Q&A with Vincent Myers, a defence scientist for Defence Research and Development Canada
By Nick Walker
Sonar scans of the sea floor (Photos courtesy Defence Research and Development Canada)
On the technology
The Synthetic Aperture Sonar system that we used during the 2014 Victoria Strait Expedition grew out of a defence industrial research program here [at DRDC’s offices in Halifax], and was made by a company in St. John’s called Kraken Sonar Systems. Meanwhile, the Arctic Explorer — the massive yellow autonomous vehicle that carries the SAS — is made by a Vancouver-based company called International Submarine Engineering. So we have technology from both coasts in this, as well as a number of government operators and scientists. It’s altogether a great demonstration of Canadian technology: our SAS is among the very best sonar systems in the world in terms of resolution and design.
We have two of these underwater vehicles. They’re much larger than your typical sonar-carrying devices because they were custom-made for sonar mapping Canada’s continental shelf, under the ice in Canada’s Far North. This technology is just one part of this huge Canadian undertaking, called Project Cornerstone, which came out of the United Nations Convention on the Law of the Sea — an effort to map countries’ continental shelves. Canada’s project is ongoing, but both vehicles are not always surveying, so we said “Hey, we have an SAS program, so wouldn’t it be a great idea to use the same vehicle to carry it and do some other types of surveying?” The more stable your vehicle is, the better your SAS image (smaller vehicles are more likely to get tossed around in the ocean). This was a perfect combination for the search for the Franklin ships.
On other applications for SAS
We’re a defence research lab, so the application that we have in mind is detecting and classifying naval mines, such as those left over from the Second World War or the Cold War in the Baltic. There are a huge number of civilian applications for this too, of course, from charting the Arctic, to inspecting pipelines, to underwater archeology. This is a nice example of military technology that could now be useful in any number of civilian fields.
On the need for better charting in Canada’s Arctic
I remember on the 2014 Victoria Strait Expedition, sometimes we simply couldn’t go where we wanted to — not because of the ice, but because there wasn’t enough data. I still find it incredible to look at the charts we had for that region, because except for a few bays where in the 1890s someone must have let a lead line down, they had virtually no soundings. It’s mostly just blank, and I’d never seen that before. The need for bathymetric charting is absolutely there.
A large amount of the Arctic Ocean is still uncharted, in my experience, and that’s going to have to be done. The Arctic Explorer is one of the tools that could conceivably be used to do that. Of course, these are major undertakings — you’d also need ships and launches and so on — but the nice thing about this machine is that it’s unmanned, so it doubles your coverage.
On how the SAS system could help fill this need
The system that we have produces bathymetry as well as high-resolution imagery. The reason it does that is not because we necessarily wanted bathymetry — although that’s a nice corollary — we need the [depth data] in order to focus the SAS on the exact right spot and altitude. It’s so sensitive to slight misalignments in the signal that the image will become unfocused otherwise. For the best-possible image, we have to produce accurate bathymetry at the same time.
On how SAS differs from regular sonar
The image resolution from regular imaging sonar systems is proportional to the size of the sonar array being used. You get a constant high resolution for a short range, and then the sound naturally starts to spread out — so as the sound propagates the resolution gets worse. What people do, then, is try to make longer and longer arrays to keep the resolution up. Or they up the frequency, which presents a problem because higher frequencies don’t travel as far. It’s a constant tradeoff between frequency and array size.
SAS works differently. Because the system is moving with the vehicle in a forward direction, you take all the data that it gives you from all the different pings that its sending, and then you process them together as if you had one very long array. You’re “synthesizing” a long array or aperture. It’s the same thing that they do in space with RADARSAT-2, except underwater, of course. In the end you collect the data and process it to create an image — as opposed to regular sonar, which creates an image by stacking pings on a screen as the sonar moves.
On the importance of positioning
SAS only started to work maybe 10 years ago. Underwater, the sound waves, or pings, have a slow velocity (1,500 metres per second), whereas RADARSAT’s pulses move at the speed of light. So the motion of a satellite is not a big deal. For us, the slow-moving wave and the instability of the vehicle make it more difficult to focus the image. You have to know exactly where your array is at every ping — up to a fraction of a wavelength, which is in the order of a tenth of a millimeter. There’s no GPS in the world that accurate, and GPS doesn’t work underwater anyway, so you have to use the actual sonar data itself to navigate. You take two pings that are overlapping, and you cross-correlate them to figure out that you’ve moved relative to you last position, say, two centimetres forward, one sideways and one upwards. It’s like the autofocus on a camera, but much more complicated.
On the challenge of icy waters
The problem with ice is the safety of the underwater vehicle: you don’t want it to surface when there’s ice above. Had it become caught under an ice flow in Victoria Strait, we might have lost it forever. Not having charts or GPS, it becomes very complicated to plan routes. During the 2014 Franklin search, we had programmed the AUV to stay about 25 metres above the seafloor — at surveying altitude — but at one point it encountered a huge uncharted mound, which rose to much less than 25 metres from the ocean’s surface. So the vehicle came right to the surface, and we were fortunate that there was no ice there at that moment. Without charts or GPS you can’t really predict that it’s going to do that. Accommodating for the vehicle to work where there’s ice cover, especially when you’re moving from area to area as we were in the Victoria Strait Expedition, becomes very complicated.
On the “shapes” in these underwater images
We don’t have any ground truth on any of these phenomena. When it comes to the shapes and patterns that we’re seeing, we’re making our best guesses as to what they are. There’s a good chance that the circles that are appearing on some of our SAS images of the seafloor are gas pockets, but we haven’t been able to dive on them with an ROV [remotely operated underwater vehicle] to confirm. It’s been suggested that they could be depressions called “strudel scour,” which would be caused, in this case, by very salty water from melting sea ice (or icebergs made of sea ice) plopping down on the seabed. Based on my experience, however, I still think the best possibility is that these are gas pockets. The massive iceberg scours [which appear like deep ruts] are the easiest to explain, and they explain why the Parks Canada researchers were concerned that icebergs had obliterated either of Franklin’s ships. That’s why the SAS, which has 2½-centimetre resolution, is a perfect technology for finding bits of ships or artifacts.