Raymarched rendering for atmosphere and spherical coordinates

[mate-h]

Objective

• Support space views for the atmosphere
• Add more accurate lighting to the atmosphere with ray marching
• Enable future support for sharp volumetric shadows through the atmosphere
• Tracking overall progress in Atmosphere PBR support and volumetric shadows tracking issue #20374

Solution

• Added spherical coordinate system alongside the sky-view lut parameterization
• Added ray-marched rendering method to support space views and accurate lighting
• This is setting the stage for PBR integration, volumetric shadows
• Could maybe re-use the ray-marching function in the aerial view LUT as well, but I didn't bother for this PR (future improvement)

Testing

• Ran the atmosphere example to ensure it still works fine with both rendering methods
• Created a freecam version of the atmosphere example and put the camera in space, below ground.

---

Showcase

Screen.Recording.2025-09-02.at.12.28.35.AM.mov

[github-actions]

It looks like your PR has been selected for a highlight in the next release blog post, but you didn't provide a release note.

Please review the instructions for writing release notes, then expand or revise the content in the release notes directory to showcase your changes.

[github-actions]

It looks like your PR has been selected for a highlight in the next release blog post, but you didn't provide a release note.

Please review the instructions for writing release notes, then expand or revise the content in the release notes directory to showcase your changes.

[JMS55]

Fix bug where sky turns black below y = 0

Which part fixes that?

Also this is #19131

[mate-h]

Thanks for the comments! I think everything has been addressed, including what to do for when the camera is inside of the earth. I update the atmosphere example here for testing but decided not to include it in this PR to avoid more scope. I would say at this stage it's ready for a final review. I updated the PR description with a new demo video.
https://github.com/mate-h/bevy/pull/13/files

[atlv24]

Fully happy with these changes now. Thanks for being patient :)

One last thing i'd consider is if view_pos never gets used directly (always immediately fed to length+normalize to get r and up) then maybe having get_view_position return a struct that has an up and radius field would make sense. Happy with it as is though, thanks for all the work!

[mate-h]

@alice-i-cecile this one is ready for a final review and merge!
Also note that once this is merged, the PR #20838 by @ecoskey will have to be rebased.

[alice-i-cecile]

I'm happy with this too :) Quite clear, and a feature that a lot of our devs will love.