Aurore glacial coverage IMTU

[sudnadja]

A depiction of Aurore IMTU:

• Aurore has a 9.14 hour orbit around Tithonus, which is a 55 jupiter-mass brown dwarf. Effective thermal insolation from Tithonus is about 1300 W m^-2 and 345 W m^-2 from Muphrid.
• Most of the light comes from Muphrid. Midday when it is directly overhead it provides about 20,000 lux, while Tithonus provides around 70 lux (mostly in red). Humans have a natural advantage during night operations compared to the Kafer, as human vision extends into the red that Tithonus primarily emits in, while Kafer cannot see this range.
• There are no prograde orbits stable around Aurore. Retrograde orbits are only stable for a short time, and only at specific distances from Aurore. The objects either crash into Aurore or are pulled into independent Tithonus-centric orbits.
• as can be seen in the image, glacial coverage on my version of Aurore is much more substantial than in either the sourcebook or the atlas of the french arm. This comes from computing average yearly insolation for each point, but neglecting thermal convective transport on the surface (that's later).
• "Satellite" communication comes from constellations that are places at the Tithonus-Aurore L4 and L5 points, which are stable long term. The cold pole (and within 30º of the cold pole) cannot see these points in the sky, though. Nova Kiev communicates to the L5 satellite and Tanstaafl communicates to L4. Communications between Nova Kiev and Tanstaafl have to go Nova Kiev-L5-L4-Tanstaafl. Communications are via near-optical laser and can be disrupted by cloudcover. There are no global lines aside from that.
• The star field in the image is accurate as viewed from near-eta bootis space, as recorded in the NSL. For the volume of space outside of the NSL, the data is supplemented with Gaia DR2 and the bright constellation stars list. In this image, the bright star to the center left of the image near the top is Vega.

 

[shaunhilburn]

Awesome. What did you render this with?


I did something similar a long time ago:



I usually leave out star fields; I like the featureless inky black like space photography. It's probably not an "astronauts-eye" view, though. You can see stars if the view doesn't have bright objects and/or the sun glare.

 

[sudnadja]

Awesome. What did you render this with?


I did something similar a long time ago:



I usually leave out star fields; I like the featureless inky black like space photography. It's probably not an "astronauts-eye" view, though. You can see stars if the view doesn't have bright objects and/or the sun glare.

Blender, with the base map converted from icosahederal to eqirectangular in mathematica and cleaned up there. Glacial coverage also computed in mathematica, after the orbit simulation is compute and net insolation computed. Climate model (well, greenhouse model) running in mathematica as well. The output of that goes into various textures used for aurore, there are separate ones for glacial coverage, the seas, land, and a height/displacement map.

The stars would realistically probably be too dim to be captured along with a full-disc world in frame as well, though I do make an attempt to compute the stars apparent magnitude (using phot_g_mean_mag if the source was GaiaDR2 for that particular star, or MagV if it was simbad - recomputed for the appropriate distance to eta Bootis, but those are two different passbands) and relate it to pixel brightness, but it's difficult to control that in blender. Additionally, in post processing in post of the render transformations the dynamic range is compressed and that brightens up the stars. I'm just winging it right now.

What did you do yours in? It looks very good.

 

[shaunhilburn]

Blender, with the base map converted from icosahederal to eqirectangular in mathematica and cleaned up there. Glacial coverage also computed in mathematica, after the orbit simulation is compute and net insolation computed. Climate model (well, greenhouse model) running in mathematica as well. The output of that goes into various textures used for aurore, there are separate ones for glacial coverage, the seas, land, and a height/displacement map.

I used Bryce 7. The Aurore surface and specular maps I drew freehand. They look like crap which is why I can't show this model from certain angles. Tithonus just has a Jupiter map. The scene file is built to scale and there are moons not shown in this shot that have procedural shaders.

The stars would realistically probably be too dim to be captured along with a full-disc world in frame as well, though I do make an attempt to compute the stars apparent magnitude (using phot_g_mean_mag if the source was GaiaDR2 for that particular star, or MagV if it was simbad - recomputed for the appropriate distance to eta Bootis, but those are two different passbands) and relate it to pixel brightness, but it's difficult to control that in blender. Additionally, in post processing in post of the render transformations the dynamic range is compressed and that brightens up the stars. I'm just winging it right now.

There is a freeware product called "Space Engine", sort of like Celestia. It lets you visit virtual solar systems, land on worlds, view the scenery and take screen captures and the like. One of its features you may find useful is "Export Skybox". Once you set the magnitude thresholds to you liking you can capture the Eta Bootis star field.

It should automatically take care of the dynamic range and anti-aliasing. I haven't tried exporting skyboxes, but I have exported planet textures. 2K horizontal resolution seems to be it's max; it will crash if you try to export at higher quality.


I like that you've worked out many of the astronomical details - which are much the same values I arrived at. One detail to consider is that large bodies in close proximity (like this moon system) have their orbital periods fall into resonances. The most stable, hence most likely is a 1:2:4 Laplace resonance, like Jupiter's Galilean moons.

 

[sudnadja]

I used Bryce 7. The Aurore surface and specular maps I drew freehand. They look like crap which is why I can't show this model from certain angles. Tithonus just has a Jupiter map. The scene file is built to scale and there are moons not shown in this shot that have procedural shaders.



There is a freeware product called "Space Engine", sort of like Celestia. It lets you visit virtual solar systems, land on worlds, view the scenery and take screen captures and the like. One of its features you may find useful is "Export Skybox". Once you set the magnitude thresholds to you liking you can capture the Eta Bootis star field.

It should automatically take care of the dynamic range and anti-aliasing. I haven't tried exporting skyboxes, but I have exported planet textures. 2K horizontal resolution seems to be it's max; it will crash if you try to export at higher quality.


I like that you've worked out many of the astronomical details - which are much the same values I arrived at. One detail to consider is that large bodies in close proximity (like this moon system) have their orbital periods fall into resonances. The most stable, hence most likely is a 1:2:4 Laplace resonance, like Jupiter's Galilean moons.

Unfortunately (or not), I'm on a mac so I can't use Space Engine without some system swapping gymnastics. I have other reasons for using my own astronomical set though, first, Gaia DR2 isn't quite in "ready to use" condition and has a lot of false hits in it that need appropriate filtering. I very much doubt that space engine is making use of it correctly either, if that is one of its data sources. I also generate sky charts that could be used in game, in certain situations, which I don't know how to do with space engine:



As you can see at the tail end of this video: Aurore Satellite Orbit StabilityI do have Memnon, Selene, Aurore and Antilochus in 1:2:4:8 MMR, though they are not stable in this configuration long term with the integrators I use - Symplectic partitioned RK and velocity verlet - I suspect because of the somewhat complex gravitational environment of eta bootis as it is in 2300 with rubis and tithonus putting muphrid in a kind of complex motion, but haven't really looked into that yet and it could just be a nuance of those integrators.

 

[sudnadja]

I used Bryce 7. The Aurore surface and specular maps I drew freehand. They look like crap which is why I can't show this model from certain angles. Tithonus just has a Jupiter map. The scene file is built to scale and there are moons not shown in this shot that have procedural shaders.

I use Saturn's surface for Tithonus, with the vague and unsubstantiated reasoning that the ~1500k surface temperature would have the effect of not allowing large scale cloud structures to form:



A number of people have objected to this, because, I think, they picture Tithonus as a mega-jupiter (though it actually has smaller volume than Jupiter).

 

[shaunhilburn]

I use Saturn's surface for Tithonus, with the vague and unsubstantiated reasoning that the ~1500k surface temperature would have the effect of not allowing large scale cloud structures to form:

I tried different Tithonus shaders and decided to go with the Jupiter map because the book mentioned some similarity ... "bands, swirls, and streaks similar to those in the atmosphere of Jupiter are clearly visible." - a purely artistic choice with no scientific basis.

But you're right about cloud structure regarding volatiles like water vapor or ammonia. It's more likely a Sudarsky class IV or V jovian with metal or silicate clouds.

https://en.wikipedia.org/wiki/Sudarsky_extrasolar_planet_classification

 

[shaunhilburn]

I do have Memnon, Selene, Aurore and Antilochus in 1:2:4:8 MMR, though they are not stable in this configuration long term with the integrators I use - Symplectic partitioned RK and velocity verlet - I suspect because of the somewhat complex gravitational environment of eta bootis as it is in 2300 with rubis and tithonus putting muphrid in a kind of complex motion, but haven't really looked into that yet and it could just be a nuance of those integrators.

Not stable? You have Tithonus set at 55 Mjup and several AU away from system barycenter. It must be a glitch; that setup should be rock-solid for billions of years.

 

[sudnadja]

Not stable? You have Tithonus set at 55 Mjup and several AU away from system barycenter. It must be a glitch; that setup should be rock-solid for billions of years.

I don't mean unstable to the point of completely destabilizing the orbits, but that they don't remain locked in MMRs for long periods. The same integrator does keep the jovian moons in MMRs for very long periods, so I'm not sure what's up.

I have tried other integrators with similar effects. Another problem is Aurore is relatively massive and tends to pull the other moons into eccentric orbits. I had arbitrarily set the moons to be at an inclination relative to Tithonus' orbit around Muphrid in order to provide seasons and that, I think, is somewhat destabilizing too.

The L1 and L2 Aurore-Tithonus points are not stable even for one aurore orbit, though L4 and L5 are stable for long periods. I think it's actually worse that Tithonus is so massive and that the moons are so close.

 

[sudnadja]

I tried different Tithonus shaders and decided to go with the Jupiter map because the book mentioned some similarity ... "bands, swirls, and streaks similar to those in the atmosphere of Jupiter are clearly visible." - a purely artistic choice with no scientific basis.

But you're right about cloud structure regarding volatiles like water vapor or ammonia. It's more likely a Sudarsky class IV or V jovian with metal or silicate clouds.

https://en.wikipedia.org/wiki/Sudarsky_extrasolar_planet_classification

There are at least two different versions of Tithonus out there. The original version in which Aurore was at a distant orbit (~60 hour), then the French Arm Adventures version where it is pulled in to a ~8.9 hour orbit and I think gained a lot of mass. My own version has it at a 9.14 hour orbit, because while we both use 55 jupiter masses for Tithonus, I have a different reference number for Jupiter's mass, resulting in a slightly different orbital period at the same distance.

 

[sudnadja]

Not stable? You have Tithonus set at 55 Mjup and several AU away from system barycenter. It must be a glitch; that setup should be rock-solid for billions of years.

Note how substantial a wobble Rubis imparts on Muphrid - when plotted from a Muphrid-centric frame, the oscillations in the orbits become pretty obvious: