Printversion of this page
PDF-Version of this page



Because of their low cost, autonomy and capability for long-range, extended duration deployments, underwater gliders constitute a new class of autonomous underwater vehicles. With wings and tail, they glide through the ocean without propellers or thrusters, controlling their buoyancy and attitude using internal actuators. Gliders propel themselves by changing buoyancy and using wings to produce forward motion. Buoyancy is changed by varying the vehicle volume. Gliders travel from place to place by a series of upwards and downwards glides in sawtooth pattern. Their flight paths naturally sample the ocean both vertically and horizontally. Through their use of buoyancy propulsion and low power design, gliders are capable of long-range and high endurance missions. They move at very low speed, travelling at about one half knot, but have ranges on the order   of hundreds or thousands kilometers and deployment periods of the order of months. The weak stratification of subpolar oceans also results in longer ranges as less energy is needed to penetrate through the pycnocline.



Fig.1: Seaglider predeployement test on deck (left), transfer in the boat to the deployment position (middle) and deployed (right) from Polarstern in Fram Strait (photos: A. Beszczynska-Möller).


Gliders carry a wide range of oceanographic sensors, collecting the data at a fraction of the cost of a hydrographic survey or fixed mooring. They can measure pressure, temperature, salinity, dissolved oxygen, chlorophyll fluorescence, optical backscatter, depth averaged currents and other quantities when traversing a chosen path or maintaining position as a virtual mooring. After each dive, a glider come to the surface to obtain a GPS fix, while between surface fixes its position is calculated by a dead reckoning. Gliders use satellite data telemetry to receive commands and send collected data in the near-real time.

The Seaglider is the type of glider which was developed by the Applied Physics Laboratory and School of Oceanography of the University of Washington (Eriksen et. al., 2001). It has a mass of 52 kg (in air), the fairing is 1.8 m long with 30 cm diameter, the wing span is 1 m, the same is the standard antenna mast length. It travels with the speed 0.1-0.45 m/s (~22 km or 12 nm per day @ typical speed 0.25 m/s), the glide angle is between 1:3.5 and 1:1 (16-45°). To power internal actuators and oceanographic sensors the vehicle carries lithium battery packs. The Seaglider is able to stem ocean currents up to 40 cm/s and can survey along a transect or hold against the current, acting as a virtual mooring. In deep water a typical full dive (down to 1000 m) lasts roughly 8 hours, with 3 dives per day the horizontal resolution is approx. 4 nm. Vertical resolution of measurements is 0.5-1 m. For near-bottom dives the Seaglider is equipped with bathymetry maps and a bottom avoidance altimeter. The Seaglider measures temperature and conductivity, (SBE custom sensor), pressure (Paine Corporation sensor), fluorescence and optical backscatter (WETlabs custom ECO-BB2F puck), and dissolved oxygen (SBE 43F optode). The vertical resolution of measurements is 0.5-1 m. Surface and depth averaged current velocities are inferred from two consecutive surfacing positions and the hydrodynamic glider flight model. Currently Seagliders are commercially available from the iRobot company, located in Bedford, MA.




Fig. 2: Seaglider components (left photo, Seaglider Fabrications Center, Seattle) and preparation of the glider for deployment in summer 2011 (right photo, A. Beszczynska-Möller).


Since 2008 the Alfred Wegener Institute has been employing Seagliders for oceanographic measurements in the northern Fram Strait. The Seaglider has been used extensively for long-endurance oceanographic sampling in locations spanning the globe. It has achieved mission endurance of 9 months (740 dives to 1000 m depth), travelling at half a knot on half a watt of power while traversing thousands of kilometers, though this figure depends strongly on speed, operating depth and ambient stratification. It has a low drag shape and the isopycnal hull, matching the compressibility of seawater, what further extends vehicle range. Under water Seaglider navigates using dead reckoning while at the surface it uses GPS fixes to get  a position, and receives instructions and transmits data via the Iridium satellite system.


Table 1: Summary of the AWI glider missions in the northern Fram Strait



Fig. 3: Temperature and salinity section measured by Seaglider SG127 in Fram Strait during the summer 2011 mission (right panel) and map of the trackline on derived ocean current vectors.


To read more about Seaglider specification and under water operations please visit: 

-Seaglider summary, animations and current missions at the PL UWA website... (link to
- Seaglider description and pictures at the SFC website... (link to
- Seaglider information at the iRobot website...


Acoustically navigated under-ice Seaglider


In ice-covered areas gliders cannot rise to the surface to use satellite based navigation and data telemetry. For the Arctic, it is therefore necessary to develop underwater acoustic navigation system (Lee & Gobat, 2005). The Alfred Wegener Institute is collaborating with the Integrative Observational Platforms Group (link to (APL-UW) led by Craig Lee to adopt and extend the under-ice Seaglider and acoustic navigation technology used in Davis Strait for Fram Strait. The first winter-long Seaglider deployment in Davis Strait from September 2008 to February 2009, performed under the US NSF project 'An Innovative Observational Network for Critical Arctic Gateways' (link to proved feasibility of extended under-ice mission where Seagliders navigate using acoustic signals received from moored RAFOS sound sources. In ACOBAR, both conventional RAFOS sound sources and tomographic sources provide RAFOS navigation signals. In the next step, the possibility to use a broadband tomographic signal for navigation will be evaluated.



Fig. 4: Deployment of Seaglider MK544 from Polarstern in the northern Fram Strait (left and middle) and Seaglider SG127 recovered after the summer mission 2011 when several under ice dives were performed (Photos: A. Beszczynska-Möller).


Multipurpose acoustic system for gliders in Fram Strait


Under the ACOBAR project, AWI is collaborating with the IOP APL-UW group (PI Craig Lee) to adopt and extend the under-ice glider and acoustic navigation technology used in Davis Strait for Fram Strait. Taking advantage of the ice-capable Seaglider technology and procedures developed by APL-UW, Seagliders equipped with the RAFOS hardware and capable of under-ice navigation were deployed in Fram Strait for the test summer missions in 2009-2011 and the winter mission is under preparation. The system for acoustic navigation in Fram Strait consists of moored 260 Hz RAFOS sources and tomographic sources providing RAFOS transmissions, and Seagliders equipped with a Benthos RAFOS-2 hydrophone, Seascan clock and receiver modules. All navigation processing is done in real-time onboard the glider. Source-glider ranges are computed from RAFOS receptions using the depth averaged sound speed, and the glider position, corrected for a Doppler shift, is obtained by trilateration. Ice-capable gliders also include significant enhancements for extended, fully autonomous operation, including a multi-tier system for determining the presence of overhead ice.



Fig. 5: A map of Fram Strait Observatory with RAFOS and tomographic sources indicated.


Seaglider receptions from 260 Hz sources in open water were tested by APL-UW group during North Pacific RAFOS Ranging Test in 2005 with ranges up to 2000 km. In partially ice-covered Fram Strait more difficult propagation conditions result in significantly shorter ranges. During all missions with available RAFOS transmissions (except for the winter 2010 when the glider RAFOS receiver failed), gliders collected acoustic receptions and calculated navigational solutions for testing purposes, remaining mostly in the open water. Only during the summer 2011 mission, Seaglider SG127 went for a short excursion under the ice and used RAFOS positioning to navigate acoustically. The maximum ranges of RAFOS signals received by Seagliders in Fram Strait varied between 50 and up to 300 km, depending on the type of sound source (RAFOS signals were transmitted by different types of sources, including two RAFOS systems and RAFOS from tomographic sources), sea ice extent between a source and a glider and the type of RAFOS receiver carried by a glider. For the recent glider mission in 2011, 72% of acoustically obtained positions had an error less than 10 km while nearly half of the positions were accurate up to 5 km. The glider missions carried in 2010 and 2011 indicate that for Seagliders equipped with the new RAFOS-2 receivers, a sufficient coverage of the western Fram Strait glider section can be achieved with a network of 4-5 RAFOS sources, under the condition that transmission distances comparable to the observed range can be provided.


Fig. 6: Nominal ranges (green circles) of RAFOS sources (black stars indicate RAFOS sources and green dots show tomographic sources) deployed in Fram Strait in 2011 overlaid on the sea ice concentration on August 20, 2011 (AMSR-E data provided by Bremen University).


Fig. 7: Ranges of RAFOS receptions collected by the Seaglider MK544 in Fram Strait during the autumn mission 2011. Correlations higher than 60 indicate the RAFOS receptions with quality sufficient for acoustic positioning.


An accuracy of the acoustic navigation can be increased by using broadband tomography signals. The study by Duda et al. (2006) shows that the errors of several kilometres, resulting from the use of narrowband RAFOS signals for basin-wide navigation, can be reduced to less than 50 m by using broadband acoustic tomography signals. In ACOBAR such a multipurpose acoustic system was deployed in Fram Strait in 2010. Acoustic signal sweep with 190-290 Hz were simultaneously transmitted by the triangular array of acoustic tomography sources every 3-4 hours. To take advantage of the available tomographic array, feasibility of using the broadband tomographic signal for acoustic glider navigation will be evaluated.

Printversion of this page
PDF-Version of this page