Non-terrestial plantlife: modifying textures

[ACrisp]

Hi everyone,

I've been playing with Celestia for almost a year, and have recently started working on some new fictional yet plausible worlds of my own. My current project is a habitable planet orbiting a red dwarf sun, but I've run into a snag.

While designing the planet texture, I got to thinking about the plantlife that might exist on such a world. Most "Earthlike" planets out there assume that native life will use chlorophyll for photosynthesis and hence be green. I've decided for some local flavour that the plant life on Struve 2398 A Prime (the working name, and yes, I need a better name ) should have a different colour.

And here is my problem. I've been using the Gimp and Arcsoft Photostudio (very basic program that came with my scanner) for my texture creation but for the life of me I haven't been able to find a way to change the vegetation colours on an Earth texture while leaving all other colours (oceans, deserts, ice, etc) intact. I suspect I'll need Gimp for this, so if you have any pointers, I'd love to hear them.

Thanks in advance,

Andrew

[selden]

I'd suggest doing the different parts of the landscape (ocean, rivers, ground, landcover, clouds, etc) as separate layers, or even as separate images, so that their colorations can be individually modified. Then merge them for the final image.

[Guest]

Any particular colours in mind?
Decades ago I did study a little botany, and chlorophyll is only one of a number of pigments which could be used by a plant-like ecology;

but it is the most efficient.

Xanthophylls could produce a yellow or brown colour, and there are red coloured algae as well as blue green algae. Perhaps these accessory pigments could replace chlorophyll in part or completely.

Otherwise totally imaginary plants might be purplish or even black, to maximise light utilisation- but a planet of black vegetation would be a challenging texture, perhaps.

[granthutchison]

Black seems perhaps the most likely for a red sun - yellow or red plant life would reflect most of the incoming light, and would therefore be catastrophically ill-adapted. Green or blue-green leaves would be much more efficient but would look almost black under a red sun, given the dearth of short wavelengths for them to reflect.
(I once had a bright turquoise car, which I couldn't find in the car park the night of a vivid red sunset ... eventually I realised the dark grey vehicle was mine.)

Grant

[ACrisp]

Thanks, Selden. I had a suspicion I would have to do something like that, though I was hoping for an... easier solution (okay, I'm lazy :p )

Guest and grant; as far as plant colour was concerned, I was thinking either yellow or cyan, and while the examples you'd provided are certainly plausible, my thoughts are along the lines of a different molecule unlike anything terrestrial life came up with (granted, that's a bit of hand-waving there. Struve 2398 A has a total luminosity of 0.003 sols, but thanks to bolometric correction, only one-sixth of that is visible. On a planet in an orbit where irridation is about equal to Sol's irridation of Earth, Struve A would probably be brighter than a full Moon, but with so much more infrared radiation, local plantlife might not even bother trying to drink in visible light at all. Whatever infrared-absorbing compounds that would be used (would we call this thermosynthesis instead of photosynthesis?) might be some other colour entirely.

And - no offense to your experience, grant - it's likely that even "red" and "blue" suns will still look "white" to local viewers during the day. It's only when viewed from a more remote location that their wavelength bias becomes pronounced. So we'd probably see much of the same colour variety during noontime on Struve A prime as we would on Earth, though the shadows would probably have a reddish tinge to them, and sunrises and sunsets would probably look red regardless of the amount of dust in the atmosphere.

(references for starlight behaviour derived from "World-Building: A writer's guide to constructing star-systems and life-supporting planets" by Stephen L. Gillett; ISBN 0-89879-707-1. All errors are my own).

Should have a texture ready for testing in about a week.

Andrew

[granthutchison]

And - no offense to your experience, grant - it's likely that even "red" and "blue" suns will still look "white" to local viewers during the day.

Ah, I've actually spent the last two years telling people on the forum exactly that, on and off - I'd really like the coloured illumination to disappear from Celestia.
Gillett says only that the light of a red star would appear white. But though you experience various continuous spectra as white light it's evident from everyday experience that that doesn't stop you encountering very noticeable colour-shifts in the objects illuminated - if you're a pale-skinned Celt like me, you notice that you tend to look a little like a corpse under fluorescent light (high colour temperature), but healthily pink under incandescent lighting (low colour temperature). Similarly if you go out looking for your car under fluorescent security lights, the change in perceived colour can be dramatic.
To that experimental fact, you have to add the observation that green vegetation seen from space already looks very dark indeed - almost black over coniferous forests (something discussed in a Scientific American article called Space Coloristics in July 1989). A red star will provide ten times less green light for such vegetation to reflect, so I guess I have to stand by my contention that you'd see green or blue plants as dark-grey/black from orbit.

Grant

[eburacum45]

Perhaps we don’t need to look for unknown pigments:
the two phytochromes Phycocyanin and allophycocyanin absorb light with a peak of 618 and 650nm respectively; on our planet they work in conjunction with chlorophyll in blue-green algae, rather than in competition with it.
But by selecting for plants using these pigments on a red dwarf world (like Dante) they could increase the amount of long wavelength radiation utilised.

[ajtribick]

So the Struve-light appears white but the relative intensities of different colours are different to Sol-light. Would the sky be a much darker blue on a planet in the Struve system even if it received the same intensity of visible light (visible comfort zone, as opposed to temperature comfort zone)?

And what about at blue stars e.g. Regulus?

[granthutchison]

So the Struve-light appears white but the relative intensities of different colours are different to Sol-light.

This is complicated by something called "physiological colour constancy", in which your brain tunes colours to partially compensate for the spectral distribution - so white still looks white, although with a bit of concentration it's evident that the white you see under incandescent light is a yellower white than the white you see under flourescent light. But your brain doesn't compensate for the relative brightness of various coloured patches seen next to each other in a biased spectrum: so if you were looking down on the Struve planet I'm predicting the clouds would look a bright (but slightly yellowish) white, orange deserts would also appear very bright, green vegetation would be so dark in comparison as to appear black, and the oceans would also be very dark in shade (but since they're intrinsically more reflective than vegetation, they'd maybe look dark blue). I've wondered how we could do this automatically in Celestia, but the colour-constancy thing has always thrown me ... we don't want to simply adjust the RGB channels, or we'll throw off the central white balance.

Would the sky be a much darker blue on a planet in the Struve system even if it received the same intensity of visible light

Gillett suggests the sky might be red, but I think he's forgotten the inverse wavelength-to-the-four scaling of Rayleigh scattering - which means blue light is scattered sixteen time more effectively than red. That would partially or completely compensate for the spectral slope of the red star, so I'd guess you'd see a nice continuous spectrum that looked white.
(Hmmm, those oceans are beginning to look grey, aren't they? Depends on how much of the blue we see from space is reflected sky colour, and how much is light that has entered the water and been reflected out again from the depths - my recollection is that the latter is the main contributor when you look straight down on water, and the latter when you look at it on the diagonal.)

For blue stars the spectral slope and the Rayleigh bias will reinforce each other, so you'll see a more blue sky than you do on Earth - I'd suggest a even touch of violet, but our eyes are extremely insensitive to violet, so I don't know.

Grant

[ajtribick]

Thanks, so basically yellow-tinged clouds in a grey sky... except if the habitable zone is a "ribbon" around the terminator, presumably the sky WOULD be red as the sun would be low on the horizon.

[granthutchison]

...you have to add the observation that green vegetation seen from space already looks very dark indeed - almost black over coniferous forests (something discussed in a Scientific American article called Space Coloristics in July 1989).

I just spent a little time trying to find an on-line version of this article I could link to, but no luck.
A quote might be useful (the article is written by Russians, hence the reference to cosmonauts): Paradoxically, several cosmonauts have said after their missions that they did not see any green areas. Forests are very difficult to observe because the intensity of their green colors is at the eye's discrimination threshold. During one coloristic experiment Vasyutin was observing a forested area against the backdrop of the desert on the Mediterranean coast of Libya. The color coordinates he reported corresponded not to a green but rather to a grayish-lilac colour. Subsequent careful observation of the forest showed that it had an earthy color but of very low intensity.
So if it's that hard to make out greenery with the naked eye from Earth orbit, you can see why I think that even a small reduction in the level of green-wavelength illumination will remove all the colour from green vegetation seen from space.

Grant

[ACrisp]

I stand corrected. So much for my lapsed physics studies Hope I didn't give offense, grant.

[granthutchison]

Hope I didn't give offense, grant.

Good heavens, no. I should have explained myself better in the first place. "A tone of authority is never an adequate explanation."

Grant

[Rassilon]

One good way to test it is get colored lights and shine them on objects of various colors...

What I want to know is say a star emits a deep red starlight...Would the oceans reflect a red spectral color or would this blend with the color of the oceans from the start....if the ocean was say blue in white-yellow spectrum light...

Would one somply combine red with blue as an average for the color of the light reflected? or would the light remain red?

[granthutchison]

What I want to know is say a star emits a deep red starlight...Would the oceans reflect a red spectral color or would this blend with the color of the oceans from the start....if the ocean was say blue in white-yellow spectrum light...

You're going to have a specular highlight that's star-coloured, and an ocean that's ocean-coloured, which is a combination of the intrinsic colour of the ocean liquid and any reflected sky-colour. In the example we've been discussing I'd anticipate a steely, grey-blue ocean (blue water colour plus the reflected sky colour) and a yellow-white specular highlight (since we know that the light from a red dwarf is going to look as white as the filament of a lightbulb). If you had a very red brown dwarf in the sky it would produce a red specular highlight on the ocean.

Grant

[Rassilon]

Thanks Grant, so it seems I am color correcting right with the current project I am working on...I am combining the color of the starlight with the overall color of the texture using BlendTexture...This is being done with various phantasy stars I have included such as dark-violet to green or pink stars...May be silly but quite an experiement to see the results of combining these star colors with the color schemes for each planet...Not to mention sunsets...Sunsets I also combine the overall color of the atmosphere with the spectural color of the star...

I tried this with a G2V star and a planet with an atmosphere similar to Earth normal...This combination similated the Rayleigh scattering to a point...giving the correct sunset hue...looked exactly like the one Celestia uses for Earth...

It was as simple as g class spec color: 1.0 1.0 0.8 + blue sky 0.6 0.6 1.0 divided by 2!

Also I am guessing this would be the same for combining spectural light color with the overall color of the texture when creating custom stars...Just use the same color as the spectural light emitted by Celestia and place in the Color tag in your ssc file and combine with BlendTexture true...This I also am doing with my current project...

This however is a bit more difficult when Celestia already does this for various stars such as R N S C T or L classes which emit a dark red that is impossible to blend with any other color correctly...so if you tend to create a custom star I would suggest using a G class star unless the star is dark in color...Then use a R L or even T class star...Adjusting the spectural sub slass to compinsate for temprature as much as possible...

[ACrisp]

Okay, update on my texture:

I settled on a cyan colour, taking a 4k Earth Map and adjusting its colour balance in favour of cyan to get the desired effect, and using a second 4K Earth map (uncoloured) for other parts. As this is my first texture, it's taking me a lot longer than I expected to assemble it, but here are a couple of sample images:



A series of islands in the northern hemisphere.



The planet itself (still "unnamed") with the islands near the top.

Plans are to construct a large (roughly Asia-sized) continent in the southern hemisphere and on the other side of the planet viewed in the second image. Given the planet's slower rotation (due to tidal braking from Struve 3298 A) I'm envisioning a large desert in the core of said continent. The continent's position (and the relative freedom of movement the ocean currents have elsewhere) would essentially prevent icecaps from forming at either pole. Still figuring out tectonics, so I'm not sure where mountains would form on the continent, but the islands were "formed" by two small continental plates moving together, building up the islands we see in the first image gradually (with occasional volcanism to speed things up). After the planet texture is done, I might get to designing a custom cloud template that would suit the planet's climate.

Critiques?

Andrew

[kikinho]

I created a planet with a yellow atmosphere with yellow clouds, and with a yellow ocean reflecting the sky color. It's orbiting a M5V star I created with a radius of 0.50 Rsun. The color is yellow because the atmospheric presure is 6 times greater than Earth, and the light will be more difficult to reach the surface, so I think we would see the sky with a low wavelength color. Or we would see white or purple instead? I put green plants and the color appear very well in Celestia.

[kikinho]

I made this planet:




Here you see the yellowish color due to the reddder color of the M0V star and due to the high atmospheric pressure, that is 6x greater than Earth. Because the light is more likely to be a lightbulb, the clouds and sky are yellowish, and the atmospheric pressure that block more part of the light, leaving lower waveleghs to reach the surface.

No blue color is present because red stars emit many times less blue color than G stars, and much more red, so the sky would be grey, but red stars have yellowish glow, so the sky is a white/yellow with some red in color. A pinkish sky.

My biologist teacher asked me that chlorophil can also absorve redder light with about the same efficiency as absorve the light of the Sun. This is true? If it is, I won't change the color of plants in my planet, but I'm planning to darken it, because red stars also emit less green color.

Do you agree with this? I'm doing this planet to my add-on and I want to make a more realistic possible Earth-like planet orbiting a M0V star. So I need more informations. Is there a website that talk about this?

[buggs_moran]

New info on plant coloring...
http://www.spitzer.caltech.edu/Media/ha ... /20070411/