2. Description of the RAFOS Floats

3. Sound Source Deployment

4. Float Deployment

5. Float Performance

6. Float Tracking

7. Acknowledgements

8. References

1. Sound source moorings

2. Isopycnal float summary

3. Isobaric float summary

4. RAFOS float ballasting/temperature performance

5. Float clock and ARGOT information

1. BOUNCE I and BOUNCE I I launch locations and sound source locations

2. Float duration chart

This is the final data report of all RAFOS (acoustically tracked) float data collected during the 1994-1997 Boundary Current Experiment (BOUNCE) study of the Deep Western Boundary Current (DWBC) in the North Atlantic Ocean. Principal investigators for the project were Amy Bower and Robert Pickart of the Woods Hole Oceanographic Institution (WHOI) and William Smethie of the Lamont-Doherty Earth Observatory (LDEO). The overall objective of the program, funded by the National Science Foundation, was to obtain the first comprehensive description of the North Atlantic DWBC’s variability over a large path segment from Cape Hatteras to the Grand Banks. The experiment was comprised of CTD, tracer, and RAFOS float observations to achieve both Eulerian and Lagrangian descriptions of the DWBC. The three main objectives of the Lagrangian float study were 1) to determine fluid parcel pathways in the DWBC and identify regions of exchange with the interior, 2) to estimate the mean speed and variability of fluid parcels at two different levels in the DWBC, and 3) to study the kinematics and potential vorticity dynamics of fluid parcels in the DWBC at the Gulf Stream (GS) cross-over point near Cape Hatteras. Thirty floats were deployed: 15 were designed to be isopycnal floats, and 15 were isobaric floats. The isopycnal floats were ballasted for the s _t = 27.73 density surface (approximately 800 decibars (db)), roughly the level of the upper chloro-fluorocarbon (CFC) maximum of the DWBC (associated with the Upper Labrador Sea Water). The isobaric floats were ballasted for 3000 db, the level of the deep CFC maximum in the DWBC (associated with the Nordic Seas overflow water). Two separate float deployments took place: November-December 1994 on the R/V Endeavor (EN257), and May-June 1995 on the R/V Oceanus (OC269). CTD casts were taken at nearly all float deployment locations. The float missions were set to be two years in length. The floats were tracked using seven moored sound sources, built by Webb Research, Inc. Four of the sources were deployed specifically for BOUNCE on EN255 in October 1994, and a single replacement source was deployed near Bermuda from R/V Weatherbird I I in June 1996. The remaining two southern sources were originally set by Kevin Leaman (University of Miami) for a different experiment in April 1992. Only the replacement source near Bermuda has been recovered.

The RAFOS float is an acoustically tracked subsurface Lagrangian drifter (see Rossby et al., 1986, for a complete description of the RAFOS system), which is programmed to listen for signals from moored sound sources. The RAFOS floats determine the time-of-arrival (TOA) of these signals, from which, given the speed of sound in water, its position can be determined. The TOA of the acoustic signals, as well as temperature and pressure measurements are stored in the float’s micro-processor memory. Also stored in the float’s memory are confidence limits for each TOA, which indicate the quality of the TOA signal heard. The sound sources in this experiment were programmed to transmit an 80-second-long continuous wave tone, which linearly increases its frequency from 259.375 Hz to 260.898 Hz. The individual sound sources broadcast this tone twice a day, and broadcast at different times (beginning at 0030, 0100, and 0130 UTC, and then twelve hours later). The floats in this experiment listened for these signals once a day (beginning at 0000 UTC). The float temperature sensors were built by Yellow Springs Instrument Company and were calibrated to ± 0.01° C. Float pressure sensors were built by Data Instruments and calibrated to ± 1% at 2000 psi.

Two types of float, isopycnal and isobaric, were used in this experiment to seed the upper and lower cores of the DWBC. The WHOI float group (Jim Valdes, Bob Tavares, and Brian Guest) ballasted all the floats in the WHOI ballasting tanks. Isobaric floats are ballasted with a solid drop weight that forces the floats to be neutrally buoyant at a desired pressure surface, in this case 3000 db. More detail on the ballasting procedure can be found in Anderson-Fontana et al. (1996). Isopycnal floats are identical to the isobaric floats, but with the addition of a "compressee" attached with the weight package, outside the float body. The compressee is designed so that the entire float package has nearly the same compressibility as seawater (Rossby et al., 1985), thus allowing the float to follow water parcels along s _t surfaces. The isopycnal floats were placed in the upper DWBC to follow the s _t = 27.73 density surface, which starts out at about 800 db north of the Gulf Stream and about 1200 db south of the Gulf Stream. It would have been preferable to use isopycnal floats for the deep floats also, but due to the weak stratification at this depth, this was not technically feasible.

After the float completes its mission (in this case, two years), the float is programmed to drop its external ballast, rise to the ocean surface, and telemeter its data to ARGOS receivers aboard the NOAA Polar Orbiting Environmental Satellites (POES). Through ARGOS, the data are relayed to a ground station and transferred to a Global Processing Center. At the Global Processing Center, the data are processed and then transferred via the Internet to WHOI. The float data, including temperature, pressure, TOAS and respective confidences, are converted from hexadecimal to decimal, and are then ready for editing and tracking.

Five sound sources were deployed for this experiment, shown in Figure 1 and listed in Table 1 as SS1-5. Standard sound sources manufactured by Webb Research, Inc. were used. Before the floats were launched, sound sources 1 through 4 were moored in the Sargasso Sea between Georges Bank and the Blake-Bahama Outer Ridge. Three of the sources were concentrated around the Cape Hatteras region, to ensure tracking where the DWBC passes under the Gulf Stream, a focal area of this experiment. The acoustic range of the sound source is maximal in the sound channel, at roughly 1200 m, and decreases above and below that depth. The shallow floats in this experiment were able to hear the sound sources at greater ranges (~1500 – 2000 km) than the deep floats (~1000 km) because they were closer to the sound sources and the sound velocity channel. The sources were placed in locations that ensured maximal range for the deep set of floats. A few of the deep floats made it south of the Blake-Bahama Outer Ridge, beyond the range of most of the sound sources placed in this experiment. It was fortunate that two sound sources (SS6 and SS7, Figure 1 and Table 1), moored by Kevin Leaman in 1992, continued to transmit long past their expected life span of 2 years and were still able to be heard.

* The drift rates for these sources are unknown, and assumed to be zero for this experiment.

Figure 1. BOUNCE I and I I float deployment and sound source locations. Shallow floats are marked as circles, with the float numbers to the left; deep floats are marked as squares, with the float numbers to the right. Floats that were not heard from after launch have an ‘X’ after their numbers. Sound sources are marked with the F symbol. A grayed sound source label indicates that the sound source was not used to track the floats. Bathymetry is shown in 1000 meter contours.

Initially, all 30 floats were to be deployed along eight of the nine planned CTD sections on the November 1994 R/V Endeavor cruise (BOUNCE I ). The isopycnal floats were to seed the upper CFC core, and the isobaric floats the lower CFC core.

Due to unusually bad weather, and the subsequent reduction in the number of sections from nine to five, only 24 of the 30 floats were deployed along the sections during BOUNCE I. The remaining six floats were deployed during BOUNCE II in June 1995. Figure 1 shows the launch locations of the floats during the two BOUNCE cruises. A summary of BOUNCE I is as follows: four floats were deployed along the first, second, and fifth sections, and six floats were launched along sections three and four. (The first CTD section is the easternmost section; section numbers increase in order to the west.) All but one of the floats were set to their planned target pressure or density: deep floats to 3000 db, and shallow floats to s _t = 27.73. To one deep float (b281) launched at the southern end of section 4 were added four 3/8" washers to increase the target depth to about 3320 m. There was not enough room on the eyebolt to accommodate all the weights, so some of the washers were attached to this float with an additional tie-wrap. Summaries of the float launch and surface times and locations are found in Table 2 for the shallow, isopycnal floats, and Table 3 for the deep, isobaric floats.

The strategy for choosing the launch sites in BOUNCE I was generally as follows. Real-time CFC data were not available before the floats had to be launched, so a combination of historical information and data from upstream sections (after section 1) was used to place the floats in the CFC maxima. Historically, the deep CFC maximum is found very close to the 3500 m isobath, so one deep float was launched at this isobath on each section. The second deep float was then launched up to 50 km offshore from the 3500 m isobath. This distance was decreased as the cruise progressed because it was observed in the first CFC section that a 50-km spacing put the float outside the CFC maximum. If three floats were deployed at one depth along the same section, the total distance spanned by the floats was in general less than 50 km. The deployment strategy for the shallow floats was to get one float as far inshore as possible. The 1500 m isobath was chosen as the site of the most inshore deployment, which leaves about 700 m under the float. The other shallow floats were deployed 25-50 km offshore from the float deployed at the 1500-m isobath.

During BOUNCE I I , the remaining six floats were also deployed along hydrographic sections across the DWBC. In summary, four floats were launched along section 2 in a manner identical to BOUNCE I , with 2 floats ballasted for 3000 db and two for s _t = 27.73. This was done under the assumption that most of the floats launched during BOUNCE I had probably drifted west of 60° W, and the new floats would in essence extend the along-stream extent of the sampling. The two remaining floats, one deep and one shallow, were launched along section 5 at two locations also seeded during BOUNCE I (the 1500 and 3500 m isobaths), to increase the potential for some of the floats to reach the DWBC/GS cross-over.

All float launching was done by hand, with one person lying on the fantail and the other guiding the float down vertically. The launch tube was not used. The weather conditions at launch varied from flat calm to moderate seas.

The 30 RAFOS floats were deployed in the DWBC for 730-day missions. Out of the 30 floats deployed, 16 floats surfaced on time after two years (6 shallow and 10 deep). One of these 16 floats was ‘deaf’ (a shallow float), returning temperature and pressure records, but no TOAs to track the float. Of the remaining 14 floats, six (3 shallow, 3 deep) surfaced early, presumably due to lost weights. In most of these cases, there is evidence in the pressure records of bottom contact. Three shallow floats surfaced early due to a low battery. One deep float surfaced early due to the float sensing over-pressure. Four floats (3 shallow, 1 deep) failed to transmit entirely. Summaries of the float missions are described in Tables 2 and 3. The duration chart in Figure 2 describes visually the float missions in time. In total, 73% of the mission was accomplished: 84% of the deep missions, and 62% of the shallow missions.

From these results, it seems that contact with the bottom can lead to detachment of the drop weight or compressee from the float. In the case of the floats with compressees, it also seems that contact with the bottom leads to either leaking or attachment of sediments, because these floats tended to go too deep after they hit the bottom. Some shallow floats (e.g. b267 and b272) managed to hold on to their compressees for two years even after they hit the bottom, but either due to a leak or picking up sediments, these floats sank below their target depth.

The floats were able to hear signals from sound sources 2, 3, 4, 5, 6, and 7. Sound source 1 was never heard by any float. Sound source 6 was weak and not used for tracking because there were always suitable stronger sources that could be used during an interval that a float heard sound source 6. Therefore, only sound sources 2, 3, 4, 5 and 7 were used for tracking the floats. Overall, the TOA signals from sound sources 2, 3, 4, 5, and 7 were sufficient to track the floats. There were a few instances, however, where a float was unable to be tracked because the float passed through the base line of a sound source combination with which it was being tracked.

To determine when sound source 4 failed, and thus get back as many of the good TOAs as possible for tracking, a subset of deep floats was chosen that had clear TOAs from SS4 and other sound sources. By tracking the floats using the good TOAs, an accurate daily position could be found. The difference between the float position and the SS4 position yielded daily distances between SS4 and the float. Dividing the calculated distance to SS4 by the sound velocity resulted in an expected TOA for SS4. Where the actual TOA record diverged from the expected TOA record pinpointed the day SS4 failed. The TOA records from SS4 were used for tracking up until this day - 13 February 1995. Figure 2 shows the usable time segment of SS4 for this experiment.

While tracking the floats, it became apparent that sound sources 2 and 3 had clock drifts. These sound sources have not been recovered, so all that is known with certainty is the initial source clock offset. Calculating the sound velocity using the first and last TOAs from SS2 and SS3, and the distance between the float launch and surface positions, and the respective sound source, resulted in unrealistic sound velocities at the end of the mission. In contrast, reasonable sound velocities resulted from the same calculations with sound sources 4, 5, 6, and 7. The clocks of sources 2 and 3 apparently drifted over the two-year mission. The drift rate was calculated by adjusting the TOA times at the end of the mission until a satisfactory sound velocity resulted, and transforming the total offset into a linear drift rate.

Different sound velocities were used to track the deep and shallow floats. For the shallow floats, sound velocity was first calculated uniquely at each float’s median temperature and pressure value, using the UNESCO sound velocity polynomial (Fofonoff and Millard, 1983). That value was then averaged with the mean sound velocity at the average sound source depth (1500 m) northwest and southeast of the Gulf Stream, calculated using the CTD data from EN257. Using this method for the deep floats resulted in sound velocities that were generally too high. In this case a sound velocity was estimated for each deep float using the first TOA after launch and the distance between the launch position (GPS) and the sound source positions. These estimates were averaged to obtain a value of 1.495 km/sec, with greater weights being given to estimates for which the elapsed time between float launch and the time of the first TOA was relatively short.

Anderson-Fontana, S., M. Prater, and H. T. Rossby, 1996. RAFOS float data report of the North Atlantic Current Study, 1993-1995. Graduate School of Oceanography, University of Rhode Island, Technical Report No. 96-4, Narragansett, Rhode Island, 241 pp.

Fofonoff, N. P., and R. C. Millard, Jr., 1983. Algorithms for the computation of fundamental properties of seawater. UNESCO Technical Papers in Marine Science, UNESCO, Paris, 44, 53 pp.

Rossby, T., D. Dorson, and J. Fontaine, 1986. The RAFOS system. J. Atmos. Oceanic Technol., 3, 672-679.

Rossby, H. T., E. R. Levine, and D. N. Conners, 1985. The isopycnal Swallow float - a simple device for tracking water parcels in the ocean. Prog. Oceanog., 4, 511-525.