Newtonian Aurora

[Yonder]

But when you are both staionary you would not need to take velocity data from the sensors so that error would not be consistent.

How do you know that you are stationary to your target if your velocity readout isn't telling you zero? If you have some sort of secondary, more accurate velocity sensor that tells you that the target is stationary and you can ignore your velocity readout, then why aren't you always using that secondary sensor?

In any case my previous list of points was not a list of things for Steve to try and model it was a simple justification as to why, when there is movement involved, your shots will be less accurate than when there is no movement involved; even when that is at a constant speed.

Phrased in another way. How do you match speed with your target (to get the stationary scenario that your weapons theoretically operate best at) without using your velocity readout?

Yeah, but if you are not going to model the effects, and they are very small effects, or effects that can be easily compensated for as part of normal operations, then the best thing to do is say that those effects aren't there.

[chrislocke2000]

Yonder

I understand where you are coming from with the constant errors in sensors but the whole point of the discussion was comparing accuracy of stationary to non stationary hence ignoring it when you were in that theoretical situation.

All the differences are tiny, the whole point is that at the ranges and accuracy requirements we are talking about you are going to have to be infinitesimally accurate in order to score a hit. That means, rather than try and model all of those tiny variances, you introduce a simple error in you hit chance to simulate all of these tiny problems.

[Yonder]

That's just it, you don't have to be infinitesimally accurate. There is a finite, though precise level of accuracy that you have to accomplish, and that level of accuracy has already been surpassed in the modern day. From laser rangefinding the moon and further bodies, to putting millions of transistors on tiny chips, to mapping Earth's irregular gravitational field we already do things that require really precise measurements. If you do the math you find that today's fidelity instruments are enough to do these operations already.

Lets say you are trying to genocide a small body, like the size of a moon, and your sensors were off by 1 cm/s per 10m km. (An earlier example, and according to Webb and Jones this is already in the low-performance range of commercially available inteferometers in a vacuum: http://books.google.com/books?id=dVOz6v5icxkC&pg=PA1741#v=onepage&q&f=false) We will also assume that we begin by approaching at 15,000 km/s with guns shooting at 100 km/s, with a gunlaying inaccuracy of .2 arcseconds (What byron claimed was modern laser accuracy).

If we shoot from 1e10 km away (around 60 AU, 50% again past Pluto's average distance from the sun) then it will take 7.665 days for our projectile to hit. That means that from our perspective we are shooting at something 66.2m km away. Our gunlaying inaccuracy gives us an error of 66.2 kilometers, and our sensor inaccuracy adds another 66.2 km of inaccuracy (a coincidence that arose from our round numbers, we are firing at 100km/s and our sensor inaccuracy is 100 times more than our gunlaying inaccuracy). That means that our total maximum inaccuracy, if each of our systems had the maximum amount of error, and they were each wrong in the exact same direction, would be 132.4 km. The moon has a radius of 1738 meters, it's easily genocided.

In fact, we could even do this with a gun with the accuracy of a modern high-accuracy hunting rifle, which top out at "sub minute of angle" or 3milliRadians. At that point we add the 66.2 km of sensor accuracy to 19264.1 km of gunlaying error. Our maximum error range is now 19330 km and considerably higher than the moons radius, or even higher the Earth's 6378 km radius, but our Astronaut with a deer rifle sitting on the space shuttle refitted with TN engines is still hitting the planet a non-trivial amount of the time with no other technological modifications.

And that's just a handheld rifle, it took me awhile of searching before I could find a single reference to the angular error of any larger guns, but I finally found a reference on Jane's ( http://articles.janes.com/articles/Janes-Ammunition-Handbook/76-mm-APFSDS-T-ammunition-for-Rooikat-armoured-car-South-Africa.html ) showing that a South African tank which entered production in 1983 has an accuracy of .3 mils, or milliRadians. This is an order of magnitude better than our rifle and at 1992.6 total error we are now always hitting the Earth, and reliably hitting the Moon. All with today's technology.

You're not going to be able to avoid Genocide with realistic accuracy penalties, or any sort of reality/physics based approach. We would either need to drastically lower the delta-Vs and some other capabilities of the ships (a ship which had difficulty making a 3,000 km/s attack run would be much less dangerous than the current Daring, which can can easily make a 15,000 km/s attack run) but I don't think that you would be able to do so without increasing the travel times and assorted logistical complexities so much that you were negatively impacting the game. We're either going to have to do some sort of handwaving defense mechanism, or just have the AIs not try to genocide the player. (A player could genocide in a pinch, but if the dust mechanics weren't changed any sort of attack on this scale would probably make any planet hurt in this way inescapably cold for thousands and thousands of years.)

I would expect a Trans-Newtonian (or whatever we'd be calling it) culture to start out with instruments an order of magnitude or so more accurate than ours, possibly even more. If the tech progression followed similar scaled improvements to the rest of the tech tree then within 30-40 years of game time races might even have the the accuracy to destroy shipyards and orbital habitats in the above scenario (which remember, was 50% farther out than Pluto).

[bean]

I agree that we can reliably hit a planet at long range.  My question is if it is practical.  You're not looking at genocide with the accuracy in question.  You're looking at ecocide.  And at something on the order of .05c, that requires probably megatons.  There is no way to launch anything smaller then an asteroid unguided at those sort of ranges.  I'm skeptical of unguided weapons at any range, but long range is by far the worst.  It's quite easy to avoid targeting.  I doubt you can get really good accuracy with passive sensors, and if you get lased, put on about 5 cm/s.

[UnLimiTeD]

Well, so planets probably are save from anything but dedicated weapons, against which someone will have to find a solution.
Homing weapons can be countered by simple countermeasures, and guided weapons will give away the gunners position.
Both of those are also expensive.
Shipyards are another ballgame.
How can we prevent those from being wrecked several Au out?
I suppose it would be logical to only have a skeleton drydock up there, which is easy to just shoot through without hitting, and assemble the components on the surface; that at least weakens the problem.

[chrislocke2000]

If we shoot from 1e10 km away (around 60 AU, 50% again past Pluto's average distance from the sun) then it will take 7.665 days for our projectile to hit. That means that from our perspective we are shooting at something 66.2m km away. Our gunlaying inaccuracy gives us an error of 66.2 kilometers, and our sensor inaccuracy adds another 66.2 km of inaccuracy (a coincidence that arose from our round numbers, we are firing at 100km/s and our sensor inaccuracy is 100 times more than our gunlaying inaccuracy). That means that our total maximum inaccuracy, if each of our systems had the maximum amount of error, and they were each wrong in the exact same direction, would be 132.4 km. The moon has a radius of 1738 meters, it's easily genocided.

I think I understand your logic here but doesn't it assume that your ship itself is flying a perfect course to the planet at the point of firing your weapon and as a result the error component is only in the aiming? I would have thought that you might expect a similar error in the ships navigation as well which would then introduce the largest error in targetting?

[jseah]

I agree that we can reliably hit a planet at long range.  My question is if it is practical.  You're not looking at genocide with the accuracy in question.  You're looking at ecocide.  And at something on the order of .05c, that requires probably megatons.  There is no way to launch anything smaller then an asteroid unguided at those sort of ranges.  I'm skeptical of unguided weapons at any range, but long range is by far the worst.  It's quite easy to avoid targeting.  I doubt you can get really good accuracy with passive sensors, and if you get lased, put on about 5 cm/s.

It could just be a couple of nuclear missiles minus engines and targeting.  IE. unguided nuclear shells

[samtmj]

Dear All,

New to Aurora forums.  Been playing for a year though.  Some thoughts on the matter of extreme range planetary bombardment.  I don't mean to offend, but though it is an attractive option, isn't it rather impractical?

Now imagine WE'RE all hostile aliens just entering the solar system and about to target earth with extreme range ballistic bombardment.  Remember we would have VERY little pre-existing astrogravitational data on the system.  What? Take time off for detailed survey in a hostile system with active defenders??? Given that our firing location is just beyond the orbit of PLUTO;-

Our problems are, assuming we can even manage to LOCATE EARTH from so far away (what a miracle that was!):-

1) All newtonian sensor data is an image of the past thanks to the speed of light limit.  It takes approx 4 to 7 hours for light to travel from SOL to PLUTO; but EARTH moves at approx 108,000 km/hour around the sun.  Thats like a 430K km to 750K km difference by the time we are pointed at it.  Sure, lets LEAD THE TARGET, but that only takes us to the next problem:-

2) To "lead" the target is difficult given the extreme time lag, because we would need to know the object's EXACT astrogravitational movements, a study that will take years (365. 25 days???) of observation.  But remember, the earth MAY NOT take the EXACT SAME path around the sun for each year due to the following problem:-

3) Predicting the exact planetary orbit given its many variations due to gravitational effects from other solar system bodies and "wobbling" due to LUNA.  Then you need to study the intricate sequence of planetary alignment, but planetary alignments are kinda unique each year so you would need to study and predict ALL the significant solar system bodies to predict their effects on earth orbital variations.

4) Okay now thats done, we ALSO need to factor in the gravitational effects of all those solar system bodies ON THE PROJECTILE, including all other minor bodies that are possibly in the way (planetoids, asteroid belts, comets).  Assuming we've done that, next:

5) What about the effects of solar wind and its unpredictable variations blowing on the projectile? We'd have to study SOL's internal structure, internal convections, solar flares, solar spot activity.  Not to mention that solar wind interaction with the magnetosphere of each major body may have a small effect.  Also, we must assume that:

6) The further away the target is, the more unreliable the sensory data (degree of uncertainty increases with range).  Remeber there are such things, (no matter how minute) as gravitational lensing of light rays, interference from nearby gravity centres, diffraction, echoes, interference from solar wind (and especially actively from defenders), etc.  So we roughly know where a planet WAS to the nearest +/- XXX km, but that may not be precise enough. 

7) Finally we'd have to hope there are no collision whatsoever (not even a glancing blow) with micrometeorites / space dust / comets / asteroids (unlikely) that will put our projectile off course.  Natural stuff are easily avoided (except the space dust and small stuff), but what about:

The last problem is a valid counter-defense in response to any such projectiles that we might launch.  Moving at relativistic speeds, even a collision with a human-seeded "cloud" of floating micro-debris in space early on is enough to veer the projectile sufficiently off-course.  Heck those humans could have already pre-emptively put up millions of square kms of those defensive stuff along the "predicted" optimal line of fire the moment they detected our ships warping in-system.  Remember we're fighting in THEIR territory now, an unfamiliar  battleground they have extensively studied and know very well more than we do.

9) And would you think they'd just let us conduct an accurate system survey in peace? Heck, bouncing all those EM signals in every spectrum off every major solar system body , blasting and lighting us up with all sorts of EM/sensor energy.  Putting up false "ghost images".  Partial cloaking of earth itself? I'm sure that'd have a profound effect on the accuracy of any survey.

Better we crack out those MK III 100 MSP self-guided drone missiles with the 999 radiation yield dirty warheads??? At least their courses are self-correcting with minimal sensor time-to-target (plus other interferences to be avoided) lag.

Regards,
Sam

[Steve Walmsley (OP)]

I haven't read the recent posts in detail yet but based on a quick skim...

Firstly, bear in mind the issue we are looking at is not just game-related but one that humanity will have to face for real at some point in the future, even if the firing ship is one human nation firing at another.

In Aurora terms I think it would be reasonable to assume that a ship with advanced technology way out beyond Pluto would still have a reasonable chance of hitting something as large as a planet. I haven't yet decided what parameters would affect such a long-range shot but I will read the previous posts in detail before I do. For the sake of argument, lets assume its possible. However, the debate should not really be about whether it is possible but if it is practical.

For starters, if you want a really high velocity shot, then your ship needs to be moving fast before it fires. You can either enter the system at high speed, or accelerate after you arrive. If you arrive slowly, it would takes years for the interstellar journey, so lets assume you arrive at a reasonable speed. If you haven't surveyed the system, you are very likely going to arrive on a course that doesn't allow the shot. You could even arrive on the far side of the system facing the wrong way. In that case, lining up the shot is going to take a while. But, lets assume you eventually manage to line it up without being attacked by the defenders.

Next there is the question of do you really want to destroy the population anyway? In standard Aurora, it is fairly easy to wipe out a planet with nukes. It doesn't happen too often though because you would like to either steal the industry or use the planet afterwards. While a massive kinetic strike wouldn't result in radiation, it would result in a LOT of dust and it would destroy any industry.

Finally, there is the chance of the defenders intercepting the projectile. Anything moving at extremely high speed could be destroyed by throwing a pebble in its path. It wouldn't be too hard to intercept the projectile with some type of debris field from a shrapnel warhead, or you could simply detonate a large nuclear warhead nearby. Aurora has FTL sensors so you may be able to detect it in time to intercept it. A small object such as a 1-2kg railgun projectile would be virtually impossible to detect with normal Aurora active sensors. However, as I mentioned earlier in the thread, I am considering adding some type of new sensor that detects an object based on mass x speed. That would probably allow detection of even small objects with a very high velocity.

So in summary. Lining up the shot won't be easy. It would be unusual that you would want to destroy the ecosphere of a planet and the defenders would probably be able to intercept the shot anyway.

Steve

[Panopticon]

The question on my mind though, is could they intercept a thousand shots? I assume a railgun projectile is cheaper than a nuke for example, so for planetary attacks long range bombardment still works, simply to remove defenders ammo if nothing else.

I find for me, when I invade a planet, the desire to keep the infrastructure intact drops rather quickly as the shipyards keep pooping out little ships that annoy me, or as the ground war just keeps grinding along. There are a number of RP reasons to annihilate a population as well. not to mention the special NPRs seem to love doing it when they get a chance,

[Mel Vixen]

So in essence you would station some shrapnelclouds in orbit around your planet and it should stay save enough?

[UnLimiTeD]

We already have that kind of shrapnel cloud^^
No, a permanent debris field would be hard to maintain, expensive, and spread to thin.
However, a Shrapnel-field fired in the direction of an attacker would be able to disintegrate a large amount of projectiles.
Which is a good point, finally getting back to the actual gameplay implications.
The Ammo is an interesting point, I think kinetic weapons should cost ammo with the power they have now, and massively so.
Though I suppose a 10 Ton Magazine could already store 8000 1 Kg projectiles.
For planetary bombardment, it might make a difference.

[Elouda]

I think one issue that gets consistently overlooked in this thread (though I admit to not reading it all) is the effect of atmosphere on kinetics. Any 1kg projectile, or even 100kg projectile travelling at those speeds will be vaporised even by the relatively thin atmosphere in high orbit, or at least break up enough that damage becomes minor. Larger, slower projectiles are more resilient (and presumably might even be worth shielding, though if thats sensible gameplay wise is for Steve to decide), but also easier to intercept.

Here's an interesting little calculator and an even more interesting peice of documentation on how its calculated for those of you interested;
http://impact.ese.ic.ac.uk/ImpactEffects/

[UnLimiTeD]

Wow, thats a great page.
Just let it calculate the effects of a 1 feet projectile the density of iron, at 5° angle and 10000km/s, and the airburst (the projectile would vaporize in the stratosphere) has a strength of 2.08 x 1013 joules. Decent, but that's a large projectile, probably the size of a guided slug, and the energy would spread over a few kilometers. Looks like it might be more damaging to the environment than Industry. The site doesn't calculate if a projectile of this high speed, but this small size, would actually make an impact at ground level at all, it's probably meant for asteroids and not space warfare.^^

[Elouda]

Yes, but the energy and damage calculations are still valid and interesting.