3 ms·
The calculated daylight even downplays the actual light at the high (or low) latitudes quite a bit. E.g. at latitude 60 there's a "nominal" midsummer night of a
by jampekka 2y ago
The calculated daylight even downplays the actual light at the high (or low) latitudes quite a bit. E.g. at latitude 60 there's a "nominal" midsummer night of about four hours, but it doesn't really get dark, as the light from the refraction is quite strong even with the disk not being visible.
- sunshinesnacks 2y agoI think the section “Atmospheric refraction and the solar limb” near the end of the article addresses this.
- madcaptenor 2y agoNot exactly; even after the sun is fully below the horizon there's some light. See the interactive graph - the times aren't long enough for twilight. But you can use a larger value of "a" in the equation above it to make that change.
- sunshinesnacks 2y agoYeah, you’re right. Something like civil twilight, as you mentioned above, should get it closer to what is being discussed here.
- madcaptenor 2y agoThat's called "twilight". In the section "Atmospheric refraction and the solar limb" there's a modified sunrise equation with a variable "a" and text suggesting a correction of 50 minutes of arc there. If you replace 50 minutes with: - 6 degrees, you get the times of "civil twilight" (roughly speaking, when you don't need light outside). At 60 degrees north at midsummer the minimum altitude of the sun will be about -6.5 degrees so almost all of the nominal night is civil twilight. - 12 degrees is "nautical twilight" (horizon clearly visible) - 18 degrees is "astronomical twilight" (sky is dark enough for all astronomical observations). (It's possible that those are defined as 6 degrees + 50 minutes, etc.)
- qwertox 2y agoAnd it's not just "sunset + X minutes" where X would be a constant. In the winter this X is much shorter than in the summer. For me X is then a "travelling" value which in the winter is at 1/3 between sunset (0°) and civil dusk (-6°), and in the summer it goes up to 2/3. Calculated as the moment when the blinds go down. I'm using https://astral.readthedocs.io https://astral.readthedocs.io for this.
- madcaptenor 2y agoWhat's the reasoning behind using a different value in winter than summer? Your winter value roughly corresponds to -2°, your summer value to -4°.
- qwertox 2y agoYou made me check the code and actually I have it the other way around (and with different ratios, winter=-6°, summer=-2°), because else the blinds would go down too late in the summer (after 22:00).
- jameshart 2y agoEven in the UK around 51-52°N you get a few nights around midsummer where it never truly gets dark - no darker than ‘nautical twilight’ at least.
- pcthrowaway 2y agoI just pulled up Whitehorse, Yukon (at the 60th parallel), and sure enough, the longest day of the year is essentially always day: https://www.timeanddate.com/sun/canada/whitehorse https://www.timeanddate.com/sun/canada/whitehorse. Almost 5 hours of civil twilight, but civil twilight is still daytime in my book.
- Sharlin 2y agoTwilight is caused by scattering rather than refraction. There's sunlight bouncing around in the atmosphere, keeping the sky non-black even after sunset.
- MarceliusK 2y agoEven during the "night," the sky stays illuminated enough that it never truly feels dark. It definitely challenges the idea of what nighttime means and makes the summer feel almost magical, but I imagine it can also mess with your sleep patterns!