If someone wanted to grow large quantities of food on Mars and they wanted to specify sunlight rather than nuclear or other sources, the two options I can think of are greenhouses and photovoltaic/LEDs.

Consider square kilometer types of areas:

Greenhouse windows would have to support the mechanical pressure of the whatever atmospheric pressure the plants required (let's say 1/3 Earth unless you have more authoritative numbers). Without Earth's warm air, the outside surface of greenhouse windows would be exposed to the infrared temperatures of space and the panes would have to do some thermal insulation also. These plants would suffer if it got too cold inside, and for optimum growth rate would the temperature need to be regulated carefully? Then there are concentrators to take the weaker Martian sunlight and focus it on individual plants for efficient use.

Compare to photovoltaics - could they be made robust against the low pressure and cold more easily than greenhouse windows?

Would stacking plants in 3D with LEDs be easier than spreading them out over kilometers of greenhouse floors? Would humidity and temperature control be easier and more efficient and total kilograms of water needed be lower in one configuration over the other?

For an example of what 3D stacking means, see The Daily Mail, and Modern Farmer and National Geographic.

There are several related questions here, but I believe that if an authoritative source comparing greenhouses to PV/LEDs exists, it will likely have considered all of these.

enter image description here


  • $\begingroup$ Zubrin wrote quite a bit on this, and he makes a good argument that growing plants with light generated from photovoltaics is untenable, since (iirc) Rhode Island's solar incidence is greater than the world's energy supply. Incidence on Mars is not that bad, 3d stacking is possible with sunlight (more widely spaced), and the high CO2 easily gotten on Mars can counteract the decreased crop yield. Greenhouse effect still works, and domes can be weak due to lower required pressure. $\endgroup$
    – user19742
    Commented Jun 19, 2017 at 3:15
  • 1
    $\begingroup$ The very small efficiency of a photovoltaic/LED combination should be considered. 10 % or even less for light energy in to light energy out. $\endgroup$
    – Uwe
    Commented Jun 19, 2017 at 11:17
  • 2
    $\begingroup$ @Uwe we're comparing two things. Photosynthesis is about 1% right? A professor of mine a zillion years ago taught us never to use the word "inefficient" unless in a direct, quantitative comparison to something that could be demonstrated to be more efficient. Otherwise it is not a helpful concept. $\endgroup$
    – uhoh
    Commented Jun 19, 2017 at 13:02
  • 4
    $\begingroup$ @ uhoh I did compare only the efficiency of a window against the photovoltaic/LED combination. The low efficiency of photosynthesis is valid for both cases. $\endgroup$
    – Uwe
    Commented Jun 19, 2017 at 18:22
  • 3
    $\begingroup$ @Uwe LEDs allow you to shift and concentrate wavelengths. Photosynthesis does not use the wavelengths where sunlight is the strongest. i.sstatic.net/oi8Yk.png It pretty much only uses red and blue. PV/LED allows you to collect the intensity in the broad range in the middle of the visible spectrum and shift it into narrow peaks where chlorophyl can actually use it. See this question. It also allows you to shift in time to optimize use patterns, and shift in space by concentrating light dynamically on the plants vs the space between. $\endgroup$
    – uhoh
    Commented Jun 20, 2017 at 2:33

2 Answers 2


Solar irradiance on Mars is about half that on Earth.

So a greenhouse would have to be twice as large as on Earth to get an Earth-equivalent amount of light on the plants, using solar concentrators. If you build the concentrators outside the greenhouse, the greenhouse can be smaller but you'd need a high structure for the concentrator (as high as the greenhouse is wide, at first approximation).

Let's assume all the building materials have to be transported from Earth, so weight serves as a proxy for cost.

On Earth, greenhouses are easy to build: just a frame and panes of glass. On Mars, the structure has to be pressurized, making the structure heavier.

A pane of glass that can contain 0.33 bar is going to be thick: a load of 4 kN/m2 already needs 2x 25 mm of glass, and 1/3 bar is 33 kN/m2. Plastic may be thinner, but not 100 times thinner.

PV panels, on the other hand, don't need atmospheric pressure to function. So you can build light frames with panels on top. Even roll-up arrays might work. I can easily see a 10:1 weight advantage for PV panels.

The cheapest combination would then be a small pressurized building full of plants using 3D stacking, powered by a field of PV outside. You could even stick the plants inside a tunnel and have them protected from radiation.

  • $\begingroup$ Even if you double the greenhouse area, you're still getting the same amount of sunlight per cm^2 of plant's leaf. And while you could grow some berries or hazelnuts, forget corn, potatoes or wheat, these simply need more light density. OTOH a hybrid method with combination of LED and sunlight should work fine. $\endgroup$
    – SF.
    Commented Jun 19, 2017 at 9:27
  • $\begingroup$ @SF. It's a good point, but this question is pretty much living in "order of magnitudeville". We can assume that there will be genetically modified crops that grow in alternate levels of sunlight such that the growth rate would not suffer more than linearly with reduced sunlight, and probably some enhancements are on the way that will help further. And there's always the possibility of reflectors such that two square meters of area feeds one square meter on the greenhouse floor. You still loose the factor of 2 but the plant receives terrestrial levels of sunlight. $\endgroup$
    – uhoh
    Commented Jun 19, 2017 at 9:59
  • $\begingroup$ I was hoping for an answer that had some links to places where we can see this worked out, or it worked out here. I'm not sure "expensive" is the proper metric unless you really mean the mass-related expense of flying it there from Earth. What are the relative "shipping" weights of a square meter of PV panel that might last 15 years on Mars, and a square meter of greenhouse window that would work on Mars and last 15 years? Then we can add a factor reflecting the ratio of food yield per square meter of PV versus window. I have a strong hunch this is the correct conclusion, but more is needed $\endgroup$
    – uhoh
    Commented Jun 19, 2017 at 10:05
  • $\begingroup$ @uhoh: ...or square meter of "mirror" foil plus lightweight scaffolding of PVC pipes. Good point with the reflectors. The greenhouse would also almost certainly be plastic, not glass, if shipped. If it was to be glass structures, likely parts for building a glass furnace would be shipped rather than glass panels. $\endgroup$
    – SF.
    Commented Jun 19, 2017 at 11:27
  • 1
    $\begingroup$ @uhoh: Common polyethylene, as used in Polytunnels. I don't know about any "space uses", but on Earth they last 4-5 years. Personally, I'd imagine the greenhouses to be much smaller - probably many domes several meters in diameter; that way both losses in case of a breach would be minimized, and maintenance breaks (depressurization) would be better "distributable", plus holding ~0.3 bar inside wouldn't cause enormous stress. $\endgroup$
    – SF.
    Commented Jun 19, 2017 at 13:24

This is a pretty difficult question to solve:

The problem with using a greenhouse on Mars would be the high levels of harmful solar radiation on the surface due to Mars' thin atmosphere and lack of magnetic field. Using solars panels on the surface of Mars would also be difficult on a large scale due to dust accumulation.

And like you mentioned, there are many more factors going into this like thermal insulation and the low-pressure environment.

A possible solution would be to use solar energy satelites to beam power down to an underground 3D stacked farm using a microwave transmitter and reciever. While such a setup is more viable on Mars due to the low atmospheric pressure, the transmitter would still end up being large and difficult to set up.

  • $\begingroup$ One of the people (or robots) who are tending the plants and eating the food (ok not a robot) can just go outside and sweep the windows. Glass that's already thick enough to hold the pressure difference and thermally insulate can also be engineered to absorb any harmful UV or X-rays, if it doesn't already. Putting PVs in orbit doesn't address the question as asked. $\endgroup$
    – uhoh
    Commented Jun 19, 2017 at 10:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.