2024 Eclipse – Twilight Darkness Measurements

[Captain Scarlet]

Here in SW Ireland our experience of the recent 2024 April 8th Solar Eclipse was that it was to end in the Atlantic just short of us. If lucky, just before sunset we’d get to actually see about a 35% eclipse very low to the horizon. I was lucky: I glimpsed it for about 4 seconds in a gap in the horizon-clouds. But the closest to totality, somewhere about 90%, was actually going to happen shortly after the sun had set, when about 5 degrees down. In other words, during Civil Twilight.

So, I thought, Twilight should be for a while significantly, perhaps even noticeably, darker than normal. I decided to take SQM-L sky-darkness measurements in the minutes leading up to maximum coverage and compare them to what I would expect on a normal evening.

In the event it wasn’t naked-eye noticeable. But it certainly shows up in the data.

Since early 2020 I’ve been collecting darkness data from my now-home site using a Unihedron SQM-L. Every time I’ve noticed a clear dark sky, I’ve nipped out and recorded the darkness, often several times. Using that data, together with concurrent Moon altitude and phase, Sun altitude and angular proximity of the band of the Milky Way, I’ve built a simple statistical regression model to be able to predict what darkness I can expect on a given night. Now that I have nearly 400 data points, the model is usefully accurate. One “darkness-factor” I have noticed but haven’t yet included in the model is “time of night”, to the extent that early evening after dark is definitely brighter at zenith than late evening when everyone’s turned their lights out, so it can certainly be made more accurate. For a moonless night, I can expect to predict a darkness value to within 0.4 magnitudes, and within about 0.2 in astronomical darkness. Milky Way proximity seems to make a surprisingly large difference, up to 0.4 magnitudes.

The chart shows darkness-value (y-axis) plotted against the altitude of the Sun (degrees below horizon). It shows that a typical evening displays a steady darkness-progression as the Sun sinks through the various Twilights, levelling out at whatever the local LP allows. An aside, regarding darkness for places with different levels of Light Pollution: I’ve found that as the Sun sinks through the Twilight Zones, (6, 12, 18 degrees down), measured darkness will be the same regardless of location until the local “LP” level is reached, whereupon the darkness “gets off at its bus-stop” and stays at a certain twilight level. But that’s an article for another day, I think.

No prizes for guessing which the “Eclipse Readings” are. They’re the ones at the far lower right. The red points are the actual darkness readings (LHS), blue points are divergence from predicted modelled values (RHS). When I started taking the readings, the Sun was 3.5 degrees down and the Eclipse was around 75-80% eclipsed. By the time of my final readings, it was 4.9 degrees down and the eclipse closer to 90%. The blue diamonds show that twilight went from over a full magnitude darker than expected, to fully 2 magnitudes darker. This was not noticeable by eye, as the eye was adapting much faster than the light was changing. Another aside: In a full total eclipse, which I experienced near Falmouth in Cornwall in 1999, as totality happened the darkness fell suddenly much faster than my eye could adapt. It really was someone turning the dimmer switch down!  It was like a curtain falling. I’ve not read that phenomenon described for this one yet.

Anyway, an interesting little snippet of data.

Thanks, Magnus

Edited April 10 by Captain Scarlet

[Stephenstargazer]

Congratulations! It occurred to me that the only possible effect this side of the pond was a dimming of twilight. As it was raining here I stayed inside and followed news coverage. How sad that a response of so many was to get out a bright phone during totality ☹️ 

Edited April 10 by Stephenstargazer