We broke 12 hours of daylight Friday, 2 days before the equinox. Why? And is this just due to my being close to 65 degrees North, or is it a more general anomaly?

There are two parts to this peculiarity. One is latitude combined with the finite diameter of the sun, which can be calculated. The other is the refraction of the atmosphere, which varies from day to day and can only be estimated.

Let’s take latitude first. Sunrise and sunset are defined as the time that the upper edge of the sun is just visible above a flat horizon. “Equal days and nights” (which is what equinox means) assumes the dividing line between day and night is the time when the center of the sun is on the horizon, assuming light moves in straight lines. If the sun rose vertically, as it does at the equator, it would rise at a rate of about 1 solar diameter a minute, and the calculated sunrise time based on the center of the sun would be only half a minute after the time the upper edge first showed.

At higher latitudes, however, the sun appears to rise at an angle and sunrise and sunset appear slower. At 65 degrees latitude the sun’s path at the equinox is 65 degrees from the vertical, and a little trigonometry stretches that half minute to about 1 minute 10 seconds, or twice that in day length. Latitude alone is still not enough to allow our days to be 12 hours 15 minutes long at the equinox. For that, the refraction of the atmosphere becomes important.

The apparent break in the spoon handle is due to refraction.

Everyone is familiar with refraction, though you may not know it by that name. The optical illusion of a broken spoon in water is caused by the fact that the speed of light in water is less than that in air. Yes, the speed of light in vacuum is constant, but in any other transparent medium it moves a little slower. When it crosses a boundary between two transparent media with different speeds of light, any light rays not moving at a right angle to the boundary are bent. Air is one of those transparent media, and while the speed of light in air is not a great deal slower than that in vacuum, there is enough of a difference that the bending affects what we can see.

The actual difference in speed depends on the density and moisture content of the air, which in turn depend on pressure, temperature and relative humidity. Air near the ground is almost always denser than that above it, and this is particularly true at sunrise. The change with height is gradual, and thus the light rays are not bent sharply, as in the water-air interface, but curved along the earth’s surface. Objects far away appear higher than they are, and this certainly applies to the sun at sunrise. The amount by which the sun appears higher in the sky than it really is will depend the atmospheric density and how it changes with height.

For practical purposes the time of sunrise is calculated assuming that the upper edge of the sun is visible when the center of the sun is 50 minutes of angle—almost a degree—below the horizon. This also means that the sun at the equinox will rise not quite due east, as it “rises” while it is still physically below the horizon and slightly north (in the northern hemisphere) of east. The difference, however, is slight.

Refraction is also responsible for the fact that the sun appears to flatten as it approaches the horizon when setting or just after rising. The part of the sun closest to the horizon is more strongly affected by atmospheric refraction than is the upper part of the sun, so the two appear pushed together and the sun appears flattened, rather than round. I’ve probably overused this in Tourist Trap.