The history of the civilization along the Nile reaches back beyond 2500 BC. The astronomical contribution to this long drawn out timeline is based on the Sothic Hypothesis, which assumes the Egyptian calendar continued undisturbed for millenia.
The permanence of the Egyptian calendar has not been substantiated. Hence calendar dates earlier than the 6th century BC cannot be considered astronomically validated.
A Babylonian-Egyptian double date found on a papyrus is the earliest absolute Egyptian date. It corresponds to 2 November 473 BC. From that time on, the correlation between the Babylonian and Egyptian calendars appears satisfactory.
Ptolemy was still using the Egyptian calendar 600 years later. Though he lived near Alexandria, astronomical data from pre-Hellenistic Egypt is conspicuously absent from his work. Worthwhile ancient Egyptian observations were unavailable in his day and they remain scarce today.
Stone monuments and written texts from Mesoamerica confirm the Mayans and the Aztecs were avid sky-watchers. Documents that have come down to us include several species of calendars, dates of events, and tables of planetary periods.
The planetary periods are accurately represented, which indicates the Mayans monitored the sky carefully over long periods of time. Records of their original observations have not survived.
Historians have identified at least two Mayan calendars, but they have not been reliably linked to a modern calendar.
The context of Mayan records is conjectural. At one time dates and other text inscribed on monuments were thought to log astronomical observations, but current translations treat them mostly as mundane history chronicling the exploits of kings. Whatever the case, the full import of Mayan astronomy has yet to be revealed.
An ancient chronicle from the Chinese province of Lu, the birthplace of Confucius, cites many astronomical events. The original texts inscribed on bone have not survived, but the authenticity of preserved copies is firmly established.
Ch'un-ch'iu (Spring and Autumn Annals) and its supplement register 37 solar eclipses between 720 BC and 481 BC. Four of the eclipses could not have been seen in Lu. The other 33 records conform to the description of the eclipses computed by modern methods.
The Bamboo Annals are Chinese records of doubtful provenance. The text registers a phenomenon around 900 BC that may be an eclipse. However, none of the many attempts to date the event is convincing.
Similarly, the Oracle Bones from the Shang Dynasty (c.1350 to c.1050) record astronomical observations in the form of divinations, but they provide no absolute dates.
All things considered, the solar eclipse of Feb 22, 720 BC of the Ch'un-ch'iu remains the earliest absolute date preserved in Chinese records. The date is curiously reminiscent of the earliest phenomenon cited in Ptolemy's Almagest -- the lunar eclipse of Mar 19/20, 721 BC.
The Venus Tablet of Ammizaduga is a curious Babylonian record tentatively dated to the 2nd millennium BC. The text registers many observations of Venus over a period of decades, which normally would suffice to date the tablet astronomically. However, never in historical times has there been an orbit of Venus that fits the recorded data.
The preserved copies of the tablet are not sports from a single cache of unknown provenance. The Venus Tablet belongs to the canonical series Enuma Anu Enlil, which was still being copied centuries later in Hellenistic times.
For more than a century, researchers have tried to make sense of the observations recorded in the Venus Tablet. Typically they eliminate the records that do not fit, and alter the text they deem scribal error. In recent years, statistical analysis was enlisted to work out the most probable result. The dates are now reduced to a few contending chronologies - high, middle and low, all in the 2nd millennium BC. This vagueness is typical of astronomical chronology prior to the 8th century BC.
Enuma Anu Enlil and MUL.APIN are two sizable compendiums of astronomical records and related texts. Historians date the records mostly to the centuries and millennia before 747 BC. But like the Venus Tablet, the observations do not tally with conventional computations. Their primary purpose may be divinatory rather than astronomical, or as some historians hold, the observations may simply be crude.
The compendiums draw on an ideal calendar that rounds off the year to 360 days and the month to 30 days. If early Babylonian astronomy were based on an inaccurate month and year, then standard calculations would be ineffective in dating the observations recorded in the ancient documentation.