Andrew Wade

Robots help create 3D map of subsea canyon

News

A research project at Whittard Canyon in the Bay of Biscay has used a combination of subsea robots and ship measurements to create a 3D map of the underwater habitat.

Rich cold-water coral reef was imaged by the the Isis ROV
Rich cold-water coral reef was imaged by the the Isis ROV

According to the National Oceanography Centre (NOC), which led the project, the canyon is England’s only deep-water Marine Conservation Zone. Until now, the mapping techniques used to chart its depths have not provided huge amounts of detail, but advances in robotics have meant that more accurate measurements can now be made.

“Our robot vehicles imaged rich communities of cold-water corals, clams, deep-sea oysters and their associated fauna, including a broad range of fish species,” said Dr Veerle Huvenne from the NOC, head of the five-week expedition.

“The morphology of this canyon is spectacular. We have mapped cliffs up to 150m high and 1.6km long, in some locations down to centimetre-scale resolution. This makes us the only group in the world who currently can image vertical cliffs in the deep sea in this way.”

The project employed a combination of methods to create a ‘nested map’. This involves building up a picture in layers, with more detailed maps stacked upon less detailed ones. First, echo-sounders on the research vessel RRS James Cook were used to create a 200km map of the canyon with a 50m pixel resolution. Following that, sideways echo-sounders on an Autosub6000 robot-sub mapped the vertical walls within the canyon, delivering a resolution of 3-5m per pixel. 

The final stage of the mapping process saw a remotely operated vehicle called Isis lowered on a tether from the RRS James Cook. This recorded high definition video and collected biological and geological samples from vertical and overhanging locations. Echo-sound data collected by Isis was also used to create the most detailed map of the three, with a resolution of 10-20cm.

The expedition was part of the CODEMAP project. It was funded by the European Research Council, and received additional support from the NERC MAREMAP programme and DEFRA.