One really fascinating aspect of sonar that leads to risks/collisions like this, is that there are ‘blind spots’ in the ocean. Because sound waves pass differently through water of different temperature and salinity and both of those change on gradients in the ocean, there are mathematical solutions to where a sub would be totally undetectable given a ship’s position and radar frequency. Because these locations are obviously useful for subs, two+ can sometimes end up in the same warm pocket, which obviously increases collision chances from a naive hundred-cubed-km search space
So the submarines can't use active sonar, because their number one priority is to avoid detection... but then these subs are all parked in the same "parking spots" in the ocean. If the favored parking spots for submarines are so rare and small that collisions happen more than once, surely that undermines their stated number one priority of remaining hidden? If someone wanted to locate enemy submarines, surely they could just send a bunch of underwater drones to swim through these parking spots?
There are no parking spots. Ballistic missile submarines conducting deterrence patrols are assigned to large areas covering hundreds of square nautical miles. The sub then maneuvers inside that patrol area to remain as hidden as possible. They almost never stop as it's easier to maintain depth control with at least a few knots of steerageway. So a collision in open ocean is theoretically possible but highly unlikely.
The real risk of collisions comes with transiting through constricted navigational channels, typically when leaving or entering port. Those channels force a lot of vessels into a small area. There have also been a few incidents where submarines failed to do a thorough surface search before surfacing and collided with smaller boats.
During the Cold War there were supposedly a few incidents where NATO attack submarines were following Soviet submarines too closely and ended up colliding.
> There are no parking spots. Ballistic missile submarines conducting deterrence patrols are assigned to large areas covering hundreds of square nautical miles ... So a collision in open ocean is theoretically possible but highly unlikely.
It's misleading to describe collisions as "theoretically possible" when we have evidence that these collisions have occurred. The fact that collisions have occurred makes me think that these submarines are not randomly located in a large area where it would be difficult to locate them.
You missed the point. The collision covered in this article occurred relatively close to port, not during an open ocean patrol.
A ballistic missile submarine's position isn't completely random, but they do move around somewhat randomly to reduce the risk of detection. They intentionally avoid repeating movement patterns or lingering in the same spot. And this also reduces the risk from spies on land. If a spy sees the patrol orders he'll only know a large general area but won't precisely know the submarine's position at any time.
> You missed the point. The collision covered in this article occurred relatively close to port, not during an open ocean patrol.
So you're saying the collision _didn't_ occur at a time when the sub was trying to stay hidden, you're saying that the operators could have / should have used active sonar? Is that it?
Look, either the sub was trying to stay hidden at the time it accidentally collided with another submarine, or it wasn't trying to stay hidden. In either case it didn't do a good job. If it wasn't trying to stay hidden, then it shouldn't have collided. If it was trying to stay hidden, then its location was not as unpredictable as it should have been.
That the collision happened close to port is weird in itself. France has its SSBNs stationed south of Brest; the UK to the west of Glasgow. One would assume the individual routes to the patrol areas would be rather different. What would the HMS Vanguard do less than a days travel from Brest?
> large areas covering hundreds of square nautical miles.
That's...not very large.
> The sub then maneuvers inside that patrol area to remain as hidden as possible.
Given that the same concerns will drive “as hidden as possible” concerns in any sub, that further narrows the space. And unless the decisions made within that space are made in a truly random manner which approaches all equally good options equally (and tolerates a loose enough sense of “equally good” not to excessively narrow the space to start with), common cognitive biases, etc., will result in common decisions.
Sounds like the FAA altitude problem all over again.
As altimeters got more accurate, the odds that you were flying at exactly 28000 feet went up. If an air traffic controller fucks up and puts two planes at 28000ft, they can now collide because they're ±10ft instead of ±200ft. That was the RCA of that mid-air collision over Brazil a while back.
The more you try to control a thing the more problems you can have. A parking lot for stealth ships is... I don't want to say dumb but confusing for sure.
> The more you try to control a thing the more problems you can have. A parking lot for stealth ships is... I don't want to say dumb but confusing for sure.
These days there are special collision avoidance systems on big commercial planes:
TCAS II provides the pilot with specific instructions on how to avoid the conflict with traffic. These instructions are known as a “Resolution Advisory” (RA) and may instruct the pilot to descend, climb, or adjust vertical speed. TCAS II systems are also able to communicate with each other to ensure that the RA provided to each aircraft maximizes separation.
The footrest is below the white screen with the navigational chart on it. The radio and transponder panel is the black screen to the right of that, with the yellow and green text on it. The 0512 is your transponder code (squawk); the button to the left of that and down one row turns it off.
This resulted in the Legacy losing its TCAS ability, and the 737 it collided with being unable to see the Legacy.
TCAS isn't built off of ADS-B (it existed long before ADS-B was ever conceived), but it's the same basic idea (mode c / mode s transponders) and probably has the same theoretical security vulnerabilities.
There's a pilot at the controls so that the technology doesn't have to be perfectly secure.
If you want to walk around broadcasting fake ADS-B or Mode C signals, someone will probably find you in hours, and the only effect it will have is maybe reducing the throughput of the busiest airports if the weather happens to be particularly marginal that day.
You don’t need drones. Planes and helicopters already drop sonar buoys for the exact reason you stated.
A better solution would be for an escort when they get into crowded lanes. Surface ships can use active sonar to scan for other subs while being some what close to their sub. The only problem is other listeners can pick up on what is reflecting off your active sonar as well. Also the subs try to avoid ever being detected because they can save your signature for comparison later.
It's not like any navy has spare surface warships with active sonar available to escort submarines through constricted navigational channels. There are also environmental concerns with active sonar due to impacts on marine life, especially cetaceans. During peacetime it's usually only allowed in designated exercise areas.
I have sonar on my kayak. You can certainly use a power that is not going to hurt sea creatures. A lot of Navy ships including submarines have off the shelf sonar as well as military sonar. I’ve seen a pic of a sub with Raymarine radar they use when surfaced in channels. The subs don’t want to use their own military grade stuff because then the enemies have the frequency signature so they use off the shelf for navigating in public areas.
That's simply not how it works. The active sonar on your kayak has nowhere near enough power or resolution to detect a military submarine at sufficient range for reliable collision avoidance. Due to the inverse square law, active sonars have to put out a huge amount of power to detect submarines at any useful range. This is a proven risk to marine life.
The escort ships would know exactly where the sub they are escorting is. They could easy see the bottom or a massive blob passing underneath (an unknown sub) or even an undersea mountain. You don’t need enough resolution to tell if it’s a sub or even a whale you just need enough to say watch out.
My middle of the road Garmin can see 1,100 feet deep in salt and side scan 500 feet in either direction. You can see bottom features and schooling fish. The higher end stuff is so much better.
You're really missing the point. No navy would ever assign one of their limited number of surface warships to that mission. There aren't enough to go around. And if a ship detects a massive blob passing underneath it's already too late. 1,100 feet isn't nearly enough range for collision avoidance; a ballistic missile submarine can't turn on a dime and a collision risk can come from any direction including astern. None of your suggestions are even remotely practical in the real world.
When submarines actually need to be sure of avoiding collisions they just run on the surface and follow the same rules of the road as any other vessel.
I suspect that wil421's idea is that the "near-harbor escort" job does not need anything near so grand as a surface warship. A low-end tugboat or similar would more than suffice.
No the escort ships do not know exactly where the sub is.
On my first submarine duty, we nearly surfaced right in the path of our escort. We had to dive hard to avoid getting run over.
“Clearing baffles” is not an exact science.
If the surface escort (whether a fancy warship, or a modest Harbor Patrol boat) was running a cheap commercial sonar set in active mode, and the sub's orders were "stay right under your escort, so long as its course seems reasonable" - with details of 'reasonable' spelled out in written orders - would that be easy to execute without incident?
Longer answer:
A cheap commercial sonar isn’t going to find a submarine, even if you know where you’re looking. Submarines are incredible stealthy, and the ocean greatly effects sonar effectiveness, especially from the surface.
Due to how it works, surface sonar is nearly useless when trying to locate a submarine.
A popular joke was for us to surface right in the middle of a fleet during war games, or sneak up behind them and take pictures of people smoking and send them to the surface ship’s captain just to mess with them.
With things like towed sonar arrays, and sonar bouys, the playing field can be evened out a bit, but in all the war games I was in, the surface guys could not find us even we told them exactly where we were.
Now imagine that, with an inexperienced submarine crew doing testing, and a surface ship practically incapable of keeping track of us…
“CON, ESM, EMERGENCY DEEP! Hold high signal strength contact bearing 087!”
I'm thinking more of a busy shipping area, and the essential mission is only to provide a surface exclusion zone, which a submarine could take advantage of as it (say) enters / leaves a harbor. The sonar's job is to broadcast a "your (say) Coast Guard Auxiliary moving exclusion zone is right here" signal to the sub's passive listening equipment. The CG Auxiliary guys up top don't even know if a real submarine is in the area - vs. a practice mission, or maybe the Navy is sailing a cheap decoy past some Russian fishing boat that's lingering where it ain't welcome.
(From exDM69's comment, I'll figure that the active sonar's power needs to be dialed way down. The theory that a sub's fancy listening gear could get a good fix on a sonar signal that's too weak to give a usable reflection seems hand-wave plausible to a landlubber.)
That would avoid incident but the sub might as well run surfaced with active sonar on.
The active sonar from the escort would reveal the sub it is escorting to any passive observers. It would need a directional sonar (not a cheap commercial sonar) and avoiding prop noise that can also reflect off the sub.
And if all subs were escorted this way, the navies of adversaries would quickly pick up this habit.
How are the drones going to detect them apart from randomly running into them? Simpler approach is for an adversary to just cruise attack boats through these areas, which you would want to do anyway since once you detect them you want to sink them in wartime.
If the passive sonar filling the entire nose of a huge submarine can't hear another submarine, why would a much smaller sonar in the nose of a drone be able to detect a submarine?
The submarine with the larger more sensitive instruments in a quieter environment ought to be able to detect sound levels below that which their active opposition can.
If X amount of noise-energy hits the submarine, it only reflects Y<X of that energy, and only Z<Y of that reflection goes in the right direction to be detected by the opposition. The noise should be louder at the submarine, than at the thing detecting the submarine.
Both of these weigh in favor of the submarine detecting the searcher before the searcher detects the submarine.
No, nothing I was talking about was related to the speed of sound in water (which is where I assume you're getting 0.66 seconds/km, since it's about right). I agree the speed of signal propagation seems unlikely to matter much in most situations.
My point is in this scenario there's no appreciable difference between "if you can hear them, they've already found you" and "if you can hear them, they'll find you in 40 seconds"
The observation that they are "detecting the searcher before the searcher detects the submarine" is academic and inconsequential when we're talking about such a negligible amount of time between detections.
But there is an appreciable difference between "if you can hear them, they've already found you", and "if you can hear them, they might never find you because you can see them looking from farther away than they can see, and they might just be outright searching the wrong area".
All my arguments went to the latter kind of problem, not the former. Your 40 seconds number seems to be based on the speed of signal propagation, not the range at which things can be detected.
Even in the event where they do find you, the arguments I gave would suggest that it would take vastly longer than the speed of signal propogation to do so, because they would have to move closer to you. Giving you substantially more notice than 40 seconds.
I was taking it for granted that knowing the location of an unmanned, expendable sonar buoy or probe before you're discovered is not tactically important.
Now that you have me thinking, it would clearly allow you to stay out of detection range if you had the good fortune for their pings to initiate within your detection range but outside theirs.
I'm fully willing to admit I'm talking out my ass here, but does it not make sense to assume that both a) the active drones are going to be much faster than a submarine trying to maintain silence, and b) there are likely to be many drones converging from different directions if they're searching a known submarine parking spot? Still seems like advantage drones to me
I'm not more of an expert of submarine tactics than you are, I was just taking issue with the physical claim that they already know where you are by the time you hear them.
The path loss for active sonar is, logarithmically, twice the path loss for passive sonar. If the signal is 0 dB at 1 meter, it will be -40 dB at 10 km away (barring SOFAR, a regions these submarines probably try to avoid, and ignoring the reflections off the surface and seafloor), and the reflection (assuming an isotropic reflector) will be -80 dB. At 100 km, it's -50 dB and -100 dB.
So, if both sides can detect a -50 dB signal, the passive-sonar target will be able to detect the active-sonar probe 100 km away, while the active-sonar target will be able to detect the reflection from the passive-sonar reflector 0.316 km away. Or maybe 0.315. And that's assuming perfect discrimination.
(By transmitting a signal you can barely detect, you can reduce this discrepancy: if you reduce the signal amplitude by 30 dB at 1 m in this example, the probe can detect the target at 10 m, or maybe 11 m, but the target can't detect it until they're within 100 m.)
However, this massive advantage for passive sonar only really true if it's just as easy for the target to recognize the signal as it is for the probe. In fact, the probe has a huge advantage: it knows what the signal was. If it's just an ordinary chirp, this is of no help, because that's easy to recognize, unless you confuse it for a humpback whale song or something. But if it's Gaussian white noise, it's impossible for a single hydrophone to distinguish from the background noise of the ocean, except by amplitude. Advantage: probe.
But that's a single hydrophone. In fact, though, a point source of white noise is the easiest thing for a phased-array beamforming passive-sonar system to localize. The Triomphant is 138 m long, so if you stuck hydrophones all along its length, you could detect directional sources of sound with an angular precision of about λ/d, the one-dimensional Airy limit. The λ for your active sonar needs to be chosen to be small with respect to the targets you're looking at so it doesn't just diffract around them, so this is probably on the order of 0.1 radians, about 5 degrees.
Two can play that game, though, or thousands. The network of tiny drones can form a single ocean-sized phased array by squirting data back and forth over short-range ultrasound or laser links.
But guess who already has ocean-sized phased arrays of hydrophones for passive sonar?
Re >> "the French Ministry of Defense reported that the submarine had suffered a collision with an “an immersed object (probably a container).”"
and
Re >> "At some point, the two navies compared notes"
I get that ballistic submarines want to stay secret and use passive sonar, instead of active sonar. But I also would have thought - if a submarine mysteriously ran into something under water that they did not expect to be there, that they'd start taking a more "active" approach to figuring out what was going on. It'd be like if I tried to walk blindfolded through a large & empty conference room: if I bumped into a whole wall after walking only 3 or 4 steps, I'd be very confused!
The trouble is whom and how to ask. The submarine world is a giant, international cat and mouse game, and asking around gives away operational details that may involve compromising other things (e.g., special operations deployments) if you say "at exactly X coordinates at time T we bumped into something. Anybody know?"
Also, since the dawn of militaries, much of what occurs is posturing. You want the other side to think you are big, bad, competent professionals. "We ran into something and we don't know what" makes you less superhuman in the eyes of the public and potential adversaries.
It sounds like both subs took damage around their sonar systems, so it's possible they couldn't use active sonar after the collision even if they wanted to.
Surprisingly, no side was assigned the blame for the collision. Unlike air traffic or surface navy accident.
One had a conning tower damage, the other had a bow damage; both were supposedly slow-moving at the moment. Is there such thing as right of way for subs?
How do you exercise right of way over another submarine that you don't know the position of?
The whole point of submarines is that they're stealthy. Active sonor would give away their position. They're basically swimming blind. The only way this all works is the ocean is big enough that random collisions are statistically rare - but there's not much actually preventing them.
It did appear just like that. Shouldn't there be some proximity sensors to avert collisions with massive bodies? After all in the aftermath both subs were rendered mission incapable.
Low power high freq radio waves? These should rapidly attenuate underwater, so not much of giveaway, but could be enough for proximity scans around the sub.
Is there some spectrum of emissions from an even stealthy sub, some kind of "local presence"? The immediate massive obstacles would result in distortions to such a self-presence profile. So the sensors could be trained to detect such local distortions, almost like some fish have a lateral line of nerves supposedly helping detect current flows etc. [1].
Ultrasound, magnetic, eddy-current metal detection, conductivity detection by conducting current through water, red light, VNIR, antineutrino detectors...
Subs avoid terrain using charts and dead reckoning, including inertial navigation systems. They typically come to periscope depth occasionally to get a navigational fix. In 2005 a US submarine allided with an underwater mountain due to a navigational error.
There is little or no light at typical operating depths. Lidar can work to a limited extent in some water conditions but it's not something to rely upon for collision avoidance.
I remember (barely) from a diver friend that colors disappear below 20 meters (everything appears blue, unless a bright light is shone), and sun light disappears completely below 70 meters. You'd assume that an optical system with flood lights could theoretically work, but then it would be pretty simple to monitor for light at these depths (for counter-intelligence/retaliatory purposes)
No right of way, but there are movements called “baffle clearing” where you check your blind spots before moving.
Due to the fickle nature of the mistress that is the ocean, sometimes even properly executed movements can still be dangerous.
each carrying 4 and 6 nuclear warheads respectively. Losing such apocalyptic firepower on the ocean floor would have been a catastrophe. However, nuclear warheads are not susceptible to “going off” as a result of a collision.
What is the catastrophe if this had happened? The weapons won't explode. Is it the danger that they'll leak their radioactive fuel, or that they'll be captured by some enemy power before they can be retrieved (if they can be retrieved at all)?
Its a little weird that the catastrophe described is about the warheads, not the loss of the whole crew. Though I'm sure it's about a billion dollar loss of hardware as well.
Pie in the sky ideas here... but how far is active sonar effective for? Can you launch a probe that enables active sonar some 50-100 miles away, and radios the results back? Or would the depth prevent any type of radio communication...or is active sonar not that long-range?
This whole field is really fascinating to think about.
The effective range of active sonar varies tremendously based on many factors: power output, microphone sensitivity, signal processing, bottom contour, water temperature, ambient noise, etc. There is no typical range. Aircraft can drop active sonobuoys but they're very expensive and in limited supply. Use of active sonar is restricted in many areas due to impact on marine life.
Water blocks all radio traffic except for extremely low frequency. The bandwidth on that is too low to use as an effective navigation aid, and transmissions can only be sent from large facilities on land. Some submarines can also send up floating antennas on towed buoys but they prefer not to do so due to the risks of detection and entanglement.
Even more amazing, only a month later (or before) two satellites collided in LEO. Literally astronomical odds, and the two incidents happened so soon after one another as well.
Lot of dismissive reactions to these questions, but there actually are non sound related signals we use to detect submarines and there is absolutely research on how to do that from space.
This article is great, but just read the "Signal Processing" section if you're interested in space.
https://www.aspistrategist.org.au/prospects-for-game-changer...
Interesting article. Also, we are now able to reconstruct a conversation by analyzing the movement of the glass on a window to the room where people are talking. It doesn't seem absolutely out of reach that we could some day detect huge underwater vessels from analyzing surface perturbations.
"reconstruct conversation glass vibration" brings up a lot of relevant results from a 2014 story from MIT. Sounds like it works in principle with any medium that can have its visible vibrations monitored (bag of chips is the main example, but they talk about other options such as plants or a glass of water).
Not sure if they are, but I do know that the US Navy funded a lot of oceanographic research into predicting and understanding bioluminescence of marine organisms for this very reason. The surface waters of the North Atlantic have massive spring blooms of phytoplankton, which are sometimes bioluminescent, and could certainly give away the presence of a large subsurface vessel, possibly from space.
next time you're out on the open ocean check how far down into the water you can see, and then check out this graph [0] of electromagnetic radiation absorption of water: visible light is the best case scenario.
there's a reason sonar is acoustic and not light based
This is fairly open secrets to the degree that it is secret at all so it probably also is no big secret that passive and active countermeasures are used to prevent these techniques.
The US Navy has researched the possibility of detecting shallow submarines from space since at least the 1990s by looking for subtle wake turbulence on the surface. The results are classified so we don't know if they were ever successful. It's not completely impossible, but the signal to noise ratio would be extremely low.
Any detection from space would have to use an optics system to detect light bouncing off the submarine. This light could be either the naturally occurring visible light, or some form of active system (probably emitting radar waves).
Unfortunately light can't penetrate very far into water, so this idea is a non-starter. Light can barely make it 200m into the ocean (meaning 100m depth for light to make a round-trip and back for detection), and subs can go as deep as 300m.
Any satellite-based detection system would have to rely on catching the sub while on the surface. I'm positive this is being attempted constantly, but catching a sub at depth is an entirely different manner.
Using what technology? I can imagine in some future there being the means, but I find it hard to believe we have that level of tech available to the sort of packages we can aloft into orbit now.
Subs would be detectable using magnetometers, except that they have electromagnets in them to hide their magnetometer footprint. Still, this is not perfect.
Large subs have somewhat detectable gravitometric footprints that could be observed from space, indeed.
These subs are large, metallic, full of electric equipment, and heavy. They have magnetic and gravitational effects on their surroundings. They also have electromagnets to help hide their magnetic signatures. A network of very sensitive satelites with very precise clocks could be used to detect gravitational anomalies -- maybe, I'm not sure what kind of precision would be needed, or if that is achievable.
Closed cycle steam generation plants aren't going to exhaust much. They can't filter seawater fast enough for use and you couldn't run seawater through those systems w/o messing them up.
Subs must necessarily produce heat. That heat has to go into the surrounding (usually very cold) water. This is unavoidable. Those nuclear reactors are cooled with that surrounding water, though obviously through heat exchangers.
Let me get this straight - there's a technique for detecting changes in the surface temperature of the water from a submarine that's probably hundreds of meters below the surface?
https://m.youtube.com/watch?v=AqqaYs7LjlM