In article ,
Pascal Bourguignon wrote:
Thank you for your comprehensive answer.

You're welcome. I had fun thinking about it.

Indeed, if there's nuclear energy available, it would be simplier to
use artificial light. But storing solar energy is hard and not too
efficient (you need a lot of heavy batteries, or complex
electro-chemical processing), so if natural light can be used at least
half of the time, the better.

If you want to think about this like an engineer, you have to consider
all factors. You may find that building vacuum and meteorite proof
transparent surface structures that require supplemental lighting for
half the month is much more difficult, expensive and unreliable than
having sturdier structures that are artificially lit the whole month
through. After all, if you have to have the spare generating capacity
to provide lighting for half the month (at least - the sun isn't always
high enough in the sky to provide adequate light) you've got it for the
other half as well.

In my example of hydroponic vegetable production in deep mines in
northern Ontario, you have an example of a situation where it's much
cheaper to light a naturally heated (cost-free) area than to heat a
naturally illuminated area on the surface. What's more, it costs to
build a greenhouse, but worked out areas of the mine are already
there. Transportation to the area is time-consuming and expensive
enough that fresh vegetables arrive in poor condition especially in
winter, so it's competitive to raise them locally with free heat even
though electricity for lighting is not very cheap. It's much easier
and cheaper to control environmental factors in the mine than on the
surface, e.g. temperature, day length, light intensity, etc., because
the background environment is constant. (IIRC, the crops are grown at
a depth corresponding to 30C - about 1km underground - while surface
temperatures often go below -40C.)

Conditions on the lunar surface vary drastically from intense heat to
bitter cold over the lunar day/night cycle of 29+ days. You can't
readily dissipate heat by better ventilation as you can in a
terrestrial greenhouse, and shading kind of defeats the purpose. You
don't get conductive or convective cooling of the structure either, and
you'll either have to have movable insulation or expend a lot of energy
to keep your plants from freezing in the lunar night, when your
structure is radiating to unobstructed sky at 3K. At any rate, you'll
need a lot more technology to keep your plants alive in such a
structure than you would in one that has a more constant background.
You'll need heating, cooling, variable amounts of light, protection
from small meteorite strikes, as well as vacuum resistance, not so easy
in a transparent material. When spare parts are fantastically
expensive and may take weeks to obtain even in an emergency, you want
the simplest technology feasible, because it's likely to be the most
robust and easiest to repair. When your air supply depends on your
greenhouse, you don't want to worry about it catastrophically failing
when it gets hit by a pebble, or a bit of caulking cracks.

Overall, you're much further ahead in an underground structure or one
protected from drastic thermal variation and mechanical damage from
small meteorites by a thick layer of regosol, even if you have to light
it. You'll probably be living in exactly the same kind of structure
already, so you'll know how to build and maintain one.

You can generate some of the energy needed for lighting with solar
panels on the surface, but they are only useful for part of the month,
and you'll need another source of energy for the rest of the month.
Your colony is going to need energy for other purposes than growing
plants, so you've got to deal with the problem of intermittent
availability of solar energy anyway.

At any rate, I had fun speculating about all this. Are you planning to
write some science fiction against this background, Pascal? If so,
remember that good writing can carry the reader blindly past a lot of
leaps of faith in science and technology. If you want to correspond
with me by email, take the no-uce and yyz out of my address.