2007 report
Progress: 20%
a) Technical evaluation of the ITPs subsystems selected for DAMOCLES:
The Ice Tethered Platforms (ITPs) subsystems for DAMOCLES considered in this report, concern exclusively the CTD profilers. Actually there are two systems operating in the Arctic Ocean. One is developed by the Arctic Group at the Woods Hole Oceanographic Institution (PIs: John Toole, Andrey Proshutinsky and Richard Krishfield) in the USA. The other one is developed by METOCEAN (Canada). The WHOI system is derived from the MacLane CTD profiler. The METOCEAN system is using an ARGO float CTD profiler. The two systems are transmitting the data in near real time to a surface unit by using inductive modem along a steel cable. In both cases the surface unit is equipped with an Iridium transmitter to transfer the data to the laboratory on shore for immediate data processing.
On August 19, 2004 a WHOI-ITP-CTD-prototype profiler was deployed at about 77°N and 141°W from a 4m thick ice floe within the Canada Basin of the Arctic Ocean. The CTD profiler was installed on a weighted, plastic-jacketed wire rope suspended below the ice floe and used a traction drive system to move up and down the wire on a pre-programmed sampling schedule. At the completion of each one way profile, raw 1hz CTD data and engineering information were transmitted in turn from the underwater profiling module to the surface unit via the wire tether using an inductive modem and from the surface unit to a shore-side data server with over on Iridium satellite telephone link. On September 28, 2004 this ITP stopped transmitting over the Iridium telemetry link after 40 days of operation. These data are available in near real time via http://www.whoi.edu/itp/. Two new WHOI-ITP-CTD profilers were deployed during the summer of 2005 in the Beaufort Gyre and are still in operation.
The first METOCEAN system (the so-called POPS) operated by JAMSTEC (Japan) in connection with the NPOE (USA) was deployed on April 24, 2005 and transmitted CTD data for 22 days (until May 16, 2005) due to an unfortunate shortage in battery lifetime of the data logger. Nevertheless 14 CTD profiles from 10m below the surface down to 1000m depth were obtained by this first POPS (1 profile per day during the last 10 days). A second POPS was just deployed on April 21, 2006 at the North Pole and the first transmissions representing vertical CTD profiles from 10m depth down to 1000m depth are of excellent quality.
In the context of DAMOCLES, a POPS system was deployed during 24 hours from fast-ice in Storfjord (Svalbard) for testing the system in shallow waters but at a high repetition rate. These tests were conclusive. This POPS represents the first of an ensemble of five POPS being acquired for DAMOCLES. Three POPS were deployed during a cruise on board the Russian ice breaker Kapitan Dranitsyn from August 21 until September 15, 2006. Two additional POPS will be deployed during spring 2007 and four WHOI ITP/CTD profilers will be deployed during summer 2007.
During the training course organised in Brest on June 20 and 21, 2006, the two ITP-CTD profiling subsystems operated within the DAMOCLES project, were presented. Performances, similarities and differences of the two systems were discussed. The quality of the CTD data is quite comparable since both systems are operating with identical Seabird CTD sensors. The main differences come from the mechanical system used for profiling up and down. The WHOI system is using an electrical motor activating wheels rolling along the plastic jacketed steel cable. The METOCEAN system is using the ARGO float controlling buoyancy for profiling up and down at a lower speed (10 cm/s) compared to the WHOI system capable of profiling at 25 cm/s. The initial array operated during DAMOCLES will be composed of three POPS (September 2006) and will then be completed by two more POPS (spring 2007) and 4 WHOI CTD profilers (spring 2007). There will be additional CTD ITP autonomous profilers deployed during the DAMOCLES experiment in conjunction with the International Polar Year. Two ITPs will be installed by AWI from Polarstern during summer 2007 (SPACE). Few more ITPs will also be deployed by the WHOI Arctic group in order to complete and extend the DAMOCLES ITPS array in space and time.
Five DAMOCLES partners are directly contributing to this effort: IOPAN (2 POPS), UPMC (2 POPS), AARI (2 ITPs), SIO (2 ITPs), Tartu University (1 POPS). The unit price for each ITP CTD profilers have been fixed at about 65000 euros and each POPS at 60000 euros.
Ice Tethered Platforms specifications for the DAMOCLES Integrated Project
The DAMOCLES Integrated Project launched two calls for the development of nine Ice Tethered Platforms. The first 5 POPS type are composed of (1) a surface buoy equipped with meteorological sensors and (2) an underwater CTD unit profiling once a day for one year (possibly two) from 10m below the surface down to 1000m depth. The surface unit shall be equipped with an Iridium transmitter for data transfer in near real time, a GPS unit for localisation and an Argos transmitter as a back up. The 3 aerial transmitters (GPS, Iridium and Argos) should be -40°C proof. The Ocean profiles from -1000m up to -10m should be transferred via Iridium on a daily basis. The met data should be transmitted every hour. The CTD should be parked at -300m depth in the core of the Atlantic layer and sample CTD parameters every hour at this depth. Every profile should start with a lowering of the CTD from -300m down to -1000m followed by an ascent at an average speed of about 25 cm/s during which CTD parameters will be sampled every second. After reaching the minimum depth (-10m) the CTD profiler will return to the parking depth at -300m depth and record internally CTD parameters every 10 minutes at constant depth for the next 24 hours. There will be no data sampled during downcast.
Performances of the meteorological sensors:
Barometric sensor: pressure range 800 hPa to 1060 hPa for temperature -40°C up to 20°C. Resolution: 0.1 hPa.
Temperature sensor: temperature range -40°C up to 20°C resolution 0.1°C
After mainly three months in operation in the transpolar drift of the central Arctic Ocean, the three POPS provided more than 50 full depth (1000m) CTD profiles at a pace of one profile every other day and meteorological data transmitted every hours satisfactorily.
b) Damocles ULS Floats specifications
The ULS floats developed for Damocles (cf diagram) are intended to deliver information concerning the sea-ice thickness distribution at various locations over long period of time (> 1 year). The floats are freely drifting under sea-ice at constant pressure (isobaric). The float depths shall be selected between 50 and 100m under sea-ice in order to optimize (1) max accuracy for sea ice thickness detection (few centimetres) and (2) max detection range (>100km) for acoustic signal propagation at low frequency (780hz up to 1560hz) taking into account stratification of the water masses in the upper layer.
The ULS floats will be located under sea-ice during free drifts by triangulation from acoustic Ice Tethered Platforms (AITP), 6 times per day (every 4 hours).
The first test in situ of both ULS floats and AITPs will occur in STORFJORD (SVALBARD) using VAGABOND logistics during spring 2007. The first deployments of ULS floats and AITPs in the transpolar drift will occur during summer 2007 in conjunction with TARA and NP35 logistics. These operations will last until summer 2008 with an intermediate deployment of both ULS floats and AITPs during spring 2008 in conjunction with TARA and NP35 logistics operations.
General ULS floats specifications
The ULS floats are based on the PROVOR floats equipped with an internal buoyancy controlled device (hydraulic pump and bladders). The ULS is similar to the ASL transducer already in use on moorings under sea-ice by IOS (Canada). A high precision in situ pressure transducer (PAROSCIENTIFIC) is integrated with the ULS to provide in situ pressure and float depth within a few centimetres accuracy. The float is equipped with an acoustic SOFAR transmitter (1560Hz) to signal its position occasionally but in real time (once a day) and a RAFOS receiver (780Hz) to be located 6 times per day afterwards.
ULS isobaric floats will have the capacity for detecting sea-ice thickness in real time on occasions thanks to AITPs sending information concerning sea level atmospheric pressure by acoustic telemetry. The skylite detection algorithm installed inside the float will be based on:
1. ULS signal travel time compare to Paroscientific in situ pressure taking into account sound velocity based on in situ temperature measured by the float
2. Comparison of the last 100 measurements (1000s) to check for stability of surface geometry of the target (stability meaning that the target corresponds to a flat surface more relevant to an ice free surface rather than sea-ice except for very thin ice)
3. Atmospheric sea level pressure SLP in near real time since 1mb of SLP can offset sea-ice thickness detection by 1 cm.
Based on these 3 kinds of information, the ULS float will decide once per month or every other month approximately if conditions for surfacing are good (or not) in order to deliver information to the satellite (Iridium). It is quite clear that opportunities for surfacing will be more abundant in summer time rather than in winter time.
Detailed ULS float specifications
Operating depth 50m min and 100m max. ULS rate transmission 0.1Hz min and 1Hz max. ULS floats positioning 6 times per day (every 4 hours). Time window for ULS float listening up to 12 AITP SOFAR transmissions at 780hz will be about 1 hour very 4 hours. Each SOFAR AITP transmission will be shifted by 5 minutes from the previous transmission for AITP identification and to avoid confusion. At this point there is no need for the ULS float to know the position of the AITPs in real time since we are not navigating the ULS float like we will do for a glider for instance.
ULS float will transmit a 1560Hz SOFAR signal (10Hz modulation for 20s or 40s) once per day for real time positioning (outside the 1 hour 780Hz listening time window). The AITPs will be equipped with a 1560Hz RAFOS receiver to localize the ULS float once per day.
ULS floats will build up 10cm classes histogram up to 4m thick ice for sea-ice thickness distribution every hour assuming constant atmospheric sea level pressure arbitrarily assumed to be 1000mb (except when ULS float will need to decide to pop up to the surface or not. Then the real SLP will be transmitted to the ULS float by the nearest AITPs. Corrections for real SLP will be applied later on to the one hour Sea-ice thickness histogram based on an accurate knowledge of SLP over each ULS floats in near real time. There will be many surface stations deployed on sea-ice over a large area to infer SLP over each ULS float updated every 1 to 3 hours for meteorological purposes.
So every 10 seconds ULS float will measure (1) In situ pressure very accurately, (2) sound speed calculated from in situ temperature, (3) ULS travel time and deduce from 1, 2 and 3 sea-ice thickness non-corrected from SLP fixed arbitrarily at 1000mb
Every hour ULS float will build up the sea-ice thickness distribution histogram based on 1, 2 and 3. Every 4 hours, ULS float will also capture TOA (Time of Arrival) of SOAFR signal transmitted by AITPs (those AITPs still within the acoustic range pf the ULS float). We'll need 2 AITPS minimum, preferentially 3 AITPs to fix the ULS float. In addition ULS float will store a mean temperature every hour and keep track of time accurately for TOAs. All the raw data will be stored inside the ULS float on a data logger part of the acoustic modem provided by Aquatec.
Providers
ULS will be equipped with an ASL 420 khz type of transducer (Canada). 780Hz SOFAR and 1560Hz RAFOS transmitter and receiver including Transducers and Hydrophones will be provided by Seascan (USA). Acoustic modem (8 khz to 14 khz) will be provided by Aquatec both for ULS floats and AITPs in order to permit a downloading of the raw data without asking the ULS float to surface. This downloading capability will be integrated either in a self contained unit operated from surface or in sea glider operated at a later time.
The ULS float and the AITP integration and preliminary test will be provided by MARTEC and IFREMER. The final test (spring 2007) and field operations (summer 2007) will be undertaken by UPMC, MARTEC and IFREMER in cooperation with Vagabond, Tara and AARI (NP35) logistics.
We are planning to build up to 8 ULS floats and deployed half of them in summer 2007 and the other half during spring 2008. We are also planning to build up to 12 AITPs and deploy 6 during summer 2007 and the other half during spring 2008.
As far as budget is concerned UPMC is responsible for providing the acoustic components of the USLs including the ULS and the SOFAR and RAFOS components. AQUATEC is responsible for the acoustic modem. MARTEC and IFREMER are responsible for the development of the ULS float (integration of the PROVOR technology, the ULS/ASL technology, the SOFAR/RAFOS technology and the AQUATEC acoustic modem technology).
c) Technical evaluation of sea gliders under sea-ice
Gliders are autonomous underwater vehicles, designed to monitor for long time periods the interior of vast ocean areas at lower cost than oceanographic ships and moorings. Buoyancy control allows gliders vertical motions and additionally gliders use their shape and small fins to induce simultaneous horizontal motions. In summary, changing buoyancy together with the hydrodynamic structure allow gliders to carry out saw-tooth trajectories between the ocean surface and a defined depth along prescribed directions. In DAMOCLES, three gliders designed by ENSIETA will be fitted with acoustic modems and will be used to establish a link between the set of AITPs and the Lagrangian floats equipped with ULS. The gliders will also collect CTD data which will be transferred to transponders mounted on the AITPs. Recovery of the glider will be done after 6 months in ice free regions, by using either Argos fixes or Iridium modems.
Glider localization
In order to fulfill their mission, the gliders need to know their position. Normal GPS based localization is out of the question because to get a GPS fix, the gliders have to surface on an open water area and it is a rare exception. So, in an ice-covered area, the gliders get information about its position (its depth, its heading, its two-dimensional displacement) from internal sensors such as pressure sensor and attitude sensor and from acoustic messages sent by the AITPs and the drifting floats. Successive range measurements give independent information on the movement of the glider. This information can be used to improve the prediction of the movement and to estimate the glider position. The localization estimation algorithms task is to fuse all the measurements across different sensors and across different time instants for a sufficiently good estimate of the current position. These algorithms will be designed for gliders and a method was presented in the detailed report based on possibility theory which could be used. It would be interesting to compare this method with other algorithms proposed by HUT.
Glider navigation
As gliders must play the role of shuttles between the AITPs and drifting floats, they must reach a close neighborhood of AITPs or drifting floats for transferring data. Within this neighborhood, they will transmit the CTD profiles collected along the transects and other data collected from underwater drifting floats equipped with ULSs. Based on the localization estimate and mission specifications, the gliders have to navigate towards a particular float (to get data from it) or towards a particular AITP (to get data out to the users) or towards an open water area (when the mission time is over). The missions' specifications have been provided by a deployment plan which contains information such as initial deployment locations and estimated ice drifts of AITPs, initial deployment locations and estimate sea currents affecting the floats, initial deployment locations of the gliders and their operation plans. These mission specifications are regularly updated and provided to the glider via encoded SOFAR messages sent by the AITPS near the glider. The glider navigation can be divided into different phases:
§ Phase I: When the glider receives an AITP message, it estimates its own position using localization algorithms and then it steers towards the area where the float has been located at that given moment. Indeed, thanks to the AITP message, the glider receives new data about the latest estimated position of the float (this information is encoded to the SOFAR signal (780 Hz) transmitted by the nearby AITPs).
§ Phase II: When the glider detects the 1560 Hz signal sent by the float, the glider knows that it is in a close neighborhood of the float and it uses this signal to alter its course accordingly.
§ Phase III: When the glider gets close enough of the float, it switches on its high-frequency modem and sends short pulses periodically to the float.
§ Phase IV: As the float is able to detect this incoming signal, it takes the active role and starts to transmit a homing signal.
§ Phase V: As the glider is able to receive this homing signal, the data exchange from the float to the glider can be made. When the data transfer is successfully terminated, the high-frequency modems go into stand-by mode and the glider continues its operation.
§ Phase VI: When the glider wants to transmit the collected data (its own data and the float data), it has to reach a particular AITP (the closest AITP). Using the received SOFAR signals, the glider steers towards the area where the AITP has been located at that given moment.
§ Phase VII: When the glider gets close enough of the AITP, the glider uses its high-frequency modem to activate AITP's high-frequency modem.
§ Phase VIII: As the AITP is able to detect this incoming signal, it takes the active role and starts to transmit a homing signal.
§ Phase IX: As the glider is able to receive this homing signal, the data exchange from the glider to the AITP can be made. When the data transfer is successfully completed, the glider switches into receiving mode and continues its operation using other SOFAR messages sent by the AITP.
In order to simulate the glider and AITP trajectories in the Arctic, it is necessary to model the environment using physical data. Two different data sources can be used to model the Arctic coastline: Matlab high-resolution coastline database GSHHS which consists of a hierarchical set of databases at different resolutions and NGDC Coastline Extractor which provides coastline data files. The data source which can be used to take into account the Arctic bathymetry in the simulation of glider trajectory is ETOPO5. ETOPO5 is a collection of several gridded elevation data sets, combined into a single global coverage. The data sets are of low resolution and accuracy. ETOPO5 gridded elevation data for areas greater than 50 degrees north were provided to NSIDC by the National Geophysical Data Center. Two different data sources can be used to model the Arctic currents: the MERCATOR OCEAN global model (PSY3 resolution ¼°) which provides 3D eastward sea water velocity and 3D northward sea water velocity (1 output/day) in the Arctic ocean and the NERSC model (TOPAZ) which covers the Atlantic and Arctic basins and provides 2D eastward/northward velocity sea ice (area average) and 3D eastward/northward velocity sea water.
Trajectory simulation was done in MATLAB using Mapping Package M_MAP. M_MAP10 is a set of mapping tools which includes routines to project data in different spherical projections, a grid generation routine and a coastline and global elevation databases. This package has been used for simulation and graphical interface. The kinematic equations used for glider simulation have been solved using the Matlab solver ODE45 based on an Runge-Kutta (4,5) formula. To operate in a vertical sawtooth trajectory, the glider has to change its pitch when it reaches the 'surface' (Zmin) or the floor. This pitch change depends on the comparison between the bathymetry and the glider vertical position. If the glider overtakes the local depth or the limits of the sawtooth trajectory (Zmin and Zmax), pitch changes sign (positive for diving, negative else). The glider and the AITP drift. So, in order to follow AITP, it is necessary to change regularly the glider course. Heading changes were simulated under assumption that the glider position and the AITP position are known for each time step. To use these rules, it is necessary to estimate the glider position.
The conclusion was that Matlab is too slow to simulate the glider and AITP trajectories. That is why it was decided to make a simulator of the navigation programs using C language. The architecture of C program presents two principal units:
· SIMULATION which contains simulation programs: simulation of glider and AITP trajectories, simulation of acoustic propagation, simulation of AITP communication.
· ON-BOARD which contains only navigation and routing programs.
· These two units (which communicate via a network) are described in details in the report D8.3-4.
In order to validate the method described previously, it will be necessary to carry out some simulation tests to study the influences of different parameters (uncertainty, possibility distribution characteristics, etc). This report presents the first results, showing the glider simulated trajectory and the glider estimated trajectory (after defuzzification) with or without random noise and also the uncertainty evolution during 106 seconds. The uncertainty does not diverge and the estimated trajectory follows correctly the simulated trajectory. The maximum value of uncertainty is 600m (which corresponds to the uncertainty on AITP position in the grid). The maximum shift between simulated trajectory and estimated trajectory is 10km, which is a correct result.
The mission specifications depend on the deployment plan and define, for example, how to divide the gliders to serve certain floats and moorings. These specifications contain the instructions for individual gliders what to do and when. The main goal of these mission specifications is to complete the given tasks (i.e., to get as much data as possible to the operator) while protecting the expensive gliders at the same time. Deployment plan and mission specifications are very important for the successful use of gliders which need to be recovered and maintained after 6 months operation. It should also include procedures how to operate in case of malfunctions in the equipment or if some sort of re-deployment is needed for some of the AITPs due to unexpected ice flows.
The extreme under ice conditions set hard constraints for the design of the localization and navigation systems and the level of uncertainty is very high. This report presented a localization method which could be used in DAMOCLES project and which might be associated with other solutions. This report described work in progress, so, the final conceptual specification can only be given after extensive testing, not only in simulations but also in a real context. ENSIETA and HUT will cooperate tightly to find the best solution for glider localization and navigation.
d) Technical evaluation of bottom moored systems
Arrays of mooring in main gateways
Understanding the mechanisms that regulate the inflow of Atlantic water from the Atlantic through the Nordic Seas into the Arctic Ocean through the Barents Sea and Fram Strait, and its variation due to a mix of both local and remote forcing requires continuous monitoring of oceanic fluxes through the main gateways. To estimate the transport of mass, heat and salt to the Arctic Ocean, in the frame of WP8 we will continue bottom moored arrays that have been running successfully for about a decade (partly in EU-funded projects VEINS and ASOF-EC-N) and that have provided crucial insight in the propagation of signals from the North-Atlantic into the Arctic. These input mooring lines are located off Norway (62°N Svinøy; 2moorings, UoB), in Fram Strait (16 moorings; AWI, NPI) and across the Barents Sea Opening (5 moorings; IMR). Observations in the Barents Sea Opening and at the Svinøy section will be carried on using existing arrays with some required modifications while an optimally cost-effective observatory will be developed and first implemented in Fram Strait where intense mesoscale activity and the strongly variable recirculation of Atlantic water demand an innovative improvement of the existing system.
In 2006 the array of mooring in Fram Strait was deployed in the configuration worked out during the previous ASOF-N project. Altogether 16 moorings cover the whole section between 6°30'W and 8°40'E at the latitude 78°50'N. Additional two moorings deployed in the western part belong to the freshwater array (see Section 5). The eastern and central set of 12 moorings is operated by AWI, the western 4 moorings are served by NPI, all are exchanged on ayearly basis within similar time period. On a basis of the experience achieved during ASOF-N, all moorings are instrumented at standard levels: subsurface level (ca 50m), level within the Atlantic water (AW) layer (250m) and at the AW lower boundary (750m), the deep water level (ca 1500m where available) and the near-bottom layer. All moorings are instrumented with Aanderaa current meters (RCM7/8 and RCM11) which had proved to be the most reliable devices of previously used variety of current meters. All Aanderaa instruments are equipped with temperature sensors, most of them also with pressure sensors. At selected mooring the upper-looking Acoustic Doppler Current Profilers (ADCP) are also in use. In the subsurface layer temperature and salinity are additionally recorded by Seabird MicroCat or SeaCat sensors. To ensure the safe recovery all moorings are equipped with ARGOS satellite transmitters, Benthos transponders and double releasers, there are also iXSea transponders on deep moorings, allowing 3-dimensional positioning by the ship's POSIDONIA system.
In addition to 16 tall moorings, 6 Pressure Inverted Echo Sounders (PIES) were also deployed at the bottom along the Fram Strait section, in vicinity of mooring positions. Four of them were equipped with Aanderaa DCM (Doppler Current Meter) sensors. Besides measuring two-way travel time, all PIES hold also accurate pressure sensors.
According to recommendations for the optimal integrated observing system in Fram Strait, the system employing acoustic data transfer between moorings and NRT (near-real time) satellite data transfer was designed for the deployment in 2007. Three moorings in the eastern part of Fram Strait in the West Spitsbergen Current will be equipped with the long-range (up to 30 km) acoustic modems (HAM.node - Hydro-Acoustic Modem) developed by develogic GmbH. The additional, independent central mooring (CM) will be equipped with a profiling top (PT) including CTD and satellite transmitter to transmit the data from adjacent moorings once per day via Iridium communication. The profiling of the PT should be regulated by the noise level to avoid surfacing under bad weather conditions. A temperature threshold will be also set-up to avoid surfacing under the ice. The homing depth of the PT should be between 50 and 100 m. The profiling top will be operated by the underwater winch (UW) controlled by the central unit (CU). The profiling top and central unit will by developed by Optimare while the underwater winch was selected from available and already proved solutions. PT will include Seabird SBE41 CP CTD sensor and will measure the temperature and salinity profile during surfacing once per day. The Automatic Elevator System Type 3 manufactured by the Japanese company Nichiyu Giken Kogyo Ltd. will be used as the underwater winch. This system can operate to the maximum depth of 300m for approximately 1 year. To be implemented at the central mooring, the underwater winch will be equipped with additional buoyancy and controlled automatically by the central unit. A designed system including three acoustically linked moorings and the central mooring with PT and CU should be ready for deployment in 2007. In the second phase of the project the gliders should be able to collect data from the moorings and to transmit to satellite. This will serve as a back-up either in case of loss or failure of the CM or PT, or in case it can not be provided in time. The strong current events during the ASOF deployments, particularly in winter time under the stormy weather, resulted in significant dumping of whole moorings. To avoid those displacements, the moorings were redesigned by shifting existing buoyancy along the mooring or adding additional flotation. However, simulations with MDD revealed that the most of drag resulting in a strong dumping comes from the wires themselves. For the future DAMOCLES deployments the most vulnerable moorings will be also redesigned to use thinner cable, at least in the upper part of the mooring where currents is the strongest. Numerous simulations have been also done for the central mooring with the profiling top and central unit to find the best configuration (amount and shape) of required buoyancy (flotation) to keep the mooring stable even under the strongest current events.
In the Barents Sea Opening the DAMOCLES moorings were deployed when we recovered earlier mooring from previous projects June 2006. Near the Norwegian coast two bottom-mounted ADCPs in trawl proof frames were deployed. Then there are four moorings with Aanderaa current meters. The northernmost mooring as a downward looking ADCP from Aanderaa Instruments placed at about 50 m above bottom, and will measure the outflow of dense bottom water from the Barents Sea to the Norwegian Sea. This type of instrument is tried for the first in this section. The positions of the moorings are given below. The moorings will be recovered in summer 2007.
The overall goal in the Svinøy section was to develop a complete and sustainable, simple and robust upstream reference-system for monitoring the Atlantic inflow (AI) toward the Arctic Ocean. The two moorings in the eastern branch of the Norwegian Atlantic Current (NwAC) are now established as a bench-march concerning monitoring of the eastern NwAC through the EU-Damocles project till 2009. To establish a complete monitoring system as upstream reference for the Arctic, moored instrumentation with reference to the western branch is crucial. However, the western branch of the NwAC can hardly be intercepted by using moored current meters because of the meandering and unstable structure of the frontal jet. We have shown that the bulk baroclinic flux in the off-slope flow can be inferred from one density profile, at the eastern boundary. Thus we intend to resolve the overall fluxes associated with this branch by deploying a McLane moored profiler (MMP) at the 1000m isobath, i.e. at the western rim of the slope current where the flow is almost vanishing. Hopefully a MMP will be funded by the Norwegian Research Council in 2007. A MMP equipped with CTD and acoustic current meter will run regularly along a mooring line of about 600 m with 4-6 hour intervals, resulting in time series of density profiles. A MMP has already been deployed and tested in the Svinøy section over a 4-month period, with promising results. Applying the thermal wind equation on a series of density profiles, will results in time series of volume flux estimates associated with the baroclinic jet. The reference velocity problem in the thermal wind equation will be addressed by deploying a current meter mooring over the 2000m isobath where the average frontal jet is located. This method will be evaluated against independent measurements from an array of moored Pressure Inverted Echo Sounders (PIES) through the IPY project iAOOS from 2007.
Boundary moorings
To provide better knowledge of the exchange processes between the boundary and the interior of the Arctic Ocean with the ultimate goal of understanding and quantifying how much this exchange affects the ice cover the work focuses on moored observations along the circum-Arctic boundary. The strongest subsurface currents of the Arctic Ocean are constrained to follow topographic features and moored arrays measuring velocity, temperature and salinity will be used to quantify volume, heat and salinity fluxes along the main transport paths. DAMOCLES was not be able to equip the Arctic with the necessary number of transport arrays, but will work in collaboration with US, Russian and Canadian efforts (notably the NABOS and CABOS experiments) to ensure a good coverage at two sites along the Eurasian slope: a mooring array at roughly 32°E north of Spitsbergen and a mooring array in the Laptev Sea. Moored CTD/velocity profilers in the boundary current arrays will obtain high horizontal and vertical resolution of thermohaline intrusions and mesoscale activity.
During the Russian icebreaker Kapitan Dranitsyn cruise the bottom mooring was deployed by IOPAS. This work was done with cooperation with the Nansen and Amundsen Basins Observational System (NABOS) and the mooring belongs to the NABOS net. The mooring is situated over the Arctic Ocean continental slope, at the Laptev Sea, north of Severnaya Zemlya. The bottom depth of the mooring is 1544 m at the position 80º44.94'N and 103º29.91'E. The mooring location in the pathway of the Svalbard Branch of the Atlantic inflow and construction of mooring allows to investigate properties of the Atlantic Water as well as a flow and its temporal variability in the entire water column. In the frame of the same cruise the profiling mooring was deployed also north of Svalbard as a part of existing NABOS network. The McLane profiler (MMP) with a acoustic current meter (ACM) and a was deployed at the depth of 2000 m next to the conventional multiyear NABOS mooring located at the depth of 1010m. Both moorings are located at the continental slope in the boundary current and will be recovered in autumn 2007. The McLane profiler was selected for this mooring on a basis of earlier experience achieved during NABOS deployments since 2001.
Storfjorden moored array
The strength and properties of the Storfjorden plume will be observed through the DAMOCLES period by two ADCP moorings, one at the sill and one at the slope, equipped with MicroCat salinity sensors. An existing trawl-proof bottom-mounted ADCP system has been used to monitor the overflow at the Storfjorden sill (76º 58.08¢N, 019º 14.95¢ E), since summer 2003 under the Polar Ocean Climate Processes (ProClim) project funded by the Norwegian Research Council. ProClim will be terminated by the end of 2006 and the moored system will be maintained during DAMOCLES. During a cruise in 2006, the ADCP system was deployed at the sill on 13 August 2006 from R.V. Håkon Mosby. The scheduled recovery is in July 2007. This will constitute the first year-long data set at the site in the body of DAMOCLES.
A second system was proposed under DAMOCLES to be deployed at the shelf-break off Sørkapp of Spitsbergen on the path of the Storfjorden overflow. From the allocated budget of WP8, one Workhorse Sentinel 300 kHz ADCP (with 2x256 MB memory card, 3 extra battery packs and an external battery case) and two SBE37SM Microcats were purchased early in 2006.
The existing bottom-mounted system comprises an RDI Workhorse Sentinel 300kHz ADCP with external battery case, a SBE37SM MicroCat and a releaser installed in an aluminum trawl-proof frame, attached to a concrete block of 2.5x2.5x0.37 m dimensions. The weight of the concrete block is about 2.5 (1.6) tonnes in air (water). The frame with instruments (acoustic release, ADCP, Microcat and the battery pack) and floatation elements installed is 300 kg in air. Overall height of the installation is 86 cm. The system was successfully deployed/recovered during ProClim and produced reliable current profiles and bottom temperature record for three years. Microcat data (temperature, conductivity, pressure) is available only since December 2005. An evaluation of the salinity record derived from the Microcat (conducted recently) showed that the absolute magnitude as well as possibly the variability of the salinity is not accurately measured with the SBE37SM, which is not apumped system. This is both due to the fouling by sediments etc. and not sufficient flushing of the conductivity cell. To overcome this problem we have sent two SBE37SMs to the manufacturer to convert into systems with integral pumps. We anticipate more accurate salinity measurements from 2007 and on. As the temperature of the Storfjorden overflow water is always nearly at freezing point, it is crucial to accurately sample the salinity.
The maintenance of the above described trawl-proof system is costly. Transport and deployment are difficult and the large concrete block has to be released at the site for each deployment. A new concrete block (of significant cost) has to be ordered (with lengthy delivery schedule) before each deployment. To overcome these issues, at GFI/UIB, we designed a more sustainable new system which was scheduled to be constructed for the summer 2006 deployment, but was delayed. The new system will comprise the same instrumentations but a new stainless steel (to avoid any fouling of the ADCP compass) frame of order 1 ton of total weight which will be deployed and recovered as a whole. Transport will be more practical as the system components can be dismounted. The frame will comprise a buoyancy element to which cable of sufficient length and strength is carefully spooled. The buoyancy element is connected to the release and during recovery the buoyancy element will surface upon release. This will then be picked up from a ship and the main body of the frame will be recovered by the ship crane. Necessary buoyancy elements and Kevlar cable is already purchased and the construction is underway. In December 2006, we will conduct a test of the spooled buoyancy element in 200m deep water in Byfjorden in Bergen. The complete system will then be tested in Sognefjorden in February 2007. Two such frames will be constructed: one for the sill and the other for the shelf break deployment.
Lacking a trawl-proof system for the shelf-break deployment in summer 2006, we attempted to deploy the new ADCP system in a traditional setting (not trawl-proof). During the same cruise of the sill-ADCP deployment, we successfully deployed the shelf-break mooring. However, upon inspecting of the reported trawl-activity at the site and further spotting a trawl-vessel close to the site during deployment, we only allowed ourselves for cruise-time deployment, and recovered the system after 9 days. After anticipated success of the tests of the new trawl-proof system, we will be able to deploy the system before May 2007, from a ship-of-opportunity. If the tests prove unsuccessful, the previous version of the system used with success at the sill will be prepared for the shelf-break deployment, while improvements will be made on the new design.
Tomography array in Fram Strait
The detail description is given in the previous chapter (Task 8.2)
Freshwater array in western Fram Strait
In September 2005, as a part of the EU funded ASOF-N project, a total of six moorings were installed across the East Greenland Current (EGC) and the shelf region at 78°50' N. Four moorings were installed across the EGC, instrumented with a total of twelve current meters, four temperature/salinity (TS) sensors, four upward looking sonars (ULS) measuring sea ice draft and four Doppler Current Meters (DCMs) measuring ice drift velocity. In addition two moorings were deployed on the East Greenland Shelf, as an effort to measure the freshwater flux there. Of these moorings one was a tube mooring instrumented with two TS sensors. The second mooring on the shelf, close to the tube mooring, was instrumented with an Acoustic Doppler Current Profiler (ADCP) measuring the currents over a range of 150 meter. In this way T, S and currents should be measured also on the shelf, in the second year of deployments of moorings on the East Greenland Shelf. This setup was the basis for the final data collection with respect to freshwater fluxes during the ASOF-N project, which ended in 2006, but also provides the first freshwater flux data points for DAMOCLES.
The moorings deployed in 2005 were recovered on the first DAMOCLES cruise on Lance in September 2006. At the same time new instruments was deployed, which, in a similar way, will be recovered in 2007. The instrumentation deployed in 2006 is identical to the 2005 deployment, except that the tube is replaced with an Aanderaa TS string.
This version of the array, in our opinion, nearly represents the optimal setup for observing freshwater fluxes through Fram Strait with conventional moorings. It captures the bulk of the freshwater in the EGC in a cost effective way, and provides some basic information about the stratification and velocities on the shelf. A second tube mooring is desirable on the inner part of the shelf, featuring TS sensors and an ADCP. Due to the vast extent of the East Greenland shelf at this latitude, this second tube would add valuable information about the cross-shelf gradient in salinity and freshwater content. However, this latter mooring is still missing from the array, since the ice conditions on this part of the shelf generally does not allow a ship like Lance to penetrate the pack and perform mooring work.
Due to the sea ice constantly drifting by, it is difficult to perform observations of salinity in the very upper layers (0-50 m). At the same time this is where the bulk of the freshwater transport takes place. Having the moorings closer to surface than 50 m would represent ahazard to the instrumentation due to the risk of collisions with deep pressure ridges and icebergs. Due to the Upward Looking Sonars, which can not operate with obstacles in their acoustic beams, we are prevented from having TS strings with weak links above the ULS/RDCP top of the moorings. One solution which could remedy this shortcoming is profiling CTDs, which would profile the water column at preset intervals not interfering with the upward looking sonars. This solution is prohibitively expensive, though. The present solution is to use stratification as provided by climatology, but corrected for our actual TS reading at nominally 50 m. The CTD sections performed during cruises to the region also provide valuable information about the upper layer stratification.
