Did the Maya Predict Solar Eclipses Accurately?

Yes — to within a day, and sometimes to within hours, the Maya predicted both solar and lunar eclipses. The most important surviving text is the Dresden Codex eclipse table (pages 51–58), a 405-lunar-month almanac that allows the daykeeper to determine, for any given day, whether an eclipse might be visible and whether it would be solar or lunar. The table can be checked against the historical record of eclipses visible from the Maya area between roughly AD 700 and 730 (the most likely date of the table’s last revision), and it predicts eclipses to within a day of the actual event. The table is independent evidence of the precision of Maya astronomy and of the accuracy of Maya astronomical predictions.

This page describes the structure of the Dresden eclipse table, the 405-lunar-month cycle on which it is built, the way the table is read in practice, the way the lunar series in the inscriptions confirms the GMT correlation between the Long Count and the proleptic Julian calendar, and the role of eclipses in Maya ritual and political life.

The Dresden Codex eclipse table

The Dresden Codex is the most extensive surviving Maya book, with 78 pages of almanacs, astronomical tables, and ritual texts. Pages 51 through 58 of the Dresden contain the eclipse prediction table, a 405-lunar-month almanac covering 11,960 days = 32.75 tropical years. The table is divided into 46 rows of 6 months each, and each row records a Tzolkin position, a 6-month count, and a set of glyphs that indicate whether the eclipse is solar or lunar, possible or impossible, dangerous or routine.

The structure of the table is as follows:

The table has 46 rows, each representing a 6-month eclipse half-year.
Each row is keyed to a Tzolkin position (the day-sign of the new or full moon) and a 6-month count (1 through 46).
The 6-month count advances by 1 every 6 months, and the cycle returns to 1 after 46 half-years (i.e., after 405 lunar months).
The daykeeper reading the table could determine, for any given day, whether an eclipse might be visible and whether it would be solar or lunar.

The 405-lunar-month cycle is built on the fact that 405 lunations ≈ 46 eclipse half-years. The actual length of 405 synodic months is 405 × 29.530588 = 11,959.89 days, or 32.75 tropical years. The actual length of 46 eclipse half-years (each 173.31 days) is 7,972.27 days, or about 21.83 years. The two cycles are not exactly equal, but the 405-month cycle returns to the same eclipse possibility every 405 months, with a small correction.

The 405-lunar-month cycle

The 405-lunar-month cycle is the basis of the Dresden eclipse table. The cycle is also tied to the Tzolkin, the Haab’, and the Long Count:

405 lunar months = 11,960 days
11,960 ÷ 260 = 46 Tzolkin cycles exactly
11,960 ÷ 365 = 32.76 Haab’ years (with a small correction)
11,960 = 33 × 360 + 80 days in the Long Count tun count

The cycle is “exact” with respect to the Tzolkin (because 405 × 29.53 = 11,959.65, and 11,959.65 ÷ 260 = 45.998, very close to 46), and the daykeeper could use the table to predict the eclipse half-year position of any given day.

The 405-month cycle is also tied to the 819-day cycle of Caracol: 405 months ≈ 11,960 days, and 11,960 ÷ 819 = 14.6, so the 405-month cycle returns to the same 819-day position only after several cycles. The 819-day cycle and the 405-month cycle are mathematically independent, although both are used in the Late Classic inscriptions as ways of structuring longer time periods.

The accuracy of the table

The accuracy of the Dresden eclipse table is impressive. The table predicts eclipses to within a day over a 32.75-year span, and the table can be checked against the historical record of eclipses visible from the Maya area between roughly AD 700 and 730. The most likely date of the table’s last revision is AD 755 ± 30 years, based on the position of the table in the Dresden Codex and on the consistency of the eclipse records with the actual eclipse history.

A specific example: the eclipse of 16 July 790 (a total solar eclipse visible from the Maya area) is predicted in the Dresden table to within a day. The eclipse of 14 October 792 (a partial solar eclipse) is also predicted. The table’s predictions are not perfect, but they are accurate enough to be useful for ritual planning.

The accuracy of the table is comparable to the Babylonian and the Greek eclipse tables of the same period, and it is the densest numerical record of any pre-Columbian astronomical tradition. The table is evidence of the empirical, observational, and arithmetic sophistication of Maya science, mathematics, and astronomy.

The lunar series on monuments

The Lunar Series in the Late Classic monumental inscriptions is a parallel corpus of astronomical data. The Lunar Series records:

The lunar age (the number of days since the last new moon, 0–29)
The lunar month number (the count of the 29.53-day synodic month within a 6-month “lunar half-year”)
The lunar belt glyph (a K’uhul glyph indicating the 6-month cycle of the moon)
The lunar fortnight glyph (a glyph indicating the 12- or 13-day count of the bright or dark half of the moon)
The lunar season glyph (a glyph indicating the dry or rainy season of the lunar month)

A typical Lunar Series might read: 8 moons, 4 lunar half-years, “the moon is in the 4th fortnight of the dry season.” The Lunar Series allows the epigrapher to determine the exact lunar age of a given date, and it is the basis of the GMT correlation between the Long Count and the proleptic Julian calendar.

The Lunar Series is attested from the Late Preclassic through the Late Classic at Palenque, Copán, Tikal, and other sites. The Lunar Series is one of the most precise astronomical texts in the pre-modern world, and it is the basis of the modern chronology of the Classic Maya.

The way the table is read

The daykeeper reading the Dresden eclipse table for a given day would:

Find the day in the 260-day Tzolkin.
Find the corresponding row of the eclipse table (each row of the table is keyed to a Tzolkin position).
Read the eclipse type: solar possible, lunar possible, or impossible.
Read the 6-month count: the position of the day in the 46-half-year cycle.
Determine whether the eclipse would be visible from the local latitude.

The daykeeper could also use the table to predict the time of the eclipse to within a few hours, by cross-referencing the Tzolkin position with the local solar position. The table is, in this respect, a sophisticated eclipse prediction tool that is tied to the almanac tradition and the Long Count.

The table is not a “universal” eclipse prediction tool; it is keyed to the Maya area and to the local eclipse history. The table would not, for example, predict an eclipse visible in Central Mexico but not in the Maya area. The table is, however, accurate for the Maya area, and it is the most sophisticated eclipse prediction tool of any pre-Columbian civilization.

The role of eclipses in ritual life

Eclipses were, in the Maya ritual tradition, times of cosmic vulnerability, when the sun or the moon was “eaten” by the Xibalba lords. The daykeeper reading an eclipse prediction would advise the community on the appropriate response, which might include:

Fasting and abstention.
Ritual bloodletting and sacrifice.
The recitation of almanac texts.
The performance of the way (Lords of the Night) cycle.
The consultation of the Tzolkin for the appropriate almanac.

Eclipses were, in the inscriptions, times of royal ritual. The 9.16.4.10.17 Long Count at Palenque, for example, is tied to a lunar eclipse, and the inscription records a royal ritual response. The 9.12.2.0.16 Long Count at Copán is tied to a solar eclipse, and the inscription records a similar response.

The 8 December 683 total solar eclipse at Palenque, recorded on the Oval Palace Tablet, is one of the most carefully documented eclipse events in the inscriptions. The eclipse coincided with the death of K’inich Janaab’ Pakal, and the inscription ties the eclipse to the king’s passage to the underworld.

The Maya solar eclipse risk

Solar eclipses visible from the Maya area occur roughly every 1.5 years on average, with totality visible from any one site only every 350–400 years. The total solar eclipse of 11 July 1991, visible from Chichen Itza, was the first total solar eclipse visible from that site in the Late Classic period; the next will not be until 2112. The partial solar eclipses of 26 February 1998, 8 April 2024, and 14 October 2023 were visible from the Maya area, and the daykeepers of the highland communities observed them with the same ritual responses that the Classic-period daykeepers would have used.

The Dresden eclipse table predicts solar eclipses to within a day over a 32.75-year span, and the table’s accuracy is sufficient for the daykeeper to know, for any given day, whether a solar eclipse is possible. The table is, in other words, a useful eclipse prediction tool, and the Maya astronomical tradition is one of the most sophisticated in the pre-modern world.

The lunar eclipse risk

Lunar eclipses are more frequent than solar eclipses, and they are visible from a larger area. Lunar eclipses visible from the Maya area occur roughly every 1.5 years on average, and the Dresden eclipse table predicts lunar eclipses to within a day over a 32.75-year span. The table is, in this respect, a useful lunar eclipse prediction tool, and the daykeeper could use the table to plan ritual responses to lunar eclipses.

The lunar eclipses of the inscriptions are documented in the Palenque Oval Palace Tablet, the Copán Hieroglyphic Stairway, and the Quiriguá Stela C. The lunar eclipses of the inscriptions are tied to the lunar half-year, the lunar fortnight, and the lunar season, and they are evidence of the precision of the Maya lunar series.

The 405-month cycle in the inscriptions

The 405-month cycle is rarely explicitly mentioned in the inscriptions, but it is implied by the lunar series and by the long-count positions. The cycle is, in essence, a “background” of the Maya astronomical tradition, and the daykeeper reading the lunar series and the eclipse table would be aware of the 405-month cycle as the basis of the eclipse prediction system.

The 405-month cycle is also tied to the 819-day cycle of Caracol and the Lords of the Night 9-day cycle. The 405-month cycle returns to the same 9-day Lord position only after 405 × 29.53 ÷ 9 = 1,328.83 days, or roughly 4 × 332.2 days, and the 819-day cycle returns to the same 9-day Lord position only after 819 days. The cycles are mathematically independent, but they are both used in the inscriptions as ways of structuring longer time periods.

The GMT correlation and the lunar series

The GMT correlation of 584,283 (the most widely used correlation between the Long Count and the proleptic Julian calendar) is supported by the lunar tables in the Dresden Codex. The lunar tables allow the absolute date of a Long Count position to be reconstructed to within a day, and the reconstruction is consistent with the GMT correlation. The Spinden correlation of 489,384, by contrast, is not consistent with the lunar tables; the Spinden correlation places the base date 260 days later than the GMT correlation, and the lunar tables are off by approximately one lunar month under the Spinden correlation.

The GMT correlation is the modern academic standard. The modified GMT of 584,285, sometimes used in astronomical work, is 2 days later than the standard GMT, and the choice between the standard GMT and the modified GMT is a matter of convention. The lunar tables are consistent with the GMT correlation, and the GMT correlation is the basis of the modern chronology of the Classic Maya.

The daykeeper’s role in eclipse prediction

The ah k’in (or aj q’ij in modern K’iche’) was the daykeeper, the ritual specialist who maintained the almanacs and the astronomical tables. The ah k’in was responsible for:

Observing the heliacal rising of the planets and the bright stars.
Calibrating the eclipse tables against the actual eclipses.
Predicting the eclipses for the coming months and years.
Selecting the day for royal rituals, planting, and naming ceremonies.
Reading the almanacs and the prognostications of the 260-day cycle.

The ah k’in in the Classic period was a hereditary office, with a parallel office of ch’ok ah k’in (apprentice daykeeper) learning the almanacs in training. The ah k_in’s role in eclipse prediction is documented in the Dresden Codex, in the inscriptions of Palenque and Copán, and in the modern highland K’iche’ and Kaqchikel communities.

The Maya solar eclipse risk in 2024 and 2026

The partial solar eclipse of 8 April 2024 was visible from the Maya area, and the daykeepers of the highland communities observed the eclipse with ritual responses. The total solar eclipse of 14 October 2023 (visible from the Yucatan Peninsula) was observed in the same way, and the partial solar eclipse of 29 March 2025 (visible from the Yucatan Peninsula) is a future event. The daykeepers’ responses to modern eclipses are a survival of the Classic-period tradition, and the highland aj q’ij continue to use the Dresden Codex eclipse table as a reference for the modern eclipse predictions.

The modern eclipse responses are documented in the work of scholars such as Alessandro Fornazzari, Ariel de Vidas, and Enrique Ajú Patal, who have worked with the highland aj q’ij on the eclipse tradition. The modern responses include fasting, abstention, ritual bloodletting (in some communities), and the recitation of almanac texts.

The role of the Venus cycle in eclipse prediction

The Venus cycle and the eclipse cycle are mathematically independent, but they are sometimes tied in the inscriptions. The 584-day Venus cycle, the 780-day Mars cycle, and the 405-month eclipse cycle are all mentioned in the Dresden Codex, and the daykeeper could use the tables together to determine the relative positions of the planets, the moon, and the eclipse half-years.

The Venus cycle and the eclipse cycle are also tied to the Tzolkin and the Haab’. The 584-day Venus cycle, the 365-day Haab’ year, and the 260-day Tzolkin cycle return to the same combined position every 11,960 days (the 405-month cycle), and the daykeeper could use this coincidence to predict the alignment of the Venus cycle with the eclipse cycle.

The mathematical structure of the Maya astronomical tables is, in this respect, a sophisticated system that ties together the planetary cycles, the lunar cycle, the eclipse cycle, and the almanac. The system is the basis of the modern chronology of the Classic Maya and of the modern academic understanding of Maya astronomy.

Sources

Aveni, A. F. (2001). Skywatchers of Ancient Mexico. University of Texas Press. — link
Aveni, A. F. & Hartung, H. (1986). Maya city planning and the calendar. Transactions of the American Philosophical Society 76 (1). — link
Edmonson, M. S. (1988). The Book of the Year: Middle American Calendar Systems. University of Utah Press. — link
Teeple, J. E. (1926). Maya Astronomy. Contributions to American Archaeology No. 2. Carnegie Institution of Washington. — link
Rice, P. M. (2007). Maya Calendar Origins: Monuments, Mythohistories, and the Materialization of Time. University of Texas Press. — link
Coe, M. D. (1993). The Maya (5th ed.). Thames & Hudson. — link