WHOI-2002-08

 

 

North Brazil Current Rings Experiment:

RAFOS Float Data Report

November 1998 – June 2000

 

by

 

Christine M. Wooding, Philip L. Richardson

Marguerite A. Pacheco, Deborah A. Glickson

and David M. Fratantoni

 

 

Woods Hole Oceanographic Institution

Woods Hole, Massachusetts 02543

 

July 2002

 

 

 

Technical Report

 

 

 

Funding was provided by the National Science Foundation through

Grants No. OCE-9729765 and OCE-0136477 to the Woods Hole Oceanographic Institution.
Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

 

 

Reproduction in whole or part is permitted for any purposes of the

United States Government. This report should be cited as:

Woods Hole Oceanog. Inst. Tech. Rept., WHOI-02-

 

 

Approved for public release; distribution unlimited.

 

 

Approved for Distribution:

 

 

 

_____________________________________

 

Nelson Hogg, Chairman

Department of Physical Oceanography


 

Abstract

 

Twenty-one RAFOS floats were tracked at depths of 200-1000 meters in and around several North Brazil Current Rings between November 1998 and June 2000.  This was part of an experiment to study the role of these current rings in transporting upper level South Atlantic water across the equatorial-tropical gyre boundary into the North Atlantic subtropical gyre.  The float trajectories in combination with surface drifters and satellite imagery reveal the sometimes complex life histories of several rings and their fate as they collide with the Lesser Antilles Islands.  This report describes the float trajectories, the velocity, temperature, and depth time series, and a preliminary analysis of the float data.

 


1. Introduction

 

North Brazil Current (NBC) rings are large (400 km diameter) anticyclones that pinch off from the NBC retroflection in the western tropical Atlantic near 8N and translate northwestward along the coast of South America toward the Caribbean (Johns et al., 1990; Didden and Schott, 1993; Richardson et al., 1994; Fratantoni et al., 1995).  NBC rings have been proposed as one of several important mechanisms for the transport of South Atlantic upper-ocean water across the equatorial-tropical gyre boundary and into the North Atlantic subtropical gyre.  Such transport is required to complete the meridional overturning cell in the Atlantic forced by the high-latitude production and southward export of North Atlantic Deep Water.  The mechanisms that contribute to NBC ring formation and the structure and dynamics of the rings themselves are not well understood.  The purpose of this study is to obtain, for the first time, comprehensive observations of the NBC retroflection, the NBC ring formation process, and the physical structure and properties of NBC rings as they translate northwestward along the low-latitude western boundary. The goal is to understand the process of NBC ring generation and to quantify the role of NBC rings in cross-equatorial and cross-gyre transport within the Atlantic meridional overturning cell. Specific objectives of the float component of the NBC Rings Experiment are to:

 

  1. Measure and describe the physical structure of NBC rings after separation from the NBC, including both near-synoptic and time-evolving temperature and velocity characteristics.
  2. Determine characteristic rates of translation, mixing, and decay as NBC rings move northwestward toward the Caribbean Sea.
  3. Identify the long-term fate of South Atlantic water trapped within the ring core, and determine the effective intergyre ring transport in several temperature/density watermass classes.

 


 

2. Float Deployment

 

Four cruises were conducted during the North Brazil Current Rings Experiment.  On each cruise, shipboard ADCP, XBT, and CTD-LADCP surveys were used to locate NBC rings and to measure their physical properties (Fleurant et al., 2000a, b, c).  Subsurface floats were deployed during the first three cruises (see Table 1) and during a non-NBC cruise following cruise 2.  Two acoustic sound sources (S1 and S2) were deployed on the first cruise and retrieved on the fourth.  An ALFOS float was also launched on the first cruise in order to monitor the sources.

 

Table 1. Number of RAFOS floats launched during each cruise.

 

Cruise

Beginning Date

Ending Date

RAFOS Floats Launched

Total

Into Rings

NBC 1

11-07-98

12-11-98

14

13

NBC 2

02-06-99

03-09-99

11

9

NBC 3

01-29-00

02-24-00

2

2

NBC 4

07-06-00

23-06-00

0

0

 

Note:  All cruises were on the R/V Seward  Johnson.

 

            The floats equilibrated, on average, deeper than their target depths.  We have divided them into three main depths: a shallow layer near 250 m, a mid-depth layer near 550 m, and a deep layer near 900 m.  North Atlantic Deep Water (NADW) is located below roughly 1000 m; it flows southward on average, counter to the translation direction of the rings.  The swirl velocity of several rings extended into the NADW, but because one of the goals of the NBC Ring Experiment was to measure the northward flux of South Atlantic Water, the floats were confined to depths less than 1000 m.

 

 

3. Description of the Floats

 

The North Brazil Current Rings Experiment used three different kinds of subsurface floats.  One was an ALFOS (ALACE-RAFOS) float, which surfaced once every 10 days and transmitted like a very short-mission RAFOS float.  There were six RAFOS floats of the "standard ROM" type, and there were 22 of the latest kind of RAFOS float, the DLD2 (see Appendix A for a discussion of  DLD2 technology).

Eighteen DLD2 floats were purchased from Seascan Corporation of Falmouth, Mass., and assembled, calibrated (temperature and pressure), and ballasted at WHOI.  Four other DLD2 floats, prototypes provided earlier by Seascan that had been used for testing in the WHOI float laboratory, were also used.   Six additional floats were assembled from spare parts and parts of floats retrieved from earlier experiments.  The ALFOS float, assembled at WHOI, was used to monitor the sound sources.

The floats recorded temperature, pressure, and times of arrival (TOAs) of sound signals transmitted by moored sound sources.  At the end of their missions, the floats dropped ballast weights, rose to the ocean surface and transmitted data to WHOI via the Service Argos Inc. satellite system.

Four of the 22 DLD2 floats were never heard by Service Argos, and one (float 037) was not tracked because it went deeper than its depth limit and surfaced after only 36 hours.  Two of the six older-design floats were not tracked because of poor-quality TOAs.  Overall we obtained 21 trajectories, although some were shorter than planned (Figure 1).  Considering that we purchased 18 floats, this seems like a good record.

The majority of floats was deployed on two cruises roughly two months apart.  In order to have most of the floats complete their missions in the same time period, the later-deployed floats were scheduled for shorter missions than the earlier ones.  The DLD2 floats were set to repeat their listening schedule every 12 hours.  Two of  the standard-ROM RAFOS floats were also set to repeat listening every 12 hours; the other four repeated every 24 hours.

 

 

 

 


Figure 1.  Float duration chart.  Float numbers are shown to the left of the line that indicates submerged time.  The dark part of the line indicates when the float was tracked. Floats are listed in order of launch date.

 

 

 

 

The floats were generally set to listen starting at 0100, 0130, and 0200.  Nothing was gained by including the 0200 window, and little was lost by omitting the 0030 window. However, the tracking program is hard-wired for 0030, 0100, 0130, and for 17 columns per record.  This meant that an extra step was needed, splitting the longer records, and shifting all TOAs and correlations right four columns, inserting a "null" 0030 window.

 

Columns before conversion:

     1   | 2 3 4 5 | 6 7 8 9 | 10 11 12 13 | 14 15 16 17

count |   0100  |   0130  |     0200        |    T&P

 

Columns after conversion:

     1   | 2 3 4 5 | 6 7 8 9 | 10 11 12 13 | 14 15 16 17

count |   0030  |   0100  |     0130        |    T&P

            (null)

 

Neither kind of RAFOS float stored the complete temperature or pressure values.  This means that a target value must be assigned, and the appropriate rollover value applied until the result is within range of the target.  The DLD2s used the last-recorded pressure and temperature of the mission as a default target (for earlier floats, the target was assigned).  These values were often at the maximum or minimum.  The floats also experienced a wide range of temperatures (especially those reaching the Caribbean or going ashore).  For these reasons, a number of floats (18 of 21) needed second and sometimes third targets to get the pressure and temperature correct, before they could be processed.

The floats were ballasted to drift at a preset depth.  This involved weighing the floats in air and in water at different pressures (see Tables 2 and 3).  Two floats (037 and 038) surfaced early due to exceeding their pressure limits.  On average the floats equilibrated 73 m deeper than their ballast depths.  RAFOS floats 431 and 432 were always shallower by 25 to 75 m than their target depths.  Floats 156 and 193 had mean pressures close to their targets.  The DLD2s were all deeper than their target depths throughout their missions.

 

 

 

 

 

Table 2. RAFOS float summary – launch and end positions

 

 

Launch

Surface

Target

Equil

Mean

Mean

 

Float

Date

Lat

Lon

Date

Lat

Lon

Depth

Depth

Temp

Depth

Ring

CTD

 

yymmdd

˚  N

˚  W

yymmdd

˚  N

˚  W

m.

m.

˚ C.

m.

 

 

 

DLD2 Floats

 

022

990219

9.179

52.439

000513

12.633

52.287

450

550

7.38

530

R3

2-42

023

990219

9.179

52.439

000513

7.981

47.296

800

910

5.21

907

R3

2-42

024

990219

8.780

52.481

000513

13.063

74.758

200

250

11.90

268

R3

2-41

025

990219

9.179

52.439

000513

15.086

70.016

200

250

13.90

282

R3

2-42

029

981208

8.841

56.242

000530

10.018

77.339

200

275

11.70

269

R1

 

030

981206

6.405

50.521

000527

8.756

48.164

450

555

7.44

531

R2

1-52

031

990213

10.266

57.203

000507

16.681

52.500

450

540

8.34

512

R2

2-23

032

981209

9.218

56.373

000531

19.544

66.109

450

620

7.29

592

R1

1-64

033

981206

6.405

50.521

000527

12.544

58.603

200

320

10.24

311

R2

1-52

034

981208

8.841

56.242

000530

13.571

64.200

450

600

7.38

527

R1

 

035

981206

7.112

50.771

000527

12.177

60.228

450

540

7.56

502

R2

1-53

036

990219

8.780

52.481

000513

12.209

52.102

800

860

5.31

843

R3

2-41

037

990302

5.498

48.321

990304

5.298

48.366

800

1000

Not tracked

R4

2-68

038

981206

7.112

50.771

991104

1.685

23.879

200

240

15.03

248

R2

1-53

039

990305

2.671

47.336

000814

1.367

35.457

800

880

4.87

879

NBC

2-80

040

981208

8.841

56.242

 

 

 

800

 

Never heard

R1

 

041

981206

7.112

50.771

000527

6.782

46.751

800

890

4.98

860

R2

1-53

042

981206

6.405

50.521

000527

12.881

50.788

800

870

5.22

860

R2

1-52

043

990411

8.314

52.608

 

 

 

200

 

Never heard

NBC

 

044

990219

8.780

52.482

 

 

 

450

 

Never heard

R3

2-41

045

990213

10.266

57.203

 

 

 

200

 

Never heard

R2

2-23

046

990213

10.266

57.203

000507

9.547

55.230

800

840

5.36

891

R2

2-23

            Older-design RAFOS Floats

 

 

156

000220

10.020

57.103

000714

13.779

63.230

200

220

12.95

223

R5

3-52

193

000220

10.020

57.103

000714

14.478

61.948

450

520

9.79

491

R5

3-52

430

981209

8.901

56.177

000607

9.936

84.087

200

 

Not tracked

R1

1-63

431

981205

6.231

49.509

000528

4.795

1.957

200

180

18.65

148

R2

1-50

432

981206

7.112

50.771

000528

11.110

49.987

1000

950

4.98

935

R2

1-53

433

981209

9.218

56.373

000227

14.360

68.639

200

 

Not tracked

R1

1-64

 

Floats 022 - 046 were DLD2 floats that listened and recorded T/P every 12 hours.  Float 43 was deployed by John Whitehead from R/V Oceanus.  Floats 156 and 193 were older, re-powered RAFOS floats that listened every 12 hours and recorded T/P every 24.  Floats 430 - 433 were older-design RAFOS floats that listened and recorded T/P every 24 hours.  R1 refers to Ring 1, etc.   These designations are internal to this report.  The station number in the last column denotes the cruise and the CTD taken at the time that float was launched (Fleurant et al., 2000a, b, c).

 

 

 


Table 3. RAFOS float summary – performance

 

 

Startup

Expected

Actual

Offsets (secs)

 Messages

 

Track-able

         Sources

Comments

Float

Date

Surface

Surface

Initial

Final

Received

%

Months

Heard

 

037

990221

000814

990304

 

 

2/2

 

0/0

 

 surfaced after 36 hrs

 out-of-range to east

 never heard

 

 "exponential"pressure

 ended in Caribbean

 ended in Caribbean

 

 never heard

 

 

 

 

 

 

 

 

 *TOAs end at surface

 

 ended in Caribbean

 ended in Caribbean

 never heard

 ended in Caribbean

 battery  failure?

*on surface, no TOAs

 never heard

 ended in Caribbean

 ended in Caribbean

038

981204

000527

991104

0.0

0.23

329/334

98

6/11

4

045

990213

000508

 

 

 

 

 

 

 

031

990213

000508

000507

0.0

0.68

450/450

100

15/15

5

046

990213

000508

000507

0.0

-2.02

450/450

100

15/15

6

024

990218

000513

000514

-1.0

1.19

432/450

96

5/15

3

025

990218

000513

000514

-1.0

1.45

446/450

99

9/15

4

022

990218

000513

000514

-1.0

1.58

445/450

99

15/15

5

044

990218

000513

 

 

 

 

 

 

 

023

990218

000513

000514

-2.0

2.76

440/450

97

15/15

5

036

990218

000513

000514

-1.0

1.55

341/540

63

15/15

6

030

981204

000527

000527

0.0

2.07

537/540

99

18/18

5

033

981204

000527

000527

0.0

0.86

539/540

99

18/18

4

035

981204

000527

000527

0.0

2.00

540/540

100

18/18

6

041

981204

000527

000527

0.0

4.04

481/540

89

18/18

5

042

981204

000527

000527

0.0

3.76

539/540

99

18/18

6

431

981205

000528

000528

-23.0

-206.4**

215/271

79

2/18

2

432

981205

000528

000528

-24.0

-258.0

246/271

91

18/18

5

029

981208

000531

000530

0.0

2.67

538/540

99

3/18

2

034

981208

000531

000530

0.0

-0.32

538/540

99

16/18

6

040

981208

000531

 

 

 

 

 

 

 

032

981209

000601

000531

0.0

3.30

466/540

86

15/18

4

433

981209

000601

000227

-24.0

-194.8

212/223

95

0/18

2

430

981209

000602

000601

-22.0

-158.5

67/271

25

0/18

0

043

990409

000702

 

 

 

 

 

 

 

156

000215

000714

000714

4.0

0.19

151/151

100

3/5

4

193

000215

000714

000714

2.0

-6.77

150/151

99

4/5

4

039

990221

000814

000814

0.0

1.22

441/540

82

18/18

6

 

        All dates are year-month-day.

*      The first Argos transmissions for these two floats were on land.

**    The final offset for 431 was reduced to -67.31 to match the TOA end date of 990414.

 

 


One advantage of RAFOS floats over the previous SOFAR floats is that two end points are available to check the acoustic tracking: the launch location and the surfacing location, obtained from Service Argos' calculation of the locus of the transmitter.  The surfacing location was not useful in two cases because the floats were on land before they began transmitting.  Float 430 stayed submerged for 100 days; it was probably picked up around day 250.  Float 430 first transmitted from Costa Rica.  Float 431 stayed submerged for 120 days, and was picked up after 200 days.  It first transmitted from Point Three Points in Ghana (so-called because it is the land closest to the Equator, the Greenwich Meridian and sea-level).

Before we received all messages from 432, it was picked up and taken to Martinique.  From there it was returned to Woods Hole.

Five floats (029, 031, 032, 034, 046) shared identical zeroeth messages (roughly one-half of their first output records).  This had the effect of making them look like they had recorded 450 messages in 449 days.  In order to produce the correct result, the start dates had to be moved one day earlier than shown in the launch logs.  It is not clear why these floats failed to overwrite the default information in the zeroeth message.

Several floats (see 031, 033, and 035) show a clumping or quantizing in their temperature records.  This is a result of being located in layers of nearly constant temperature and salinity (Schmitt et al., 1987)

Float 433 had no useable TOAs after 80 records, and only heard one source.  It stopped collecting data and began transmitting after 223 days and gave its final status as "on the surface."  The TOA record is typical of the case where the battery voltage drops.

RAFOS floats can detect when they are on the surface by recrossing a minimum pressure and also exceeding a maximum temperature.  Because of past problems, for the NBC experiment the minimum pressure was set to 1000 counts, very close to (and sometimes above) the surface.  This explains why 430 and 431 continued "in mission" after reaching the surface, and suggests that the "on surface" status of 433 was caused by low voltage.

The pressure record of float 046 repeatedly sweeps the same 150-meter range.  No oceanographic explanation has been posited and these results are attributed to instrument problems.

The DLD2 floats can store message-numbers up to 2048, but the earlier floats can only store values up to 256.  The checksum is used to decide whether a message-number is greater than 256 or not.  This creates uncertainty when you want to fill in gaps in the received data with messages with bad checksums.  This was an issue for floats 431 and 432.


4.  Description of Sound Sources

 

The NBC floats used sources deployed by scientists from three different countries for five different experiments (Table 4, Figures 2, 3).  The oldest source used, 185, was deployed October 19, 1994, as part of the Deep Western Boundary Current Experiment (BOUNCE) (Hunt and Bower, 1998).  Source 185's clock had drifted at a faster rate than was deduced at the end of BOUNCE, so an expedient offset was applied the one time it was used.

 

Table 4:  Sound source summary

 

 

ID

Pong

Corr

Rep

Start

End

Offset

Drift

Depth

Lat

Lon

S.V.

Exp.

 

Time

(min)

Rate

Date

Date

(sec)

(sec/rep)

(m)

˚ N

˚ W

(m/sec)

BOUNCE

185

0100

0

12

941019

none

0.561

0.0057

1500

36.687

58.263

1484

French

B1

0100

0

24

950101

none

0.0

0.005

1000

7.028

49.925

1480

DBE

191

0100

0

24

950212

none

0.0

0.0

1000

-0.327

32.643

1482

NBC

69

0100

0

12

981108

000622

0.0

0.0043

650

13.000

57.886

1484

NBC

77

0130

-30

12

981110

000620

0.0

0.0113

700

12.996

51.107

1484

MOVE

403

0035

25

24

000117

none

0.0

0.0

1100

21.938

62.570

1484

MOVE

404

0100

0

24

000129

010103

-3.0

0.0

1110

15.324

51.526

1484

MOVE

406

0130

-30