The obvious thing is to build decoys which look like warships for the initial scanners (at long enough distances, pretty much anything...) and trick the enemy into arranging the interferometer to scan those mysterious maneuvering disturbances.

Are there any particular ways of determining what the maximum sensitivity would be, for a given exposure time? It might be useful in terms of determining just how quickly one could sweep a detection cone over a patch of sky.


Sorry, I'm guilty of imprecise use of language. I was using "slow down" as short-hand for "stop doing any big or fast maneuvers relative to each other".


I wasn't even thinking of relativistic speeds.

Well, it might be impossible to sneak up to someone who is monitoring the sky with your 200 m space cruiser, but what about sneaking up with missiles. For example you want to destroy enemy military installation on Mars. You fire dozens of cold missiles from mass drivers from base disguised as mining outpost in asteorid belt on an intercept course to Mars and fire up the rocket motor only few hundred km from target for final attack run.
How close it might be possible to sneak before enemy spots your drifting missiles? Let`s assume those missiles are 10 m long and 1 m in diameter.

Well, I forsee another problem if you don't actively cool the side facing the sun. The side facing the sun will warm up to an equilibrium temperature that depends on distance and albedo of the objects. High albedo means it warms up less. Higher albedos mean that it's going to be brighter on the visual spectrum, however. You can also cool the side facing the sun, but you're going to have to be careful which direction you eject the heat, or you're going to be visible from that.

---

"preemptive killing of cops might not be such a bad idea from a personal saftey[sic] standpoint..." --Keevan Colton
"There's a word for bias you can't see: Yours." -- William Saletan

How would this all effect your use of missiles and other weaponry in space? Unless you're close up, your missiles will probably have to have sensor systems of their own to compensate for the movement of the enemy ship, unless you're maintaining a constant communication between them and the mother ship for targeting. Lasers and mass drivers presumably don't have the same problem, but you're still stuck with trying to hit a moving, distant target.

I'm guessing that that means that space combat beyond a few light-seconds between moving ships would be a non-starter.

---

"You can't hammer tin into iron, no matter how hard you beat it, but that doesn't mean that tin is worthless."
-Jon Snow, A Game of Thrones

"I prefer my history dead. The dead sort is written in ink, the living in blood."
-Rodrik Greyjoy, A Song of Ice and Fire

Long range terrestrial missiles already combine command guidance for mid-course update with an independent terminal seeker head. There's a sharp limit to the size and cost of sensors you can put on a missile.



Realistic mass drivers will likely throw at less than 1% of c, so command-guided projectiles (likely just cold gas thrusters) would be highly useful in hitting anything at significant ranges, if you can build them to take the launch acceleration.



Even a few light seconds would make hitting with lasers and particle beams (and that's assuming you can collimate them at such distances) very difficult if the enemy is introducing small random variations into their maneuvering thrust to throw off your predictive targeting.

OK, so your miniscule 1 metre wide missile frontal cross-section is going to be just as visible as a natural meteor or asteroid of the same size. So what? You're acting as if an inert 1 metre asteroid would be easy to detect at astronomical range.

Just because it immediately followed Sky Captain's post doesn't mean that it's a response to his post. The real response to his post is: doesn't matter, since the point where it needs to fire up the rocket for the final attack run is the point where fast defense missiles can take it out. If it doesn't fire up a rocket, then it's probably not going to come close enough over astronomical distances to be able to hit. Also, radar is probably going to be effective well before that point.

Inert 200m asteroids are relatively easy to detect at astronomical ranges. People find them all the time with fairly small telescopes. There's going to be a dedicated instrument for search for them that uses only a 1.8m telescope (actually 4 of them, but not connected in an interferrometric system).

---

"preemptive killing of cops might not be such a bad idea from a personal saftey[sic] standpoint..." --Keevan Colton
"There's a word for bias you can't see: Yours." -- William Saletan

It seems to me that far more astronomical objects are probably detected by reflection of sunlight than by passive IR emissions, so I'm afraid the line "people find them all the time" doesn't strike me as sufficient justification to assume that nobody can evade detection with a 150K hull at astronomical ranges no matter what stealth technologies he employs. Just what is a maximum realistic sensitivity?

Let's try some numbers: suppose the missile is a blackbody radiator at 150K: I'm too lazy to look up numbers right now but IIRC, the Stefan-Boltzmann constant is 5.67E-8 W/m2·K4. So we'd be looking at a luminosity of 28 W/m², even if we assume emissivity of 1. If it's using some kind of stealth coating that brings its emissivity down to, say, 0.25, then we're talking about roughly 7 W/m². Assuming the missile head radiates equally in all directions, at a range of 10 million km the light intensity would drop to 5.7E-21 W/m². That's pretty damned faint. You would need a 100 metre wide telescope collecting for hours in perfect optical conditions to pick up even 1 picojoule of light. Of course, that's assuming a point-source: if there's a big flat surface pointed at the receiver it becomes easier, but still, I don't see it as the easy feat that you seem to think it is. Not unless they conveniently put a big flat mirror on the front of the ship. And you have to already know where to point your telescope!

In any case, the "stealth is useless in space" people are quite frankly full of shit. Whatever the detection range is, you could cut the detection range for IR emissions in half by cutting your surface emissivity in half. If you can pick him up before he picks you up, that's an advantage.

It might still work as a means to launch a surprise attack especially if the defender do not expect an attack because the general locations of all enemy ships are known and none of them are on intercept course. Success or failure depends how far missiles are spotted. If the missiles are detected only few hundred km away when engines are fired and are incoming with speed of 15 - 20 km/s it might simply be not enough time to react especially if defender is caught off guard.

I think everyone agrees if you fire up your multi terwatt fusion engine of your space warship you are going to be detected even over astronomical distances, but for small inert objects there might be a decent chance to infiltrate undetected.

Even if you have a fusion drive with a ridiculously massive output in the hundreds of terawatts, space is huge. Unless you're mapping the entire sky at the right time you get twitchy, you may still miss it, and the ranges that have been mentioned are actually pretty insignificant. Tens of millions of kilometres is nothing for the Solar system, given that Neptune is billions of kilometres from the Sun. In all probability hiding a ship which generates hundreds of terawatts against the blackness of space would be an insurmountable obstacle (along with managing its heat), but you still need to be looking at it to see it.

Let's examine the problem of minimum detectability. A starship will probably have two basic long-range sensory systems:

A) Radio antennae: A hostile starship will be a significant source of radio and microwaves (if active,) and large radio receiver surfaces are much easier to build and deploy than large receiver surfaces for lower portions of the EM spectrum.

B) IR/optical telescopes: A hostile starship will tend to glow brightly in the IR portion of the spectrum, and will also reflect ambient light. If it's got a large, powerful engine, and that engine is active, we can expect to see the drive plume as well.

You could speculate that a starship might also carry an X-ray/gamma ray telescope, since matter/antimatter annihilation produces well-characterized gamma rays. A large enough starship could carry a neutrino detector, since just about any conceivable high-density energy generation scheme will produce copious neutrinos as a byproduct. However, the highly penetrative nature of neutrinos, and the fact that the cosmos is loaded with neutrino sources would suggest that the signal-to-noise ratio of a neutrino telescope is going to be abysmally low. Gravity wave detectors are also out for much the same reasons, although both of these instruments may make a certain amount of sense aboard fixed early-warning platforms.

So let's examine the optical telescope. Our starship's optical telescopes aren't going to be very big, since huge mirrors with the fast focal-ratios needed for wide-field imaging are difficult to accurately figure, large telescopes require large housings, and tend not to be very easy to swing around. And we want the widest field possible.

Fortunately, NASA's provided us with an ideal example of a long-range, wide-field imaging telescope. The Kepler Mission features a wide-field telescope with about 1 meter of effective aperture. It feeds photons to a 95 megapixel camera watching a ten degree wide swath of the sky. For an optical telescope, this is a stupendously wide field of view. It works out to be 105 deg2. However, the area it must cover is 20,626.48 deg2. A 'mere' 196 images. Depending on how long your exposure time is, and the time required to slew the telescope over to the next patch, this might take a while.

An extremely sensitive sensor, such as the one used on the New Horizons spacecraft's LORRI imager picked up Pluto in a 1 second exposure. Pluto can be modeled as a 4.2 TW light source (Average albedo of 0.575, radius of 1151 km,) for an irradiance of just 1.89x10-8 W/km2 at the 4.2 billion kilometer distance that New Horizons was at when it first imaged Pluto. The New Horizons imager has an aperture of a mere 0.2 meters. Meaning Pluto illuminated one pixel of the detector with 2.38x10-15 W of power. Our starship-mounted telescope has about 25x the light gathering area, so it could match this feat at roughly five times the distance.

So, we could take 1 second exposures and be golden, right? Well, not exactly. We want to reduce the noise we get in our images. Some we reduce by circulating coolant through our telescope. Much of it we'll reduce by stacking multiple exposures, so we can mathematically eliminate the noise. So we'll take 15 exposures, then bump the telescope to the next patch of sky and wait however long it takes for the vibrations to damp out. We'll assume this takes another 15 seconds. It'd probably take much less, but this gives us 30 seconds of time spent on each patch of the sky. So our single telescope will take 98 minutes to scan a hemisphere's worth of sky. Put four on the top side of the hull, and we could be done in 25 minutes. Bear in mind that the sensor we're talking about is tuned to a very low-light environment. Less-sensitive sensors will require longer exposure times. If we took a modern off-the-shelf astronomical imaging CCD which requires around 450 seconds of total exposure time (30 seconds per raw frame x 15 exposures to stack) to image similarly dim objects with similarly fast optical systems, it would take over a day for our one telescope to do this "all-sky" scan.

So this tells us we could spot every object the size of Pluto within a 21 billion kilometer radius, and could do it in under 100 minutes if our starship had an especially cheap sensor suite. How about something smaller, like a Star Destroyer? Take two admirals, Jamie and Adam. Jamie would like to find Adam's ISD before Adam can get close enough to mess Jamie's shit up. (I've picked the ISD because it can be modeled as a white-painted triangle with an area of 0.812 km2. Conveniently, the albedo of white paint is 0.5 to 0.9, and Star Destroyer white is on the lower end of this . . . so we'll arbitrarily give it the same albedo as Pluto.)

So, if we start Adam out at the orbit of Pluto, his ISD looks like a 409 kilowatt light-bulb. Plug in the minimum irradiance derived above, and he's detectable at a mere 6.56 million kilometers away.

Since there's nothing of merit out at Pluto, let's start him somewhere closer, say the orbit of Mars. Now Adam's ISD looks like a 286 megawatt light-bulb, and could be seen by our sensor at 173.5 million kilometers away. However, his apparent size at that distance is 0.0019 arc-seconds. The optics of a 1 meter telescope won't resolve a point-source of light to any better than 0.25 arc-seconds. We'll see him, but as a dim point of light.

Now let's add in his thermal emissions. Let's assume that Adam is trying to be clever and is trying to slip in with engines and reactors cold. However, because his crew likes living, he's got a base temperature of 273.15K for a black body irradiance of 315.64 W/m2. In the far infrared, his ISD looks like a 256 megawatt bulb, assuming the side facing Jamie happens to be all radiator. The distance he becomes detectable in the far-IR is actually a bit closer than the distance he becomes detectable in visible. Useless at Mars, but out at Pluto, he'll be detectable over 130 million kilometers further away in far-IR than he is in visible. If he's got some freaky cooling and insulation, along with magic stealth that gets him down to Darth Wong's postulated 7 W/m2, he still looks like a 5.7 megawatt bulb in far-IR, and our telescope sensor will spot him at 24.5 million kilometers off.

Astronomically speaking, these are pretty short distances. If Adam drops out of hyperspace around Pluto and accelerates towards Jamie at Mars, he's going to have enough burn time to be going fast enough to force an engagement by the time Jamie can pick him up. And remember, at these distances, Adam's ship is going to appear to be a star-like point in our survey telescope. If he's coming head-on to Jamie, we'll have to filter him out from all the other dim star-like points out there. So we may not actually catch Adam until a few iterations of the scan after we might've first spotted him, at which point, his growing brightness will make him stand out.

Scaling for detection range is usually not that linear. I don’t know how the hell you’d cut IR emissions without some sever penalty in mass for insulation and cooling gear. Real life anti IR coatings work well, but only in an atmosphere because all they do is converted IR energy into different wavelengths which are easily absorbed by the air.

---

"This cult of special forces is as sensible as to form a Royal Corps of Tree Climbers and say that no soldier who does not wear its green hat with a bunch of oak leaves stuck in it should be expected to climb a tree"
— Field Marshal William Slim 1956

Current state of the art is to be able to scan 6000 square degrees of sky down to apparent magnitude 24 over a single night(Pan-STARRS). 68 of these installations would result in a revisit time of an hour. ~4100 installations would result in a continuous staring system (exposures are approximately 1 minute long). Naturally, there would be times when parts of sensors would be overloaded due to a brighter object (such as the moon).

Apparent magnitude 24 is a flux of ~7e-18 W/m2. Assuming I did my math correctly. 4 TW over 10 billion km is still around 3e-15 W/m2. For reference: if the Earth didn't massively swamp the signal, the space shuttle launching would be 7 times stronger of a signal than would be the minimum of being detected. Also for reference: said sensitivity is massively better than Wong thinks the sensitivity is. It's close to a femtojoule of energy required for detection. And this is today's technology.

I'll admit the Pan-STARRS is a visible light telescope, but even if a IR scope is 1000x less sensitive, it's still a detection out to 10 thousand km, given the same assumptions that Wong is making for emission in the IR band. Also, you're wrong about cutting detection range in half by cutting emissivity in half. Since detection range would follow the square law, cutting emissivity by half would reduce range by ~30%. Of course, emissivity is roughly inverse to albedo. By reducing your emissivity, you're raising your albedo (unless you go for a highly polished chrome exterior, anyway), making it easier to spot you. by visible light.

---

"preemptive killing of cops might not be such a bad idea from a personal saftey[sic] standpoint..." --Keevan Colton
"There's a word for bias you can't see: Yours." -- William Saletan

As far as the claim that body heat & life support would give away a ship, how about making most of the voyage in a cryogenic form of suspended animation ? You'd still show up when the ship fires up the reactor, then warms up and reanimates the crew, but that could let you stealth in close enough that the enemy couldn't avoid action or get help to them in time. And robots or downloaded human intelligences wouldn't need life support at all.

I've also noticed that discussions of stealth in space usually seem to assume a solar system like the modern one; an uninhabited wilderness. The ability to detect that something is artificial ( but little else ) at great range is going to be a lot less useful for identifying a warship in a solar system filled with thousands or even millions of ships, drones, cargo pods, stations, and so on.


Unless it's a drone firing the laser; that would only tell them where the drone was.


Here's a scenario that occurred to me. You have two sides about equal in power. The aggressor launches a fleet in a major assault using mass drivers ( so no drive plume, which I understand gives away the mass ), one large enough that the enemy would have to commit most of its forces to stop. It ALSO launches, say, five fleets of drones that can't be told at long range from actual ships. The real fleet and the drone fleets are all sent at separate targets; targets far enough away that it's physically impossible for the defending side to get it's mobile forces to them in time to defend against more than one attack. That means the defending force has only a one in five chance to be anywhere it can do any good.

---

"There are two novels that can change a bookish fourteen-year old's life: The Lord of the Rings and Atlas Shrugged. One is a childish fantasy that often engenders a lifelong obsession with its unbelievable heroes, leading to an emotionally stunted, socially crippled adulthood, unable to deal with the real world. The other, of course, involves orcs." - John Rogers

About sensitivity and scanning times: in a solar system like ours, most faraway (>1 A.U.) asteroidal or planetary bodies would look close to the plane of the ecliptic, and in a universe with no FTL these would make the easiest staging areas, so wouldn't it make sense to disproportionately have your most sensitive devices spend most of their time scanning a few degrees above and below that plane? You may detect most of what you're looking for in that plane. Although a thoughtful military would make a point of building bases far outside the plane of the ecliptic for exactly that reason.

That's a very interesting question; I don't know enough to answer it fully. Overall, we can scale known systems to get a feel for things, as GrandMasterTerwynn is doing, but there's an issue of quantization that can put further limits.

Suppose there is a blackbody of temperature T. From Planck's law, the spectral number density should be
[1] n(ν,T) = [2ν²/c²]/[exp(hν/kT)-1] (s-1m-2sr-2Hz-1)
Integrating over all frequencies, assuming a planar infinitesimal face emitting to a hemisphere (π sr)
[2] N(T) = [4πk³T³/c²h³]ζ(3) = bT³ (s-1m-2),
where b = 1.521E15/(m²sK³).

The typical sci-fi long, narrow ship shape actually makes sense, as we would want to minimize the cross-section. Assuming we have active cooling that dumps heat to the back of the ship so that the visible parts are kept at 100K, at 1AU distance we would have about 5.4E-3 photons per second per m² of emitter per m² of detector. With too few photons, the detector might not even distinguish it from noise, although with a bit more it might deduce that the particular patch of sky deserves more scrutiny if it has data on the astronomical objects that should be visible in that region.

Relativistic effects are ignored here. They have the unfortunate effect of concentrating the emissions toward the front of the ship and Doppler-shifting them to be more energetic. This would force the ship crews to be rather patient.

Hooray more informative posts.

What I'm getting generally so far: strategic stealth is possible, but tactical stealth is not. In a tactical situation around effective weapons range (a few light seconds max) you're going to be spotted. But as long as you're not maneuvering like crazy with main thrusters, you can very likely conceal broad moves.

Now I see it would make sense to have two kinds of sensor systems. One with a capability to quickly scan the whole sky looking for drive plumes and abnormally moving faint IR sources and another one for detailed imaging once the anomaly is found. Basically once your quick scanning scopes have found something that potentially could be a hostile ship or missile you point your superduper interferometer array at it to get a closer more detailed look.

Also in a solar system where two or more hostile factions are fighting I`d expect the whole solar system would be seeded with hundreds or thousands of cheap expendable sensor platforms to monitor all known enemy military installations, industrial sites, shipyards, spaceports thus making it extremely difficult to launch a major attack undetected.

Actually, people make a good point about a cluttered, inhabitated system: telling the various contacts apart from long range would be an absolute headache, emphasizing the importance of command and control ships/installations used to process all the data, or early detection of incoming hostiles.

I suppose it all depends on the scenario: the limitations and tactics of the situation would be different for two warring planets in the same system, vs. an extrasolar incoming fleet (which is unlikely if you want to be "hard", as in - no FTL )

I think the detectability of a coasting ship is probably a bit of a red herring: the real question here is how feasible it would be to hide the energy emissions of your engines. Because that's what's murder on the idea of stealth in space. Trying to make a coasting spacecraft inconspicuous is one thing, trying to make an active rocket inconspicuous is going to be a lot harder. And if you can't change your velocity without being observed your enemy will always know where you are by doing some math.

Darth Wong mentioned hiding your drive plume behind a cold umbrella. If it's tightly collimated enough I suppose that could work, but the big problem is it requires you to be sure that all your enemy's sensors are in the direction of the umbrella. If he has any eyes in your rear you're detectable. I imagine that any competent space navy would seed their solar system with lots and lots of observation platforms in lots of different orbits for precisely this reason.

So, I think the real question is how can you change your velocity without being seen to do so.

The only good possibility I can think of offhand is to launch stuff from a mass driver on an asteroid, and use the asteroid's body as a heat sink. Any other ideas?

Real spacecraft will spend most of their time coasting. You launch yourself in the direction you want to go, burn and adjust velocity until you're lined up with the pre-computed path, and then you coast until you're close. Most of the burn would have happened really far away from the target site.

Obviously, it doesn't work that way in most space opera, where the difference between atmospheric flight and space flight is rather blurred.

Yeah, but if you can't hide your burn the idea of sneaking around seems pretty problematic, as figuring out where the ships are the rest of the time and where they're going is just an exercise in math.

I mean, say you've just spotted a bunch of ships making rocket burns over at Pluto, and your computers tell you that their courses are likely to intercept your big naval space station or whatever. You'd have to be clinically retarded to get caught with your pants by a surprise attack from those same ships a month later.

If your space station is in the inner solar system, than pretty much anything leaving Pluto is as likely as not going to look like it's heading for you. How easy would it be to tell whether a fleet 100 AU away is heading for you or for another base controlled by a third power 2 AU away? Also, you're discounting the possibility of the attackers using IR-emitting drones to make it look like a bunch of ships are going in a bunch of different directions.

I can't do the math myself, but I'm pretty sure that orbital mechanics is highly deterministic, so as long as you had a good track on the ships you could plot out their courses fairly precisely. Somebody with more knowledge about the subject can probably address it better though.


But the enemy still sees a bunch of rockets headed toward a critical installation. He'd be an idiot to just brush that off. Maybe you could get him wondering about how to divide his forces since he doesn't know where the attack will come, but that gets into a second issue:

The problem with decoys in general is that you can deduce a ship's mass from how fast it's accelerating and how much energy is going into the drive plume. The decoys all have to be the same mass as your ships, and the same engines as your ships, in which case you probably almost might as well make them ships as they'll probably cost a good chunk of the cost of a ship. If you can send decoys off to "attack" twenty of the enemy's bases to confuse him, for that kind of money you could probably have built real ships and attacked 10 of his bases for real instead of just one.