Looks like it was just getting darker. And that's exactly what I'd expect at that time of day.
As a photographer, I can tell you that most people will not notice a 3 stop change in light levels (unless they occur quite quickly). That's an 8:1 range of intensity. It's hardly surprising that you did not notice it was getting darker outside during the time of your measurements.
I agree that it's probably because of the sun going around the earth
, but I was a bit surprised that it was picking this up with such a relatively clean signal (and it was only a fraction of a percent of change in the signal). Of course, I can't leave it alone because now I want to know, so I set up the diode clamped in a vise to look out the window for the next 24 hours and I'll have the data tomorrow. If I don't have a stroke or unless my wife kills me for mouthing off to her (or one of the cats disturbs the setup), I'll post the data.
I was aware from past scientific and photographic work that it takes a goodly change for humans to notice light differences, but I didn't realize it was quite as much as 3 stops. This measured change was only a fraction of a percent of the signal, so I had no hope of detecting it by eye.
Here's a tidbit from when I was a student. I was setting up a Fabry-Perot interferometer and I was in a dark room and looking at the sodium doublet. After a little bit I was a bit surprised to realize that my eye could rather easily see the color difference between these two orange lines (that's about half a nm difference in wavelength if I recall correctly). Of course, if you saw these colors one-at-a-time, you'd likely declare them identical. But if you can compare colors, luminances, etc., you can detect much smaller changes.