Solar panels cannot operate at night.

If we put self-adjusting mirrors on the near side of the moon, they could reflect sunlight back to the dark Earth, allowing solar panels to operate at night.

How much of the moon would have to be covered in mirrors in order to allow Earth-based solar panels to operate at, say, 25% of daytime capacity?

(Let's assume that the reflected light is focused on the whole Earth, roughly evenly distributed.)

Also, which location(s) on the moon's near side would be more optimal/efficient (if any), and how much "light pollution" would this create?

  • 2
    $\begingroup$ Minor frame challenge - why not put the solar panels on the moon, and then have a microwave reception dish on Earth? $\endgroup$
    – jdunlop
    Nov 25, 2021 at 0:10
  • 1
    $\begingroup$ Any megascale construction project would serve well as a weapon. Beaming power from the moon to power relay satellites in orbit of Earth has been proposed by multiple individuals, and has the advantage of working during, for instance, a new moon, when the far side of the moon is illuminated and no light would be reflected to Earth. $\endgroup$
    – jdunlop
    Nov 25, 2021 at 0:24
  • 5
    $\begingroup$ "so that sounds like a dangerous death ray to me" Well, the same happens with the mirrors oriented in such a way that focuses the light on the target. Haven't you watched that 007 documentary? No microwaves there. $\endgroup$ Nov 25, 2021 at 0:40
  • 1
    $\begingroup$ If this is a plot to combat global warming, then that might be counter-productive. Almost all the warming on Earth originates from the sun. Projecting even more of the suns rays onto Earth would thus make the problem worse. And no, this does not apply to photovoltaic in general. Solar energy is about making better use of the suns energy which arrives on Earth anyway, but thermodynamically it's are a net-zero. $\endgroup$
    – Philipp
    Nov 25, 2021 at 11:28
  • 1
    $\begingroup$ @jdunlop and about 100 times greater inverse square losses compared to geostationary orbit, a horrible thermal and dust environment, many times higher costs for access from Earth, and the solar arrays and transmitters you went to such expense to construct on the moon are in the dark half the time, negating any advantage over solar panels located on Earth. $\endgroup$ Nov 25, 2021 at 14:29

6 Answers 6


You're doing it wrong

First, bear in mind that optics used to be a normal branch of physics where you had to learn all the impossible things you can't do. Now, well, this is by no means the strangest thing I've read recently: Anti-Solar Cells: A Photovoltaic Cell That Works At Night

Next, well, if you want to bounce light off the Moon to a point on Earth, you should just build a solar panel there, convert the light to microwaves or laser, and beam it straight to the point that you want to receive it. Otherwise you can't narrow down the effect of any one mirror to an area on Earth that looks smaller than the Sun from the same perspective on the Moon. And nobody on the Moon is paying to build mirrors that spread their benefits out on that many non-paying customers.

  • $\begingroup$ "you can't narrow down the effect of any one mirror to an area on Earth that looks smaller than the Sun from the same perspective on the Moon" Certainly not with flat mirrors. But what if you use parabolic mirrors to focus the rays of the sun to a smaller area? $\endgroup$
    – Philipp
    Nov 25, 2021 at 11:23
  • 4
    $\begingroup$ @Philipp it can't be done. It's called the "law of conservation of entendue". It turns out that if you found a way to break it you could use it to build a perpetual motion machine, so we can be pretty certain it's not possible. $\endgroup$
    – N. Virgo
    Nov 25, 2021 at 11:29
  • $\begingroup$ The anti solar cell might work if it is a thermocouple taking energy from the heat flowing in the sky from the ground warmed up during the day, but with such a small gradient the efficiency would be small. Notice that the article makes a claim on the amount of energy produced without specifying the amount of radiant area to produce that energy. $\endgroup$
    – FluidCode
    Nov 25, 2021 at 15:41

No. The moon is simply too small to be useful as a mirror.

As you're no doubt aware, moonlight is already reflected light from the sun. Lunar dust is fairly dark, with an albedo of about 11-12%. In other words, about 8/9ths of the light that falls on the moon doesn't reach Earth. With a perfect mirror, you could deliver all of that.

The problem is, the moon only delivers about 0.1 lux on a typical night. 1 lux, which you might get with a perfect moon-sized mirror, is roughly the light thrown by a candle on a surface a meter away. An overcast day might be lit to 1000 lux.

No matter how many mirrors you put on it, the moon won't light up your solar panels. Might I suggest batteries?

  • 6
    $\begingroup$ In fairness, the OP specified that the mirror would be focused on Earth, rather than naively coating the moon with mirrors, so there'd be more energy than just a perfect albedo. Still probably not enough, but more than 1 lux. $\endgroup$
    – jdunlop
    Nov 25, 2021 at 3:11
  • 9
    $\begingroup$ Your calculation works for white paint, but not a mirror. If all the mirrors on the Moon include a single point on Earth within the area that the Sun reflects to, then a person at that point will look up at the Moon and see the Sun perfectly framed inside, like an eclipse in reverse, providing day in the night. Minus a little ring maybe if it's an annular eclipse. The trade-off is most of the Earth wouldn't see the Moon any more, and you have to totally cover the visible portions in mirrors that adjust position every minute. $\endgroup$ Nov 25, 2021 at 11:56
  • 1
    $\begingroup$ When you say "In other words, about 8/9ths of the light that falls on the moon doesn't reach Earth" I think you mean that 8/9ths never leaves the moon. The light that does leave the moon goes in all directions so the proportion that reaches the Earth is much smaller than this! $\endgroup$ Nov 25, 2021 at 17:39
  • $\begingroup$ "No matter how many mirrors you put on it, the moon won't light up your solar panels." - not a correct statement - the moon can be a solar concentrator like these power plants - just arrange mirrors in a proper way. Will it make sense, no it won't, is it possible - yes it is. $\endgroup$
    – MolbOrg
    Nov 29, 2021 at 12:31

That would be awfully unpractical. In addition to this, bear in mind the moon rotates on its axis in 27 days and then some as well, and your panels would spend half of that time in the dark, so they're very roughly useful only 12 hours per day (earth night time), 13.5 days out of 27, so even before putting any additional physics or engineering into it, you have a 1/4th efficiency.

Off the top of my head I remember Gerard K. O'Neil doing in The High Frontier some (quite optimistic) maths to calculate at which point orbital energy generation breaks even (assuming locally sourced material from transient asteroids) economically. Here you have the advantage of h24/d365 energy generation beamed to Earth via microwave. Ground reception antennas would be no larger than photovoltaic farms and less disruptive for the environement. Zubrin talks about it as well in The Case for Space, if you're looking for more references. It's a very common discussion subject when it comes to industrialization of space and you will find plenty of references around. What if we put them on the moon, what if we put them at Lagrange points, what if we have to haul the material from Earth, what if we source it locally? etc.


Can mirrors on the moon allow solar panels to operate at night?

Not really.

You may have looked up in the night sky before and wondered "Where'd the Moon go?"

The lunar cycle is about a month. Since it's not quite a month, we'll call that a moon-month. There's a less fun, more astronomical term for it that I won't be using. Over a typical moon-month, the Moon will appear half-lit or less for about half of the time, and not always the same half mind you. And at the start of a moon-month, it will be just completely dark. Not great.

There is no point on the Moon that will be illuminated consistently every night. You'd have to cover the near-side of the Moon from end to end to get light almost all of the nights, and then you'd still get nothing on a new moon. Bottom line is that instead of idle solar panels you'll have idle Moon mirrors. Think about that before you whip out the checkbook.

Another reason the Moon might be absent is clouds. I'm not sure if a good enough mirror could allow the light to peer through a hefty cloud layer, but that's definitely something you should find out before you invest trillions in Moon mirror technology.

And also frame challenge.

The problem of idle solar panels need not be one. Energy usage isn't constant over a day, nor day to day. On our modern power grids, we always want to produce as much power as we consume. When we have excess energy, we store it. Solar panels that don't work at night may simply be a feature, not a bug. They work and store excess energy by day, and the stored energy is used by night.

I know the Sun is a pretty cool (but actually quite hot) celestial body, and solar panels are certainly useful, but they're also not a silver bullet. And that's fine, we have other ways to generate power, and you should too. We call that the energy mix.

Now if you're specifically looking for a pretty cool (but actually quite hot) celestial body, available day and night, come rain or shine, to complete your energy mix, you're in luck because there is one. It is called Earth. I'm partial to geothermal energy and I think it is criminally underexploited. I mean, it's just there for the taking, and you barely have to dig for it, and it's good for both electricity and heating.

Anyways, the point is, if your solar panels are down for the night, other power plants can pick up the slack. I'd admit, it's not quite as cool as Mirror Moon, but a healthy energy mix works. And if you're starting to design convoluted space contraptions to lit a few solar panels, you're designing your power grid wrong.

  • $\begingroup$ Thanks for the answer. I think it's a pretty good answer overall. I've looked into it, and I now think that the north and south poles of the moon are in perpetual daylight (or nearly so). $\endgroup$
    – cowlinator
    Nov 29, 2021 at 23:45

Sure. Well, almost.

There are lots of answers here saying why that would be a bad use of your resources, which is certainly true, but not answering the question. So I'll do that here.

The disk of the sun has an angular diameter of about 32' (arc-minutes) in the sky. If we cover the near side of the moon with adjustable flat mirrors, the reflected beam from each mirror would also have a beamwidth of 32'. If we aim all the mirrors at the same spot on earth, those mirrors from earth look just as bright as the sun (assuming 100% reflective mirrors). The disk of the moon is about the same angular width as that of the sun, so during full moon this produces the same amount of lighting. In that area it would be like day. The spot on earth that the 32' wide beam from the moon illuminates is some 3500 km wide. (Which happens to be the same as the diameter of the moon, as the moon and the sun look the same size in earth's sky.)

Now, the earth is bigger than that, so that 3500 km wide spot won't cover the entire earth. The diameter of the earth is some 12000 km, which is 3.67 times the diameter of our spot. The angular area of the earth's disk as seen from the moon will be 13.45 times as large as the area of our spot. To distribute that light evenly over the whole disk, it will be 13.45 times dimmer, so your solar panels will only work at 7.4% of daylight output. But if you are willing to make some compromises such as not illuminating oceans or Antarctica, you should be able to get your percentage up over 25% for a large part of the time, and at some times closer to 100%.

Of course this only works when the moon is in the sky at night, which isn't always the case. There is one happy accident which is that the illuminated part of the moon's disk is always the same as the dark part of earth's disk as seen from the moon, so if you want to illuminate the dark part of earth uniformly (at 7.4% of daylight illumination) the moon's phase nicely cancels out. If the moon is at, say, first quarter, then the earth as seen from the moon is at last quarter, so only half of the visible earth disk is in nighttime and needs illuminating, so we can focus the mirrors on the moon only at that area of earth. It doesn't work perfectly, as the spots we can work with are 3500 km wide, but it works to a first approximation. If you want at least 25% of daylight illumination, well, you need at least 25% of the moon's disk to be in sunlight.

This plan assumes you will cover the entire visible side of the moon in mirrors. That will give you an illumination of 7.4% of the sun if you want to cover the entire night side of Earth, up to 100% if you are happy with a 3500 km wide spot. If you are happy with a 3500 km wide spot and 25% of daytime illumination, you only need to cover 25% of the moon's surface with mirrors. The most economically effective (well, least ineffective) place would be the center of the moon's near side. Or if you want to make as much as possible from the crescent moon, cover a band over the moon's equator on the near side.

As for light pollution, that depends on your definition. The moon will look as bright as the sun, depending on how you point the beams. Is that light pollution or is that just extended daytime? It surely will wreak havoc on the day-night rhythm of plants and animals, but with a bit of environmental care many species could probably adjust.

But if your goal is energy production it is much better to put those solar panels themselves on the moon. A lot of the light reflected to earth won't end up on a solar panel after all. Or even better, put the panels in space where they don't suffer from nighttime. Or just put them on earth and invest in some kind (any kind) of energy storage, that will be much cheaper than putting stuff on the moon or in space.

  • $\begingroup$ @downvoter, care to explain what you think is wrong with this answer? $\endgroup$
    – JanKanis
    Nov 29, 2021 at 11:47
  • $\begingroup$ not a downvoter, and it is a nice attempt, at least you did try, but mirrors on the moon can do a better job at focusing, should they, not so much, can they, yes. $\endgroup$
    – MolbOrg
    Nov 29, 2021 at 12:35
  • $\begingroup$ @MolbOrg No they can't really focus much better, due to conservation of etendue. Using curved mirrors can project a 3500 km wide sharp image of the sun on earth while a flat mirror projects a slightly fuzzy 3500 km wide image, but that doesn't make much difference. Covering the moon with individually pointed flat mirrors approaches the results of a curved mirror well enough to not make a difference. It is effectively a segmented curved mirror. $\endgroup$
    – JanKanis
    Nov 30, 2021 at 11:00
  • $\begingroup$ Yes, a smart word for a simple fact that one can't focus ligth to a higer temperature than a source has(sorry for mixing stuf, but for simplicity), this case sun, cuz it would break the universe(joke). If you would remember it that way, you would notice the mistake in your assumption. Concentrating is possible and it is exactly how concentrated solar power, those towers, work. There will be other limits as well, but making a 350km spot, u even do not scratch those theoretical limits. If it is not enough, I may try to be more specific. $\endgroup$
    – MolbOrg
    Dec 2, 2021 at 2:08

Light - and the energy in it - doesn’t magically grow, so if you want to feed one square meter of solar cells on earth, you need (at least) one square meter of mirror to catch and reflect the equivalent amount of light. If you are fine with 25%, you need 25% of it.

If a mirror distributes light to a larger area, it will be less intense, so nothing is gained.

Looking at it from the other end, if you want 25% light, it would have to be 25% as bright as daylight. The moon simply cannot provide this by reflection- just go outside and compare the brightness of moon and sun.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .