If the OP wants a an exact date, I got nothing. However, the ever evolving science of chronodendrology allows us to date buildings (bridges, camps) within years.
Chronodendrology uses the fact that trees have widere or narrower treerings in their stems, depending on how plentiful a year has been (warm/cold, hours of sun, amount of rain, etc). While the circumstances differ from location to location, the patterns remain distinctive. “Lessee…this preserved treestump form the foundation of this bridge, preserved in the bog it was built over, shows five cold years in a row, followed by an extremely warm one? Gotcha! 1005-999 BC! Judging from the rest of the yearrings, the tree was hacked down in 978 BC. So that’s the year this bridge was built.” "
There is one bit of bizarrerie that’s worth noting for its sheer wierdness. Unfortunately I have forgotten precisely what the event was, but it was dated to the year of the “heliacal rising of Sothis” during the 1st or 2nd dynasty of Egypt. What one of these was, is simply the event on a multi-hundred-year cycle when Sothis (Sirius) rises precisely at sunrise on the Summer solstice. Evidently the time of rising shifts by a small amount each year on the Egyptian solar calendar, to the extent that it takes several hundred years (and I don’t recall the precise number) to get back to the point where Sun and Sirius rise simultaneously on the date of the Summer solstice. The net result is a date somewhere around 3000 BC which, though purely astronomical, pins down the date of some Pharaoh’s fourth year of reign.
Actually, the Sothic cycle is a different animal than precession. The Egyptian calendar had only 365 days in a year, with no leap years, so New Year’s Day would drift through the seasons over a cycle of roughly 365.25*4 = 1,461 Eqyptian years. There are some complications due to the difference between the sidereal and tropical year–which is due to precession–but the primary driver was the 365-day year. Note in the wiki article, however, that there is inherent uncertainty in dating even to the year via the Sothic cycle, much less to the day.
That’s exactly the problem with any non-astronomical dating from before the early Christian Era–we may have a reliable record in the local calendar, but it’s impossible to exactly align the local calendar with our modern calendar. We can’t even align the early Julian calendar with the later Julian calendar, because of the problems you mention.
As an aside, though, I’d be interested to know the earliest event for which we have a written day-month-year record in the local calendar, even if we can’t translate it.
Any solstice-based references are complicated by the fact that it’s very difficult to pin down the precise date of the Solstice. Based on simple naked-eye observations (i.e., not using any day-counting), the best you can get is within a week or so. One can, however, determine the Solstice very accurately (that is to say, you can be very certain that the true date of solstice is somewhere in that week). The equinoces, on the other hand, have the opposite problem: One can very precisely pin down the date of either equinox, but accuracy suffers. In other words, you can be very sure that you’re marking the same day as equinox this year as you did last year and the year before, but you can’t be sure that that same day is the correct one. Only once you start counting days can you be sure of both accuracy and precision for both the solstices and equinoces.
I’m surprised that more pieces from Chinese history aren’t popping up- haven’t they had accurate calendars for longer than most of the rest of the world?
I had hoped that an ancient eclipse would be recorded in terms of an ancient calender, which would let us time-sequence events occurring earlier. I guess the Lydians and the Medians didn’t carry Filofaxes.
Great answers gang; the tangents have been interesting as well.