Sun observation models - Revision history

Flerf Whisperer: Tighten: ad copy only; details live on sunset.observable.science

2026-09-20T06:35:24Z

Tighten: ad copy only; details live on sunset.observable.science

@@ Line 1: / Line 1: @@
-'''Sun observation models''' are side-by-side predictions of what the Sun should look like during a day or a sunset if the Earth is a globe versus if it is flat. Flerfs love to say the Sun "goes away" into perspective. These tools ask them to name the numbers: angular size, elevation, path, and whether the disc actually shrinks the way a nearby lamp would.
+'''Sun observation models''' compare what the Sun should look like on a '''globe''' versus a '''flat''' Earth — angular size, elevation, path. Flerfs say "perspective." These tools make them pick numbers.
-This page collects publicly available models. Each section is one tool or paper trail. Add more as they show up.
+== Sunset Visualiser ==
-== Sunset Visualiser (sunset.observable.science) ==
+'''Claim it kills:''' on flat Earth a nearby Sun sliding toward the horizon must '''shrink'''. On a globe it basically doesn't; it sets because the horizon tips past it. Real sunsets match the globe pane.
-=== The punchline ===
+Four views (globe/flat × camera/side), local-plane or Gleason-map flat paths, location/date/refraction controls. Maths and in-browser tests linked from the site.
-=== What the page shows ===
+* [https://sunset.observable.science/ Open the visualiser]
-Four panes, always:
+* [https://sunset.observable.science/maths.html Maths]
-* '''Globe · camera''' — what a wide lens sees, horizon low in frame; a long-lens "loupe" beside it holds a dashed ring at the Sun's starting size so shrinkage (or the lack of it) is obvious.
+<!-- Add the next model as a new == section. -->
 == See also ==

Flerf Whisperer: Create page; first section: Sunset Visualiser (sunset.observable.science)

2026-09-20T06:32:09Z

Create page; first section: Sunset Visualiser (sunset.observable.science)

New page

'''Sun observation models''' are side-by-side predictions of what the Sun should look like during a day or a sunset if the Earth is a globe versus if it is flat. Flerfs love to say the Sun "goes away" into perspective. These tools ask them to name the numbers: angular size, elevation, path, and whether the disc actually shrinks the way a nearby lamp would.

This page collects publicly available models. Each section is one tool or paper trail. Add more as they show up.

== Sunset Visualiser (sunset.observable.science) ==
'''Sunset Visualiser''' is an interactive, open-source comparison of a sunset on a spherical Earth and on a flat one, live at [https://sunset.observable.science/ sunset.observable.science]. Author: [https://www.youtube.com/@jamesdebono James DeBono]. Source: [https://github.com/james-debono/sunset-visualiser GitHub]. Companion pages: [https://sunset.observable.science/maths.html the maths], [https://sunset.observable.science/tests.html verification tests].

=== The punchline ===
A flat-Earth Sun that stays a few thousand miles up and slides away toward the horizon must '''shrink''' (lose angular diameter) as range increases — the same way a car's headlights look smaller down the road. A globe Sun at ~1 AU barely changes angular size over a few hours of rotation; the disc sets because the observer's horizon tips up past it, not because the Sun flees into the distance. Real sunsets match the globe column. They do not match a perspective lamp.

=== What the page shows ===
Four panes, always:

* '''Globe · camera''' — what a wide lens sees, horizon low in frame; a long-lens "loupe" beside it holds a dashed ring at the Sun's starting size so shrinkage (or the lack of it) is obvious.
* '''Globe · side''' — Earth turns, Sun stays put; horizon line, line of sight, and elevation angle drawn exactly (distances not to scale).
* '''Flat · camera''' — same lens and framing if the Sun were a nearby light moving away over a plane.
* '''Flat · side''' — to-scale angles; adjustable Sun height; two flat path modes:
** '''Local plane''' — Sun recedes in a straight line at the measured speed of the subsolar point (the charitable version).
** '''Gleason map''' — azimuthal-equidistant disc that flat-Earth maps actually use; Sun circles above the pole and the whole path follows from the map.

Controls cover location (globe picker / lat-lon, up to 80°), date, metric/imperial/both, refraction preset, flat Sun height, and whether the flat Sun starts at the real elevation or above the real subsolar point. Settings live in the URL so a moment is shareable.

=== Default scenario (site numbers) ===
Observer on the equator at the September 2026 equinox, eye height 1.7 m, from three hours before sunset through sunset. Clean case: Sun drops straight down at 15°/hour, and the subsolar point moves as fast as it gets (which is the flat model's best shot). Headline comparison quoted by the project for a flat Sun ~1,000 mi up:

{| class="wikitable"
|-
! !! Globe !! Flat (Sun ~1,000 mi up)
|-
| Angular diameter, 3 h before sunset || 0.5311° || 0.5311° (matched at start)
|-
| Angular diameter at sunset || 0.5311° || '''0.1775°'''
|-
| Change || +0.0094″ || '''−21.2′''' (shrinks to about a third)
|-
| Altitude at sunset || 0° || '''13.7°''' (never actually sets)
|}

The flat model's size swing is on the order of '''136,000×''' the globe's. Latitude changes the numbers, not the conclusion: even at 80°, where the Sun only creeps down a shallow slope, the flat disc still loses a large fraction of its width while the globe change is arcseconds.

=== What it is careful about ===
* Both globe distance effects over the window (rotation carrying the observer farther from the Sun, orbit sometimes bringing Earth closer) — net change quoted, not a one-sided half.
* Flat model given its strongest start (correct starting size and elevation; straight-line recession at measured subsolar speed). Gleason / map modes show their own inconsistencies (e.g. wrong speed, wrong set azimuth) on screen.
* Wide-lens edge stretch is modeled so distortion cannot fake a size change; the loupe is a separate camera on the Sun.
* Refraction adjustable and applied to both models; it can squash disc ''height'', not width, so it cannot save the flat shrink prediction.
* Atmospheric extinction (glare collapse) is why a setting Sun ''looks'' smaller to casual eyes; the loupe shows the photosphere size against the starting ring.
* Sky colors are decorative; positions, sizes, angles, and glare flux are computed. Physics lives in pure functions under <code>js/physics/</code>; the browser test suite runs dozens of checks against published values and closed-form identities.

=== Use against flerfs ===
When someone says "perspective," open the flat camera pane and watch the disc collapse while the globe loupe's ring still fits. Ask them which flat height and path they want. Then ask why the real Sun's angular diameter does not do that. If they switch to "it's just the glare," the site already separates glare from disc. If they switch to "the Sun rises up into the dome," they have left the charitable local-plane model and need a new set of numbers — which this page's Gleason mode already humiliates for free.

=== Links ===
* [https://sunset.observable.science/ Sunset Visualiser]
* [https://sunset.observable.science/maths.html The maths]
* [https://sunset.observable.science/tests.html Verification tests]
* [https://github.com/james-debono/sunset-visualiser Source on GitHub]
* [https://www.youtube.com/@jamesdebono James DeBono on YouTube]



== See also ==
* [[Perspective]]
* [[Sun]]
* [[Gleason map]]

[[Category:Evidence]]
[[Category:Globe models]]
[[Category:Flat Earth claims]]