A joint project by Australian and American marine scientists has found a new way to explore the ocean. Researchers have made an open-source framework, meant to transform how research image datasets can be managed, processed and shared. In 2024, the database set a new standard for collaboration, data accessibility and citizen science.

The program builds on previous projects, such as the Integrated Marine Observing System (IMOS) and the Understanding of Marine Imagery project (UMI project), which are opening the underwater world to both scientists and the general public. The data sources, despite being scientific endeavours, are all open source, with millions of images open to the public.

The databases are helping foster a sense of collaboration within the scientific community, with projects such as “Marimba.” Marimba required Australia’s national science agency, CSIRO, and a team at the Monterey Bay Aquarium Research Institute (MBARI) in California to work together for more than a year to develop the open source tool which was first shown to audiences in 2024.

When images taken hundreds of meters below the surface are processed, scientists can use Marimba to add geotags to the images, so they can be displayed on Google Earth. Alternatively, the images can be added to online platforms such as Seamap Australia and SQUIDLE+, where they are made available for the public and for scientists to use for viewing and further research.

“The explore map feature of SQUIDLE + is one of the ways we can access and explore images described and packaged by the new tool, Marimba,” said Franzis Althaus from the CSIRO in Hobart, who manages deep sea image data sets. She continues, “At this stage only one deep-sea data set from an expedition to seamounts (underwater mountains) south of Tasmania is available in SQUIDLE +. Marimba will help us package and publish many more such data sets.”

According to Chris Jackett from CSIRO’s Sustainable Marine Futures program, Marimba is designed in collaboration with MBARI to improve existing processes,“It streamlines the structuring, managing and processing of scientific image datasets while adhering to the globally recognised FAIR principles.” The FAIR principles – Findable, Accessible, Interoperable and Reusable – are a set of globally recognised standards for scientific research.

“In the vast ocean of scientific data, finding, accessing and utilising marine imagery has long been a challenge for researchers worldwide, compounded by diverse data formats, rapidly increasing volumes of imagery, and inconsistent workflows,” says Jackett.

“Since there wasn’t a system available that could efficiently process and package our scientific marine image datasets while adhering to FAIR principles – we built one.” Jackett said.

Marimba is able to process and show imagery from any source, “whether it’s from a Canon DSLR or a GoPro system. It can even be used for more specialised instruments like imaging flow cytometers.” As such, it holds “the potential to accelerate research and foster international collaboration,” Jackett says.