Subtask 4230 Final Report

Jordi Font, Mikhail Emelianov, Agustí Julià, Joan Puertas, Maribel Lloret


Barcelona, 15 January 2004

    0. Introduction

        The Institut de Ciències del Mar (CMIMA-CSIC), Barcelona, is responsible for MFSTEP subtask 4230 in colaboration with OGS and IFREMER

According to the MFSTEP Description of Work the contents of subtask 4230 is:

The profilers, with the customized hardware and software upgrades, will be tested in the
Balearic Sea during pre-TOP. The deployments and recoveries (after about three months at
sea) will be conducted during CSIC cruises. Tests of 4 profilers of different manifacturers, but
with similar hard/software characteristics, are envisaged. CTD casts will be taken in the vicinity
of the deployment/recovery sites. The deployment teams of subtask 4310 will participate in
these deployment operations as part of a training exercise.
Responsible : CSIC
Participants : CSIC, OGS, IFREMER
Duration : T6-T9 (3 months)
Deliverables : Deployment/recovery of 4 profilers (D7).
This report concludes subtask 4230 and can be considered as the written version of MFSTEP deliverable D7.

   1. Preparation

       The initial plans were to deploy 2 APEX and 2 PROVOR profiling floats from the CSIC R/V García del Cid and recover them with the same vessel after three months at sea. Ship time was obtained accordingly, but due to the delay in MFSTEP kick-off the dates had to be adapted (tests to be carried out six months after task 4200 start) and it was not possible to completely modify the vessel schedule to fit our needs. The test period had to be reduced to 5-6 weeks.

Trajectory of a surface drifter following (N of 39.5º) the Northern current
The NW Mediterranean is a well studied area dominated by an alongslope current (the Liguro-Provenço-Catalan or Northern Current) controlled by a shelf/slope front between light (fresh) waters over the continental shelf and saltier offshore waters. The figure on the left displays the trajectory of a surface drifter during 50 days in spring/summer 2000 following the Northern current from the guf of Lions (42.5º N) until the gulf of Valencia (39.5º N). The figure below shows a vertical section of the salinity and geostrophic velocity near 40.5º N

Typical salinity front and geostrophic velocity structure

It was decided to deploy the floats in an area between two submarine canyons (around 41.5º N, 3.5º E) over 2000m depth and offshore the shelf/slope front. At the programmed parking depth (350 dbar) the mean temperature there in autumn is 13.23 ºC and salinity 38.48 psu.

Typical CTD profile in autumn in the deployment area

Considering the typical vertical temperature and salinity structure in the area in autumn (see figure on the left), the maximum profiling depth for the floats (700 dbar), and the number of possible sampling points to be transmitted (70), the sampling strategy was decided (see other WP4 subtasks and the two tables below). One float of each model was programmed to have one cycle per week, and the other two cycles per week.

Programmed sampling for the APEX floats

  Samples during drifting at parking depth: 5 to 8 (35505) and 13 (35506)
Programmed sampling for the PROVOR floats

   2. MFSTEP-0 cruise      

        The decided deployment strategy consisted in situating the four floats in the corners of a square of 1 nautical mile of side. And performing a CTD profile from the oceanographic vessel at each release point to compare with the profile recorded by the float for calibration/validation purposes. However, during the preparation of the equipment we realised that the APEX can not be programmed to record a profile at deployment, but only after completing a cycle. To solve this problem the deployment cruise was split in two. One week before the scheduled cruise, the two APEX floats were deployed on Friday 26 September from a small boat Blue Seal in two points situated one nautical mile apart in a N-S direction. The boat remained besides the second float until observing it initiated the sampling cycle by sinking slowly to the parking depth.

ARGOS ID 35504
                 15h06 GMT
                41º 44.040' N
                03º 43.003' E
ARGOS ID 35503
                15h25 GMT
                41º 45.000' N
                03º 43.002' E
  APEX  float before deployment
  APEX  float at surface

        The MFSTEP-0 team was formed by Agustí Julià and Mikhail Emelianov (ICM-CSIC) plus Laura Ursella (OGS)

   3. MFSTEP-1 cruise

From the ARGOS localisation of APEX 2 on September 29 we calculated a possible surfacing area for both floats and decided to deploy the two PROVOR close to it.

On 2 October 2003 the R/V García del Cid sailed from Barcelona at 10h local time with a "PROVOR Team" formed by Agustí Julià, Mikhail Emelianov, Joan Puertas (ICM-CSIC), Daniel Almonacid (UTM-CSIC), Sylvie Le Bras, Serge Le Reste (IFREMER), Riccardo Barbanti (OGS) and M. Isabelle Pujol (CLS).

After performing some tests with an ARGOS direction finder (goniometer) close to shore, the vessel reached the deployment area, some 70 nautical miles to the E-NE. It was decided to drop the floats with their specific deployment case at a speed of only 2 knots, for protection.

The CSIC research vessel García del Cid in the Barcelona harbour


ARGOS ID 35505
                 16h30 GMT
                41º 36.037' N
                03º 46.466' E

ARGOS ID 35506
                 18h45 GMT
                41º 36.037' N
                03º 43.791' E

Sequence of the PROVOR floats deployment

The floats were stored inside a deployment case (paper craft that dissolves in water), thrown over board, and after laying horizontally on the water surface for a while, turned slowly to vertical, the case sunk, and the float remained in the position shown in the figure (emerged antenna and sensors, as when emitting data after every cycle) until the programmed sinking time at 19h00 GMT. The two floats were deployed 2 nautical miles apart in the E-W direction. One CTD profile down to 800 dbar was made, with a Neil Brown Mark III probe, just besides each float after completing the deployment operation (CTD #1 and #2).

The following day, October 3, at 6h00 GMT (expected surfacing time for both PROVOR floats) the vessel was situated in the point where PROVOR 1 had been released. OGS communicated from Trieste that ARGOS had localised both PROVOR floats and provided the two positions. Due to rough seas it was not possible to use the goniometer as direction finder, but only to drive the search of the float by following the fluctuations of the signal intensity. The float was seen at 6h20, almost 1 nautical mile east of the release point, and a CTD profile (#3) was made besides it. The float PROVOR 2 was observed a couple of hours later and also a CTD cast (#4) was made for data intercomparison.

APEX 2 had been heard by the on board goniometer since early in the morning, but it was not possible to see it due to increasing rough seas (24 knots wind from the SW at that moment). While PROVOR have yellow antennas, APEX have black ones and then their visual localisation is more difficult. At 10h25 a CTD cast (#5) was performed in the position where the goniometer signal was highest, and hence we suppose that the float was nearby. The last float, APEX 1, was seen after some search at 12h20, and the last CTD cast (#6) was made there.

The following table summarises the information on the CTD casts. The files with the data (pressure, temperature, transmissometer, fluorometer, salinity and quality flag) can be accessed by clicking on the profile number. Raw data have only been interpolated to 1 dbar steps, without further filtering or other processing to allow comparison with floats profiles from the original records.

cast date time GMT latitude  longitude bottom depth cast depth
02 Oct 2003
41 35.99N
 03 46.40E
1836 m
809 dbar
      02 Oct 2003
41 36.00N
03 43.66E
1434 m
808 dbar
03 Oct 2003
41 36.51N
03 47.62E
2104 m
809 dbar
03 Oct 2003
41 35.85N
 03 45.38E
1505 m
809 dbar
03 Oct 2003
41 34.98N
 03 41.01E
1400 m
809 dbar
03 Oct 2003
41 38.02N
 03 49.56E
2038 m
809 dbar

 The figures below display these raw temperature and salinity CTD data for different pressure ranges, as well as the TS diagrams for the six profiles

   4. MFSTEP-2 cruise

        The floats surfaced periodically following the programmed cycles, although some small deviations occurred with the expected APEX surfacing times. They were localised by ARGOS and the transmitted CTD data recorded and immediately incorporated into the CORIOLIS ARGO data bank. The OGS MedARGO web page provided on-line information on the floats trajectories during the whole experiment. The following figures show the trajectories as they were displayed on the OGS web site just before the MFSTEP Observational Workpackages Meeting that took place at ENEA, La Spezia on 20-22 October 2003.  The colour sections are built with the consecutive ARGOS localisations when the float is at surface, while the straight grey lines indicate the distance between two surfacing events.

35503 APEX 1
35504 APEX 2
35505 PROVOR1
35506 PROVOR2

      The recovery was planned for 8-12 November, so the floats had been programmed to remain at surface after completing their last cycle on 7 November. The figure below (also from OGS) indicates the situation after all the four floats had surfaced. The last ARGOS positions are indicated by a white dot.

Three floats (2 PROVOR and 1 APEX) were relatively close to each other, while the second APEX was further SW, some 35 nautical miles in front of Barcelona.

Due to bad weather conditions the recovery cruise had to be delayed for 24 h. The García del Cid sailed from Barcelona on 9 November at 0h GMT, with a team formed by Jordi Font and Agustí Julià (ICM-CSIC) plus Riccardo Barbanti (OGS).

With the help of recent ARGOS localisations sent from OGS and the onboard goniometer, the group of three floats were caught with an auxiliary zodiac boat at

APEX 1          07h30     41º 07.36' N, 03º 41.19' E

PROVOR 1    08h45     41º 02.97' N, 03º 41.83' E

and after 5 hours of hard search

PROVOR 2    16h10     41º 27.60' N, 03º 48.46' E

when we were almost at dawn (thanks to Riccardo "magic eyes")

After the recovery of this group of floats, two CTD casts were made in the area: one in the point where the first localisation for PROVOR 2 on 7 November had been obtained by ARGOS, and the other one 3 nautical miles to its NW to check for short distance variability in the area. As the objective of performing the CTD casts was to compare to floats profiles, and due to the imprecision on the exact position where the floats had been profiling (only surface positions are available, and maybe up to a couple of hours later) plus the horizontal drift of the float while profiling, we wanted to identify a "natural variability" at 700-800 dbar that could set the maximum accuracy we would be able to recover by comparing floats to CTD data. Few water samples were taken for salinity determinations at ICM to check the CTD conductivity sensor performance.

cast date time GMT latitude  longitude bottom depth cast depth
09 Nov 2003
41 19.10N
 03 49.19E
2409 m
785 dbar
      09 Nov 2003
41 17.12N
03 52.49E
2354 m
803 dbar

During the morning of 10 November the vessel was engaged in a different activity not linked to MFSTEP around 41º 48' N, 5º 12' E, and then steamed for nearly 12 hours to the SW to find APEX 2. We were able to recover it deep in the dark, thanks to a relatively good performance of the goniometer, a quite calm sea, a hard visual effort of the onboard MedARGO team (and vessel crew) and some luck.

APEX 2   11 Nov 2003   01h00    40º 54.62' N, 02º 38.46' E

The image below indicates the positions of all the profiles recorded by the four floats during the experiment, as they appear on the Coriolis web pages

The figures below display the raw temperature and salinity data for different pressure ranges, as well as the TS diagrams, for the two CTD profiles recorded during MFSTEP-2

    5. Preliminar data intercomparison

            The main purpose for performing CTD casts was validation of the data recorded by the profiling floats. The CTD probe used in the experiment had been calibrated in March 2003, and it will calibrated again in February 2004. A first comparison made after the deployment cruise showed a good correspondence between temperature data, but a drift in salinity that appeared to be nearly 0.02 psu higher in the floats records.

Single comparison of T and S profiles

All S profiles from floats and CTD probe during MFSTEP-1

To clarify the situation (error in the floats conductivity sensors or in the CTD probe?) water samples were taken at depth during the MFSTEP-2 casts for further salinity determination on land using a high performance Guildeline AutoSal salinometer. However, the salinometer measurements made at ICM showed an unexpected noisy character

prof.  dbar




























Different tests indicated a probable malfunctioning of this specific salinometer. A new Guildline PortaSal will be installed at ICM by March 2004 and an intercomparison experiment with the old instrument is being prepared. While waiting for the results of this intercomparison and the new calibration of the CTD sensors, we propose to provisionally consider as sufficiently correct the data recorded during both cruises by the Neil Brown Mark III probe. An extra CTD profile recorded during MFSTEP-2 (November 10, 41º 48' N, 5º 12' E) indicated that at 2200 dbar, where high stability is known to exist, the recorded salinity lies within 0.007 psu from autumn climatology (MEDATLAS). This is better than the expected performance of the floats measurements (0.01 psu).

The deffinitive assessment of the sensors performance will be completed in subtask 4240 Assessment of field tests due for April 2004
Acknowledgements. Thanks are given to the García del Cid and Blue Seal crews and the OGS and IFREMER colleagues that participated in the organisation and execution of the field work. MFSTEP - Med ARGO is funded by the EU 5th Framework Program under contract EUK3-2001-00174