This data takes the entire history of San Jose's climate and plots the maximum temperatures against the number of days elapsed since the winter solstice each year.
In order to generate a better average, each daily data point is skewed to the average time of the highest temperature (when the sun is about 17 degrees below its peak). Probably doesn't matter much (especially since it's not guaranteed to be the time of highest temperature) but I wanted to account for it.
I then generated a smoothed mean, median and "inner" mean over the 20 day period around each data point, which is what you see here.
(I also generated 1-day, 5-day and 10-day means that are... less smooth)
End result is that the lowest daily high in San Jose is usually about 7 days after the winter solstice and the highest is about 206 days after the winter solstice (i.e. about 24 days or so after the summer solstice).
Not very useful for timing the cross-quarter days. Too close to the main solar days.
Still, I plan on comparing these to the figures for London soon, just to confirm that this image is right: https://commons.wikimedia.org/wiki/File:Phases_of_the_Sun_(NHemi).png
Turns out London wasn't in the daily summary set, so I used Oxford instead.
Bit more data to process this time, since Oxford has records back to 1814 or thereabouts. But I wasn't able to round up any local climatological data so I just went with the same "number of degrees down from solar noon" as San Jose. Might be wrong.
Anyway, you can see that the curve is much more sinusoidal, and delayed. The coldest day is about 20 days after the winter solstice (as opposed to about 40 between Yule and Imbolc) and the hottest day is about 210 days after the winter solstice, which is about 28 days after the summer solstice. Not quite the 40 days from Litha to Lughnasadh, but not unreasonable.
Close enough fit to feel comfortable anyway.
End result is that it's *probably not* unreasonable to put a Californian Lammas/Lughnasadh on August 1, but Candlemas/Imbolc on February 2 is a fool's game.
Probably should have some other marker for the latter (end of the rainy season?)
Anyway, I figure that the Californian quarter days are a given (we're tracking the solar year after all), though their meanings might differ. For example, the Yule counterpart not only symbolizes the rebirth of the sun but also the beginning of rewarming just a week later. More strongly than Yule normally would I mean.
Ostara is roughly going to align with the end of the rainy season I expect, so that will likely be a relevant symbolism.
Mabon might be pretty close to the start of high fire season, so again, probably going to just lean on that a bit.
Litha is a bit of an open question with respect to symbolism, though.
Cross-quarter days are a bit harder at the moment.
There's an anecdotal pattern that I've noticed where San Jose usually has a multi-week break in rains in January/February, so depending on what I find in that upcoming analysis, Imbolc/Candlemas may stay in about the same place and symbolize that rather than the coldest point in the year.
Likewise, Lammas/Lughnasadh can still serve as the "hottest" festival at about the same time too.
Samhain might align to the start of the rainy season, but this leaves Beltane unaccounted for.
I might be able to use something like the "official" fire season declaration (which is about that time) for that, not sure.
