So, if the mirrors are required permanently, would they not make a viable strategic target, and another weakpoint required to safe guard?
This could add another layer of consideration, seeing how a mirror station would be vulnerable to enemy sabotage that would affect the colony it was assigned to, away from its garrison fleet.
Yes, that would definitely be a consideration. Something else to protect.
It doesn’t mean you have to guard all the mirrors forever - sometimes you’d need a mirror just to evaporate some ice to let the terraformers suck it out, or you’d place a temporary mirror until the terraformers finish with adding greenhouse gase.
Another use is to make enemy strongpoint planet a bit unpleasant for them without being too rude and glass it with nukes - for a decade or so, making a slow xenocide, or just before the invasion, to switch the landscape bonuses.
I don’t know if mirrors could be feasible…
E.g., we want to double the energy that arrives to a planet. This means that an “equivalent star” (i.e., another energy source) should be present in the planet sky, emitting the same quantity of energy emitted by the star. So, if we had a perfect mirror and place it just next to the star, its diameter should be the same of the star! If we need even more energy, the mirror should be increased.
It’s not a question of focussing the energy to the planet, but of the energy captured by the mirror. Focussing avoids to lose the energy intercepted by the mirror.
And the more the planet is distant from the star, the larger the mirror should be. E.g., to obtain conditions good for human beings, we should have an “equivalent star” as large as the Sun in the Earth sky: for a body around the Sun at a distance double than the Earth one, we should build a mirror as large as 2.8 million km, giving a reflecting area four times the one of the Sun!
Moreover, if the mirror is not placed next to the star, but at a point distant from this (maybe a Lagrange Point, to compensate the variable star-planet distance of an elliptical orbit), its diameter scales up enormously to reflect enough energy towards the planet, because the energy reduces with the square of the distance from the source (i.e., the star).
And finally, there should be one mirror for each body we want to heat…
Certainly too many minerals needed, as we are not playing as a Type 2 civilization ( Kardashev scale - Wikipedia ). 
A mirror in the orbit of the star doesn’t have to be big, since it would capture significantly more energy per mm/sq than if it would orbit the planet. The bigger factor would be the foci designed to transmit all of that captured energy into an extremely small distant point, aka the planet.
Years ago I read a book where humans built a whole network of mirrors in the Solar system to achieve this. There were primary mirrors orbiting the Sun and a bunch of secondary and tertiary mirrors that relayed the solar power where it was needed. And a plot point was that when hostile aliens showed up to conquer Earth, the mirror network was converted into a weapon, scorching alien battleships into fiery hulks.
Think about it the other way around. If you want to double the energy hitting the Earth, you build a mirror the size of the Earth and in the same orbit just slightly ahead or behind. The same amount of energy hits it as hits the Earth (because it’s the same size) and is reflected onto the Earth.
This shows that you absolutely do not need a mirror the size of the star.
Further, the energy that reaches a planet from a star is a function of distance from the star squared, because you’re calculating the area of a sphere with a radius of that distance, and all the energy coming out of the star is spread across that sphere’s area. However, that’s because the star is a point source that radiates in all directions. If you captured the energy at a certain radius and focused it such that it was collimated from there to the target planet/body, it wouldn’t fall off with distance squared anymore. Certainly you’d have some losses, but much, much less than distance squared.
From the surface of the planet it might look like a “2nd sun”. But that would depend significantly on how far away it was. If it was close, it could be much, much larger than the sun but less bright. If it were far away, it might be much smaller but even brighter (which might not be ideal for the Mk 1 eyeball.
Firstly, I am not definitely implementing this, but the idea is interesting.
Secondly, a mirror does not need to be anything like the size of the planet to double luminosity.
Earth is 12,742 km inn diameter and 150m km from the Sun.
So the physics problem becomes how close to the Sun can you position the mirror, and how efficiently can you beam that energy to your target destination. Also, we are thinking in terms of a physical mirror but it could be some form of field that gathers and re-directs energy, or some TN equivalent.
The opposite would be something blocking sunlight positioned between the star and the planet, but close to the star.
This would make a potential weapon if you could target it. You could argue that it needs a fire control and that restricts the range against ships, while at longer ranges it just provides heat to the planet. You could still attack your enemies though by raising the planetary temperature.
Plus if you do use this to keep a planet warm and it gets destroyed, you have a significant problem.
Wouldn’t shades need to just cover the surface of the planet they’re trying to cool? Why would they need to hang near the sun?
Esp. when you instead consider them something like a curtain instead, instead of blocking all of the sunlight, it just absorbs some of it before it reaches the target body, which reduces the incoming energy.
It feels like placing a bunch of shades around the star could result in potential secondary effects for other bodies in the system.
Because the closer you place the ‘shade’ to the sun, the smaller it has to be to block the same amount of sunlight.
Another thing mirrors could solve/improve would be planets with fairly eccentric orbits. So planets that otherwise would be great potential colonies, but due to eccentric orbits can’t get good temperature range.
Mirror could heat them up when they are far away from their star and then automatically redirect the beam (stop heating) when such colony gets closer to the sun. Hypothetically that could be automated to provide stable temperature on eccentric planets.
I think that would be a great mechanic and a fun puzzle for the game. Maybe we could even add some technology levels that would affect how much a planet temperature can change and how fast it can change?
I’ll just remind the problem that started it. Removing water ice.
If we can heat the planet with mirrors, will we need to heat the entire planet to > 0°C, or will there be a separate order “Heat ice sheet” that turns ice into water vapor on bodies with temperatures of, for example, -100°C? If we have few mirrors or they are small, it can do this quite slowly, but it will already be an affordable way to remove ice from very cold bodies.
Actually no, providing shade is the other way around: the closer to the planet the better. Because there is no focus effect anymore - you’re just shadowing some part of the stellar visual disc, and the star in bigger then the planet, so better to be closer to the latter. The umbrella needs to be really huge, yet much simpler (no focusing, nearly no requirements for precision).
If I add this, the ‘mirrors’ would have an immediate effect on temperature but removing them would remove the effect.
This is because as the mechanics currently stand, any change in temperature from orbital position, dust, albedo change, atmospheric gas changes, etc. is immediate.
The mirrors would be changing the amount of energy from the star reaching the planet, so to match all the existing mechanics that affect that, they would have to also have an immediate effect. That would distinguish them from terraformers. The effect of mirrors is immediate but temporary (ending if the mirror moves), while terraforming is gradual but permanent.
In reality (not Aurora), some planets would experience the effects more slowly. Mars, for example, reacts fairly quickly - within days - to a change in energy received. Earth, with large oceans and an atmosphere that act as heat sinks, reacts more slowly, but it still happens over weeks not years. This is why summer happens after perihelion.
I could simulate the speed of change for planets with atmosphere and oceans, but it is a lot of extra processing for no significant gameplay impact.
An alternative might be to lessen the impact of eccentricity on temperature for planets with oceans and thick atmosphere, so the temperatures don’t extend as far from the mean. This would make terraforming that adds atmosphere or water a little more effective than now.
Regardless of what does or doesn’t happen with mirrors/shades I like the idea of oceans and thick atm. mitigating the effects of eccentricity.
It would create reasons to go beyond the legal minimum of 20% hydrosphere (speaking in the normal oxgen-breather context)
I would also like to comment on the “shades”. We already have Dust, which does just that.
How crazy would be the idea of adding installations that generate Dust up to a certain level? This would allow (in a few years) to cool a planet that is too close to the sun to be cooled using the atmosphere.
Such cooling, unlike terraforming, will need to be constantly maintained (spending part of the population on Dust generation).
Theoretically, this could be done using orbital bombardment, but it would be quite expensive. 
Perhaps in this case it would be easier to add the ability for terraformers to add dust to the atmosphere.
Yes, I have been considering it since you mentioned it before. The tricky part was finding a relatively simple way to express the effect. I’ve created the concept of a new ‘dampening’ parameter applied to temp calculation and I have some preliminary code.
It is based on pressure, hydro coverage and orbital period. The first two set the level of effect and the third lessens that effect for longer orbits.
I calculate what the temperature would be without the dampening effect, then calculate the temperature at the average distance, then use the dampening factor to move the temperature closer to the average.
For example, if the dampening factor was 40%, the temperature at any given distance would be moved 40% closer to the temperature at the average orbital distance than before. A planet with an average temp of 300K and a current temp of 250K, would be 270K after dampening.
Currently, I have atmosphere having up to 50% dampening, using pressure * 10, with ocean having up to 30% - using Hydro Extent * 0.3.
So 5 atm and 100% ocean would reduce extremes by 80%. Earth (1 atm and 70% ocean) reduces extremes by 31% (10% for pressure and 21% for ocean).
Orbits longer than one year apply a reduction to dampening factor based on 1 / Orbital Period, so a five year orbit would reduce the dampening factor by 80%
Moons use the orbital mechanics of their parent body, but their own atmosphere and hydrosphere.
There are some complexities around the max not being the max if gases were frozen, or the ice didn’t melt, etc., but it can be handwaved I think. The other factor is whether players can easily understand the concept, as terraforming can be complex already, but I think providing the dampening factor should be sufficient.
I just need to play around with it a little more before implementing anything.
Love this idea.
Rise of Troy series by John Ringo
Yeah, that was the series. Shame he didn’t write more, some of the concepts were pretty cool.