.  J.
     Phys. Oceanogr., 11, 1194-1204.

Shaw,  P-T,   and   H.T.   Rossby,   1984.     Towards  a  Lagrangian
     description  of the  Gulf Stream.   J.  Phys.  Oceanogr.,  14,
     528-540.

Thompson, R.O.R.Y., 1977.  Observations of Rossby waves near site
     D. Prog. Oceanogr., 7, 135-162.

Tracey,  K.L., and D.R.  Watts,  1986.   On  Gulf  Stream  meander
     characteristics near  Cape Hatteras.   J.  Geophys.  Res.,  91,
     7587-7602.

Tracey, K.L.,  M.  Cronin,  and D.R. Watts, 1985.   The Gulf Stream
     Experiment:    Inverted  Echo  Sounder Data  Report for the June
     1984  to May  1985  Deployment  Period.   University  of Rhode
     Island,  Narragansett, RI.  GSO Tech. Rept. No. 85-3.

Watts, D.R.,  and  W.E.  Johns,  1982.   Gulf  Stream  meanders.   J.
     Geophys. Res., 87, 9467-9476.
                               50

-------
                         ACKNOWLEDGMENTS
     Sincere appreciation is expressed to Dr. Robert W. Houghton
and the Lamont-Doherty Geological Observatory (LOGO), of
Columbia University, for the loan of five Aanderraa RCM-5
current meters.  The LOGO meters were a newer-designed version
of the RCM-5 meters, specifically adapted for extended
deployment in the deep ocean environment.  The overall success
of this two-year study is directly attributable to using the
current meters provided by LDGO.

     Sincere .thanks are also expressed to Dr. James Lane of the
Minerals Management Service (MMS), Department of the Interior.
The two-year current measurement study, detailed in this report,
was part of a broader MMS study of physical oceanographic
processess along the U.S. mid-Atlantic slope and rise provinces.
In his role as the overall MMS project manager, Dr. Lane
provided frequent and able coordination support and assistance
to the EPA project officer.
                                51

-------