Teams in both institutions explore the use of imaging and optical measurements to study water masses and biological organisms.

glider In OOV, and in the context of the Oceanographic Autonomous Observations group, optical sensors are used to measure several optical properties of sea water from which bio-geo-chemical ones are inferred (microscopic algae biomass, particle concentration and dimensions, dissolved organic matter content). Some of these sensors are already miniaturized and implemented on autonomous sampling platforms such as gliders. Gliders are torpedo-shaped instruments which ride oceanic currents to move in the water, collect data, and send it back to the laboratory when they come to the surface. The spatial resolution of gliders between two surfacing points (~1km) allow to capture process at sub-mesoscale (1-10km) which is a relevant scale to investigate coupling between biological and physical processes.

UVP The Underwater Video Profiler (UVP, Gorsky et al 2000a; 2000b) captures high resolution images along a vertical profile in the ocean. It takes pictures of particles and organisms from 300 microns to a few mm and collects physical characteristics of the water simultaneously. Images are then automatically processed to extract and classify elements of interest (organic particles, living organisms etc.). This automatic image processing system integrates with more traditional sampling tools such as fine-meshed plankton nets, because organisms hence captured can be scanned on a special scanner (the Zooscan) and the images can then be processed by the same software (Grosjean et al 2004).

ISIIS At RSMAS, the In Situ Ichthyoplankton Imaging System (ISIIS, Cowen and Guigand 2008) is a vehicle towed behind a ship which captures a stream of images while oscillating between 0 and 200 m depth. It ‘sees’ organisms from a few mm to a few cm and collects physical parameters simultaneously along its track. The image processing routines are currently being set up; feature extraction is almost complete while several classification methods are being tested (Tsechpenakis et al 2008).

DISC The Drifting In Situ Chamber (DISC, Paris et al. 2012, successor to the OWNFOR system, Paris et al 2008, Irisson et al 2009) is a circular, drifting instrument which allows observation of the behavior of live organisms by taking pictures of them in a behavioral arena, open to the surrounding water. In particular, position in the arena can be extracted from the images hence collected to study cardinal orientation. The DISC is actually an experimental platform on which additional instruments can be attached, such as hydro-phones or light sensors depending whether one wants to study orientation of the organisms with respect to sound or light for example.

Teams in both institutions use satellite imagery to get a quasi-synoptic view of surface waters which combine physical (sea surface temperature, sea surface height) and biological (water color) information.


Through this partnership, teams at the origin of these innovations will be able to collaborate more efficiently to :

In particular, the enormous quantity of images generated by the instruments presented above calls for a collaborative effort to develop a common, standard database platform, for archiving, image analysis and publication purposes. Such public databases foster further research on questions that were left unexplored by the original researchers and allow meta-analyses of several datasets. In addition, a key role of these databases could be to create a community of enthusiasts around the identification of open-sea marine organisms, through pictures. Such efforts have proved very successful in astronomy for example. In consequence, the metadata format and the database access API (Application Programming Interface), at least, need to be defined together to be interoperable.

Physical data such as those collected as part of the Argo project by profiling floats (nearly 3500 floats presently active), satellite imagery data of ocean surface, time series of observations at fixed station as well as models output are generally stored in centralized, universally accessible databases. Data collection practices in biological oceanography are dramatically evolving thanks to the availability of autonomous platforms and new sensors. As part of the partnership we hope to define standards for a future system of archiving and dissemination of quality-controlled data on par with its physical counterparts.