On a flat earth, the horizon is relative to the observer, for example -
I fail to see where that picture is any indication of a horizon.
It seems to be just a picture taken with a wide angle lens camera looking down a long corridor.
It is in no way what you see when looking at the horizon.
The definition of the horizon is "The line where earth or sea appears to meet the sky."
I am a "round earth" person, too, but I will make an attempt at an explanation for the horizon on a flat earth.
If you were in a ship, sailing around the earth, just a few miles off shore from the ice wall, but close enough so that the ice wall would not be obscured by "atmoplanic" conditions, haze or fog, the horizon would be a distinct line where the top of the 150 feet tall ice wall and the bottom of the dome appear to meet directly in front of you and a few miles to either side. As you looked to the sides or rear of the ship, the view would gradually fade away in a indistinct blur.
If you were in the middle of the ocean, you would not see a horizon. You would just be in the middle of a circle, looking into a blur in all directions.
I know from experience what the horizon really looks like. It certainly doesn't look like the description above. I have a difficult time trying to imagine what the huorizon would look like on a flat earth. That was the best explanation I could come up with.
The horizon, or lack of it, seems to me to be one of the simplest and most elementary fallacies of flat earth.
No it's caused by perspective.
The further an object is from the observer, the more it moves towards the center of vision.
Like if you have two objects approximately the same size and horizontal distance from you, the one further away from you in vertical distance will be closer to the center of vision.
For example -
With bigger objects and objects placed further apart from one another, like the earth, moon and sky, the sun and stars, it takes longer for them to move towards the center of vision, but they inevitably do.
Furthermore, refraction could be causing the sun to appear to sink below the horizon, when it's really just moving further and further away from the observer.
Refraction (as well as obfuscation) happens more across vast distances and when objects approach sea level because the atmosphere light travels through on the way to your eyes is denser and more heterogeneous.
So the horizon is relative to the observer, and the objects surrounding him.
Rather than speaking of the horizon, we may speak of horizons produced by differing circumstances.
Another example -
You may criticize the photograph, say it's produced by a certain sort of camera, but it's not, and we can verify that it's a natural effect of our vision, not an artificial effect of a camera, just by taking a walk, and being conscientious of your surroundings.
Show me a photograph of a long hallway that doesn't have this effect.
There is a natural limit to how far the human eye can see, but this limit is relative, it depends on a few things, how bright things are, how thin the atmosphere is, how perceptive a persons eyes are, and the width of the objects observed, both the width of the objects themselves, and the width in between them.
The greater the width, the farther you'll be able to see them, until they merge.
Again, the horizon is relative.