Edited by: Juliet Hermes, South African Environmental Observation Network (SAEON), South Africa
Reviewed by: Clive Reginald McMahon, Sydney Institute of Marine Science, Australia; Fabien Roquet, University of Gothenburg, Sweden
This article was submitted to Ocean Observation, a section of the journal Frontiers in Marine Science
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Surface and sub-surface ocean temperature observations collected by sea turtles (ST) during the first phase (Jan 2019–April 2020) of the Sea Turtle for Ocean Research and Monitoring (STORM) project are compared against
Because sea temperature has a strong influence on climate dynamics and global transport of ocean water masses, knowledge and observation of this parameter is fundamental to achieve realistic forecasts and simulations of the coupled ocean-atmosphere system at all spatial and temporal scales. In particular, sea surface temperature (SST) is an essential parameter in meteorology and oceanography, especially in tropical regions, where it plays a major role in the formation of tropical low-pressure systems, and more generally, ocean-atmosphere interactions.
In recent decades, the international community has developed the Global Ocean Observing System (GOOS,
In the equatorial zone, the Research Moored Array for African-Asian-Australian Monsoon Analysis and Prediction (RAMA) network (
Although the ARGO and RAMA networks have significantly improved the density and quality of measurements in the tropical oceans, their coverage is not uniform and not always optimized for short- and medium-term applications. The sampling frequency of ARGO drifters (1 profile every 10 days on average) is, for instance, rather low and the spatial resolution of the RAMA network (5–15° zonal resolution) is quite loose, especially in the Southwest Indian Ocean (SWIO). With the continued deployment of new drifters, mooring buoys and satellite missions, the coverage of ocean observations is nevertheless expected to further improve considerably in the coming years. However, because of the extensive development and maintenance costs associated with such programs, researchers have also been considering new measurement capabilities to improve and complement existing observing networks at lower cost.
An appealing alternative for collecting oceanic data over periods ranging from a few days to several years is to use marine animals equipped with biologgers (aka bio-logging), that is electronic tags equipped with various environmental sensors. Animal-borne systems are indeed relatively inexpensive to operate compared to conventional observation systems (gliders, ARGO floats, ships). They can be deployed worldwide with limited human resources and therefore, used to extend ocean observations to remote and hard-to-reach areas. Furthermore, bio-logging also provides countries with limited human and financial resources the opportunity to contribute significantly to the collection of ocean observations. In this regard, animal-borne observations do not only provide unique and essential ocean data, but also have a positive impact on empowerment. Among the numerous marine candidate species available, air-breathing diving predators such as elephant seals (
The use of other marine animal species for collecting ocean data is attracting increasing interest from the international scientific community. For example,
The deployment of Argos tags on ST is a common and well-controlled procedure that has been used for more than 20 years to track the movements of all ST species worldwide (
In the tropical Indian Ocean, ST measurements were initiated in January 2019 as part of the INTERREG-V Indian Ocean 2014–2020 research project “ReNovRisk-Cyclones,” with the aim of assessing the relevance of such observations for ocean monitoring and modeling in the Southwest Indian Ocean (SWIO). During this experiment, conducted by Laboratory of Atmosphere and Cyclones (LACy), Centre d’Etude et de Découverte des Tortues Marines (CEDTM) and Kelonia, the Marine Turtle Observatory of Reunion Island, a dozen of rehabilitated loggerhead and Olive Ridley STs were equipped with Argos tags integrating TD (temperature/depth) sensors, before being released from Reunion Island. The preliminary results obtained during this experiment, presented hereafter, have confirmed the potential of this approach for oceanic studies, and led to the creation of the international research program “Sea Turtle for Ocean Research and Monitoring” (STORM).
The SWIO is home of significant cyclonic activity that regularly, and often dramatically, affects the inhabitants along the East African coast as well as over small (e.g., Reunion, Mauritius, Seychelles) and large (e.g., Madagascar) islands of this region. This basin, the second to third most active ocean basin in terms of cyclonic activity (
Although the advent of coupled ocean-atmosphere models is a major achievement in improving TC forecasts on all time scales (e.g.,
In this study, ∼115,000 temperature observations collected by loggerhead and Olive Ridley STs released from Reunion Island (55.28°E; 21.15°S) are compared to conventional ocean temperature datasets (ARGO, satellites), and used as an independent observational source to evaluate ocean model temperature forecasts. The paper is organized as follows: Section “Sea Turtle Observations” describes the data collection procedures, processing methods and spatial distribution of observations collected during the preliminary phase of STORM (January 2019–April 2020). The accuracy of temperature observations collected by STs is assessed in section “Evaluation of Sea Turtle Temperature Measurements” through comparisons with
All sea turtles equipped in this study are juvenile specimens that were accidentally hooked by longliners in the vicinity of Reunion Island. In addition to hooking, the main sources of injury are caused by boat propellers, plastic ingestion or entanglement in large pieces of plastic (
ST Morphometrics at release (Standard Carapace Length SCL, Curved Carapace Length CCL, body mass) and tracking parameters (Argos ID, tag manufacturer and model) for the 11 juvenile sea turtles released from Reunion Island (55.28°E; 21.15°S), Indian Ocean.
Argos ID | Tag Manufacturer/Model | ST Species | ST Name | Date of release | CCL (cm) | SCL (cm) | Mass (kg) |
65711 | WC SPLASH10-344D-01 | Cc. | Ilona | 2019-01-09 | 78.5 | 72 | 69.2 |
65711* | – | Cc. | Fifi | 2019-10-07 | 91.5 | 88 | 115.2 |
65712 | – | Cc. | Brice | 2019-04-02 | 70 | 85 | 65.8 |
65722 | – | Cc. | Samson | 2019-02-15 | 69 | 76 | 59.3 |
65723** | – | Cc. | Maria | 2019-07-16 | – | 73 | 45.6 |
180908 | WC SPLASH10-E-385A-01 | Cc. | Tikaf | 2019-09-30 | 67.7 | 79 | 58 |
178936 | LK K2G 376D DIVE | Cc. | Tina | 2019-05-21 | 66.4 | 72 | 47.8 |
178937 | – | Lo. | Nesta | 2019-05-02 | 61 | 64 | 42.2 |
197624 | LK K2G 376E DIVE | Cc. | Tom | 2020-02-19 | 66.9 | 73 | 57.4 |
197625 | – | Cc. | India | 2020-02-26 | 62 | 67 | 42.9 |
197982 | – | Lo. | Lazarus | 2020-03-02 | 56 | 61 | 32.6 |
All animals were considered juveniles, except for the loggerhead called “FIFI,” which was assumed to be a sexually mature male at the time of release. The average individual curved carapace length (CCL) for loggerhead (resp. Olive Ridley) was 70.8 ± 8.9 cm (resp. 58.5 ± 3 cm) and the average weight was 62.35 ± 21.7 kg (resp. 37.4 ± 6 kg). All STs were instrumented at Kelonia and released from Reunion Island, except for the loggerhead ST “MARIA,” which was equipped on board a longliner south of this island (50.05°E; 29.81°S) by a fisheries observer from the Institute of Ocean and Atmosphere (IPMA, Olhão, Portugal). On average, the Kelonia care center rehabilitates and releases ∼25 STs every year. Since the start of the STORM project, in January 2019, about half of the animals released from this center have been equipped with electronic tags recording temperature/depth data.
A key element of the STORM program is to ensure that ST handling practices minimize negative effects on animal welfare. As mentioned previously, none of the animal considered in this study was intentionally captured and a particular attention was also dedicated to tag selection. Measurements were obtained from two types of tags (see
Top panel: Photographs of
All tags have been factory calibrated by manufacturers and are given with an accuracy of 0.2° for temperature and 1% for pressure measurement. They were configured to continuously, and simultaneously, sample bivariate time series of external temperature and depth (from pressure) with a time step of 5 min. The repetition rate was set at 15 s for WC tags and 45 s for LK ones. The maximum number of messages per day was set to 500, with priority given to the transmission of time series data. For the one WC tag equipped with Fastloc-GPS technology (WC-SPLASH10-F,
The WC tags were also set to archive all collected depth and temperature measurements every 10 s. The data archive was only recovered from the tag deployed on the poached individual “ILONA” (Argos ID 65711,
Argos location processing was set to use the Kalman filter scheme (
The number of observations collected by the 11 animals between 9 January 2019 (release date of the first ST “ILONA”) and 30 April 2020 (date of the last measurements transmitted by STs “TINA,” “INDIA,” “TOM,” and “LAZARE” at the time of compiling these statistics
Number of temperature observations collected by the 11 STs of the STORM program as of 30 April 2020 indicated by month (JFMAMJJASOND) and by individual (Ndata) with total number shown in bold.
The lifetime of a tag is determined by the capacity of its battery, the charge of which depends on the frequency of transmission, but also by unexpected incidents that may affect the integrity of the tag (or of its attachment material) as well as the animal surfacing behavior itself. According to the parameterized daily allocation for transmission, each tag was expected to transmit for a period of 7–8 months. Including the tags deployed on STs “TINA” and “INDIA,” which are still active 15 and 7 months after being released, respectively, four of the tags did actually transmit for more than 7 months while four have had a duration of less than 3 months (see caption in
Approximately 115,500 depth-temperature observation pairs (
The trajectories of the 11 animals equipped in this study are shown in
The rough analysis of the vertical distribution of ST-borne data (
Vertical distribution of data collected by all sea turtles between January 2019 and April 2020. Number of observations (NDATA) between 0 and 120 m in 20 m increments at all times (dark gray) and at night (18–5, Reunion Island Local Standard Time) only (light gray). The black line indicates the cumulative percentage of collected data.
To achieve a more precise description of the vertical distribution of collected ST-borne data,
Depth (D, bottom of the layer) of Glo12 model levels (#).
# | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 |
D (m) | 0.5 | 1.5 | 2.6 | 3.8 | 5.0 | 6.4 | 7.9 | 9.5 | 11.0 | 13.4 | 15.8 | 18.4 | 21.6 | 25.2 | 29.4 |
# | |||||||||||||||
D (m) | 34.4 | 40.3 | 47.3 | 55.7 | 65.8 | 77.8 | 92,3 | 109 | 130 | 155 | 186 | 222 | 266 | 318 |
Most of the measurements are collected just below the surface (0.5–2.6 m), but a significant number of observations are also gathered between 20 and 90 m, with a secondary peak of ∼5000 data between 30 and 35 m (model layer #16). The associated mean vertical temperature profile (
Mean vertical profiles of ocean temperatures in 10 m steps, as deduced from data collected by ST SAMSON (10/02–29/04 2019) and BRICE (7/04–10/07 2019) within latitude bands LB1 [20°S–11°S] (dash), LB2 [11°S–3°S] (plain) and LB3 [3°S–7°N] (long dash) shown in
In order to qualitatively assess the realism of temperature structures sampled by STs,
Observed temperatures near the surface appear relatively uniform across the whole transect, with mean values comprised between 29.2 and 30°C in the first 20 m. SST is also relatively uniform, although slightly higher (30°C) in the SCTR area (LB2) than in the other two regions (∼29.5°C). Below 20 m, significant regional differences can be seen in the vertical structure of ocean temperature. First of all, the mixed layer depth (MLD), identified by a strong gradient in vertical temperature profiles, varies significantly with latitude. It averages at ∼25 m in the southern hemisphere (27 m in LB1 and 22 m in LB2), but deepens to 40 m in the trans-equatorial area (LB3). The rate of temperature decrease in the OML is fairly uniform in the three regions, but also changes subsequently underneath. It is thus relatively slow in the LB3 zone, with a temperature of 26°C at 100 m depth (∼−5°C/100 m), but increases significantly in the two northernmost areas to reach ∼−8.5°C/100 m in LB1 (22.8°C at 100 m depth) and ∼−11°C/100 m in LB2 (21.5°C at 100 m depth). The differences on either side of the SCTR (LB2) are expected and reflect the existence of strong upwelling and a shallow thermocline in the SCTR region (
In order to quantitatively evaluate the temperature measurements provided by STs, a comparison with ARGO profiler data, whose quality is well-established, can be particularly instructive. As shown in the previous section, however, the limited spatial and temporal resolution of the ARGO profiling network within the area of evolution of ST considered in this project (see
Comparison of various ST-borne (square), ARGO (red) and Glo12 (blue dotted) temperature profiles in the SWIO basin on
Given the limited spatial and temporal variability of tropical oceans under normal conditions (i.e., in the absence of tropical cyclone) over distances (resp. periods) of a few tens of km (resp. days), imperfect collocation of ARGO/ST observations is expected to have limited impact on temperature data comparisons. Thus, it can be seen that ST and ARGO temperature measurements are generally in good agreement at the four considered times/locations. Small discrepancies can nevertheless be observed in the first 50–60 m of the ocean in the SCTR region (
Another possibility to assess the accuracy of SST temperature observations is to compare ST-borne observations with spaceborne measurements.
Current SST observations rely primarily on the use of radiometric sensors deployed on board scrolling (e.g., METOPs) or geostationary (e.g., MSG) Earth observation satellites. These data generally consist of a combination of observations collected by radiometric sensors operating in the thermal infrared and/or microwave range—infrared sensors are more accurate than microwave sounders but are very sensitive to the presence of clouds. To date, there are a dozen operational satellite SST products available to the research community with spatial resolutions ranging from 0.01° to 0.5°. In the following, ST-borne SST data collected at or just below the sea surface are compared to three sets of subskin satellite SST products : (i) the low-resolution (0.25°) OSTIA (Operational Sea Surface Temperature and Ice Analysis) dataset, operationally produced by the United Kingdom Met-Office, and (ii) two versions of the high-resolution (0.05°) OSI-SAF (Satellite Application Facility on Ocean and Sea Ice) dataset, produced by the French Meteorological Service Météo-France.
The level 4 OSTIA diurnal SST product is based on the United Kingdom Met Office system (
OSI-SAF products used in this study are distributed by the EUMETCAST Data Centre
In the following, all
Histogram (left panel) and scatterplot (right panel) of differences between SST data collected by STs (SSTST) and inferred from the satellite products:
Overall, SST data collected by ST vary between 20° and 32°C throughout the domain. Error statistics show that SST satellite data are, on the whole, very close to the ST-borne measurements although a slight systematic underestimation can be noticed for each of these three products with respect to ST data.
The comparison of ST-borne SST measurement against OSI-SAF operational (
These values are in good agreement with the maximum margin of error expected for both ST-borne (0.2°) and satellite (0.5°± 1°) measurements, which attests to the good accuracy of both in situ and remotely sensed measurements in the area of evolution of STs. It should also be noted that considering ST data collected precisely at or slightly below the ocean surface (i.e., down to 0.51 m) does not affect comparisons with satellite subskin SST data. The dispersion thus remains the same regardless of the depth of the ST measurements (see colors in
In this Section, ST-borne data are used to evaluate temperature forecasts from Mercator-Ocean 1/12° global operational model Glo12, based on the Nucleus for European Modeling of the Ocean (NEMO) OGCM. NEMO is a state-of-the-art community modeling framework for research and forecasting in ocean and climate science, developed by a European consortium (
Comparisons between ST-borne SST data and Glo12 forecasts over the 16-month analysis period are shown in
As in
The comparison between
As in
Overall, the dispersion (
Analysis of the vertical profile of the mean bias error [TNEMO–TST, black curve)] shows excellent agreement within the first 35 m (16 first model layers). Within this layer, which more or less corresponds to the depth of the OML in this area (
These results are qualitatively consistent with Glo12 forecast verification scores performed since 2008 in the tropical Indian Ocean. As shown in
This overall tendency nevertheless shows significant variability when areas and time periods are considered independently. As shown in
Scatterplots of temperature data collected by STs (TST) SAMSON (left panel) and BRICE (right panel) and forecasted by Glo12 (TNEMO) within latitude bands
The data collected by the individual “SAMSON” were obtained during the second half of the summer season (mid-February–end of April). Overall, one notes a systematic underestimation of model predicted temperatures in the three geographical areas, which tends to become more pronounced as depth increases [mean differences between models and observations of −0.27°/−0.4° and −0.5° in LB1 (
Although the amount of data used in this analysis is clearly not sufficient to draw strong conclusions on the performance of Glo12 at the regional and seasonal scales, these results suggest a strong seasonal and spatial dependence of the model performance, especially in the first 50 m of the ocean. In contrast, the model’s tendency to significantly underestimate ocean temperatures below 50 m appears systematic and is broadly consistent with the global figures presented in
The exploratory measurements collected during the first phase of the STORM project (January 2019–June 2020) aimed to assess the capabilities of ST’s equipped with environmental tags to sample the thermal properties of the western tropical Indian Ocean. Although based on a limited dataset (11 sea turtles, ∼115,000 observations), the preliminary results obtained during this experiment confirms the quality of temperature observations provided by ST and clearly shows potential for ocean monitoring and forecasting in this tropical area. Comparisons of temperature profiles collected by STs with measurements from co-located ARGO drifting buoys are in good agreement at all sampling depths (0–250 m), while comparisons with various SST satellite products fall within the expected uncertainties with a slight overestimation of ST-derived temperature data of ∼0.2° ± 1°. The comparison of ST-borne surface and subsurface temperature observations with Glo12 ocean model forecasts also demonstrated the potential of such
A particularly striking result of this study is related to the ability of ST to collect large amount of data in the upper 100 m of the ocean, a depth that more or less correspond to the maximum depth of the mixed layer in tropical oceans (
These encouraging results have led us to carry on with this experiment and to build an ambitious program, based on the instrumentation of several dozen animals released throughout the SWIO basin. In this respect, the next steps of the STORM program will be carried out in two phases aiming, on the one hand, to assess the possibility of collecting large datasets over very short and targeted periods and, on the other hand, to extend the collection of observations transmitted by STs to the whole SWIO basin.
The second phase will begin in mid-November 2020 with the objective of assessing the capacity of the ST to collect observations in the vicinity of tropical cyclones during the 2020–2021 warm season (November-April). To this end, a dozen animals equipped, for the first time in this project, with CTD (conductivity/temperature/depth) sensors, will be released from Reunion Island at very short time intervals (10 days) within a period of about 3 months. This new experimental strategy will make it possible to increase both the temporal continuity, the number and the space-time distribution of the observations collected by STs, as well as the probability of intercepting a tropical cyclone. The results obtained during the preliminary phase of STORM have indeed demonstrated the ability of ST to collect ocean temperature data in close proximity to tropical low-pressure systems developing in the SWIO basin. In April 2019, the individual “BRICE” was or instance trapped for several days in the immediate vicinity of TC Kenneth during its cyclogenesis and initial intensification phases, while the individuals “TOM” and “INDIA” evolved in the immediate vicinity of TC Herold in April 2020. In the latter case, ST “INDIA” remained stuck for several days within 10–50 km of the storm center, while ST “TOM” evolved a few hundred km east of the storm (
Location of STs “INDIA” (green symbol) and “TOM” (yellow symbol) during the cyclogenesis and intensification phases of tropical cyclone Herold (NE of Madagascar), superimposed on AQUA-TERRA satellite images at
The third phase of STORM, to be funded by the EU under the INTERREG-V Indian Ocean program, should start in June 2021 and will consist of continuously releasing ∼60–70 ST over a 2-year period at different locations in the basin. This experiment, made possible through the involvement of six regional marine reservations in Seychelles (SE), Mozambique (MZ), Reunion (FR), Comoros (CO), and Terres Australes et Antarctiques Françaises (TAAF), will enable the collection of a unique set of high spatial and temporal resolution temperature and salinity data over most of the SWIO basin. In addition to open ocean areas, which have been the subject of previous STORM observation campaigns, this new phase will make it possible to also collect observations in the Mozambique Channel, a region where in situ observations are quite scarce. The continuous and regular release of STs during a period of 2 years will allow assessing the capability of ST-borne observations to capture the variability of the tropical Indian Ocean (e.g., seasonal and spatial variations of temperature and salinity) as well as to anticipate the onset of large-scale climate anomalies (e.g., Indian Ocean Dipole, Southern Indian Ocean Dipole, SOD) that drive ocean dynamics in this area.
In the present experiment, a high priority was given to the collection of 5-m time series of depth and temperature data. The knowledge and experience gained during the first phase of the program will be used to refine the acquisition procedures as well as to better manage the necessary trade-offs between coverage, temporal resolution and transmitted information, resulting from limited Argos message size and limited battery capacity. In the next phases of STORM, more attention will therefore be devoted to tag programming in order to better exploit their capabilities. For example, a particularly interesting capability of the WC (and other advanced) tags, which was not used in this study, lies in their ability to reconstruct and transmit high-resolution (16-points) mean vertical profiles of depth and temperature over periods of 1 hour, which will be extremely useful for model verification purposes and general ocean studies.
During the forthcoming phases of the STORM project, efforts will also be done to carefully calibrate and intercalibrate hydrographic data obtained from the different tags to be deployed on sea turtles. SMRU tags, whose measurements have been thoroughly analyzed and calibrated when deployed on marine mammals (e.g.
Finally, the very recent decision of the Executive Committee of the GOOS Observations Coordination Group (OCG) to create the new sub-network “Animal Borne Ocean Sensors” (AniBOS) opens up new perspectives in terms of biology and oceanographic applications. By offering the possibility to disseminate high quality/frequency observations of physical oceanographic data in a standardized manner, AniBoS will significantly expand the use of biologging data for research and operational applications. In this respect, another important objective of STORM will be to ensure that collected ST datasets will be widely distributed to the ocean science community. Applications envisaged in the medium-term concern, in particular, the assimilation of ST-borne temperature and salinity observations into global and regional ocean models, as well as the provision of consolidated
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation, to any qualified researcher.
Ethical review and approval was not required for the animal study because sea turtles were all equipped by qualified personnel (from Kelonia care center and Institute of Ocean and Atmosphere) holding official accreditation to handle and equip these animals. The tags used in STORM meet all the requirements of the international conventions for the protection of sea turtles and were directly purchased from manufacturers specializing in marine biology and biologging. Written informed consent was obtained from the individual(s), and minor(s)’ legal guardian/next of kin, for the publication of any potentially identifiable images or data included in this article.
OB: writing, project leader, and data analysis. MD and PG: writing and data analysis. MB and SB: data analysis. SC: data collection. ER: writing and data analysis. AV: writing and expertise. All authors contributed to the article and approved the submitted version.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We would like to thank the following contributors for their collaboration: Mathieu Barret and the staff of the Kelonia Sea Turtle Care Center; Rui Coelho and the fisheries observers from the Institute for the Ocean and Atmosphere (IPMA, Olhão, Portugal) as well as Jérôme Bourjea (UMR Marbec, IFREMER Sète) for their collaboration in deploying a tag from a Portuguese longline fishing boat, south of Reunion Island; François-Xavier Mayer and the team of the NGO Cétamada (Sainte-Marie, Madagascar) for their priceless help in recovering the tag from the individual poached in Madagascar.
In addition to the time series of depth/temperature data used in this study, additional products such as histograms of “time at depth,” “time at temperature,” “minimum and maximum dive depth,” and “dive duration” over user-defined bins and time periods can also be obtained from WC tags to investigate the diving behavior of instrumented animals. Information derived from such histogram data is presented hereafter to provide general information on the diving behavior of STs SAMSON and BRICE over approximately 7 months (February 15–August 30 2019 vs. April 2–October 31 2019, respectively). The two animals shared approximately the same morphometric characteristics (see
Nine hundred eighty-four (resp. 892) histograms containing diving data aggregated over periods of 3 h have been transmitted by ST BRICE (resp. SAMSON), which represents an accumulated time period of ∼123 (resp. 112) days. This number of files correspond to ∼58% (resp. 56%) of the theoretical number that would be expected assuming an Argos transmission efficiency of 100%. The summary of diving data is given in
Diving behavior of sea turtles SAMSON and BRICE. Number of recorded dives >10 m classified by depth layers.
Depth (m) | Number of recorded dives |
||||||||||||
10–20 | 20–30 | 30–50 | 50–80 | 80–100 | 100–150 | 150–200 | 200–250 | 250–300 | >300 | Total | Mean depth (m) | Mean duration (min) | |
1,930 | 482 | 710 | 2,328 | 328 | 64 | 71 | 11 | 12 | 7 | 5,943 | 43.46 | 30.46 | |
4,298 | 574 | 601 | 1,379 | 480 | 117 | 122 | 24 | 30 | 12 | 7,637 | 38.71 | 25.6 |
Recorded data (
Time series of daily maximum dive depths (>10 m) for STs SAMSON (left panel, 15 Feb–30 Aug) and BRICE (right panel, 2 Apr–31 Oct). The position in latitude of the two animals at the beginning of each month is indicated on the top of each panel.
Tags deployed on STs TOM and LAZARE ceased emitting in June 2020, but tags deployed on STs TINA and INDIA are still active as of 10 August 2020.
Data can be downloaded at
Data can be downloaded at
Available from CMEMS at