The History of Crescent Observation: From Babylonian Tablets to Satellite Data

The Short Answer

People have watched for the crescent to time their months for about 4,000 years. Babylonian scribes logged the interval from sunset to moonset on the evening of first visibility (NA1) for some 600 years; Islamic astronomers from al-Khwārizmī to al-Bīrūnī built visibility tables and the arithmetic Hijri calendar; André Danjon set the limit near 7 degrees of elongation in 1936; and Yallop (1997) and Odeh (2004) turned visibility into a single computed score.

The act of watching for the first sliver of the crescent is among the oldest continuous scientific observations in human history. Long before telescopes, computational engines, or satellite weather data, civilisations built their calendars, planned their harvests, and timed their religious observances around the reappearance of the crescent moon after conjunction.

This is the story of how humanity has watched for that sliver of light, and how each era's technology transformed the practice.

The Babylonians: Where It All Began (c. 2000 to 500 BCE)

The earliest known systematic records of crescent observations come from ancient Mesopotamia. The Babylonians, based in present-day Iraq, developed the world's first mathematical astronomy, and the crescent moon was central to their calendar and religious life.

The MUL.APIN Tablets

Among the most important Babylonian astronomical texts is the MUL.APIN series, dating to approximately 1000 BCE (though encoding traditions considerably older). These cuneiform tablets contain, among other things, schematic rules concerned with the first and last visibility of the Moon, framed in terms of time intervals rather than angles.

For visibility timing the Babylonians measured intervals in time-degrees rather than in altitude angles, using the UŠ (roughly 4 minutes of time, or about 1 degree of sky rotation). Their records show they understood that the crescent becomes visible only once a sufficient interval has opened up between sunset and moonset, an early, qualitative version of what we now call the Danjon limit and ARCV constraints.

The Astronomical Diaries

From roughly the 7th century BCE to about 60 BCE (the latest securely datable entries fall around 60 BCE), an astonishing span of roughly 600 years, Babylonian astronomers kept continuous records known as the Astronomical Diaries. These tablets recorded, night after night, the positions of the Moon, planets, and stars, along with weather conditions, river levels, and commodity prices. The standard modern edition is the multi-volume Astronomical Diaries and Related Texts from Babylonia edited by Abraham Sachs and Hermann Hunger.

The Lunar Six and NA1

The diaries did not merely note that the crescent had appeared; they timed it. Babylonian observers recorded a set of six standard time intervals around new and full moon known today as the Lunar Six. The most important for crescent science is NA1 (also written na), the interval between sunset and moonset on the evening the new crescent first becomes visible.

NA1 is, in effect, an ancient direct measurement of the lag time, the gap between sunset and moonset that modern models treat as a first-order proxy for ARCV (the arc of vision, the altitude difference between Moon and Sun). A long NA1 means the Moon lingers well after the Sun has set, the sky has darkened, and the crescent stands clear of the horizon haze; a short NA1 means the crescent chases the Sun down and is likely lost. Two and a half millennia before Yallop expressed the same idea as a number, Babylonian scribes were already logging the single most predictive quantity in the field, and doing so in minutes of time rather than in degrees.

System A and System B

Alongside the observations, Babylonian astronomers developed two distinct arithmetical schemes for predicting lunar and solar positions, conventionally labelled System A and System B. System A modelled the Sun's varying speed along the ecliptic as a step function, jumping between two constant velocities, while System B used a linear "zigzag" function that increased and decreased by a fixed amount each month. Both schemes let astronomers compute the Lunar Six intervals in advance, so that a predicted NA1 could be checked against the observed one. This pairing of arithmetic prediction with systematic observation is the deep structural ancestor of every modern visibility engine.

These records together represent the oldest empirical dataset on crescent visibility in the world, and 20th-century researchers returned to them repeatedly. Tom Bruin used the Babylonian and later historical data when constructing his graphical visibility criterion in the 1970s; Bradley Schaefer drew on long historical observation series in his photometric modelling of the visibility limit; and Louay Fatoohi, with F. Richard Stephenson, mined the Babylonian first-visibility records directly to re-examine the Danjon limit and the reliability of ancient sightings.

Why It Mattered

For the Babylonians, the first sighting of the crescent determined the start of a new month, much as it does in the Islamic calendar today. When the crescent was not sighted on the expected evening (due to clouds or a very young moon), the current month was extended to 30 days. This system is directly ancestral to the Islamic practice of completing 30 days when the crescent is not observed.

Greek and Roman Contributions (500 BCE, 500 CE)

The ancient Greeks inherited much of Babylonian astronomical knowledge and reframed it within a geometric and philosophical framework.

Hipparchus and Ptolemy

Hipparchus (c. 190 to 120 BCE) compiled the first comprehensive star catalogue and developed methods for predicting the Moon's position using epicyclic models. His work was the foundation for Ptolemy's Almagest (c. 150 CE), which remained the dominant astronomical reference in the Mediterranean, Middle Eastern, and European worlds for over 1,300 years.

Ptolemy's lunar model, while geometrically sophisticated, fixed the time of conjunction well, but neither it nor any ancient theory modelled the brightness and contrast limits that decide whether a thin crescent can be seen. It could not determine the ARCV or crescent width with the precision needed for visibility forecasting, the parameters at the heart of modern criteria.

The Contribution of Optics

The Greeks were also the first to systematically study optics and vision. Euclid (c. 300 BCE) and later Ptolemy wrote treatises on how the eye perceives light. Their work on the limits of visual detection laid groundwork that would be developed a millennium later by Islamic scholars.

The Indian Bridge (400 to 800 CE)

The path from Greek astronomy to Islamic astronomy did not run in a straight line; it passed through India. Greek geometric models reached the Indian subcontinent in the early centuries CE and were absorbed, transformed, and combined with an indigenous Indian tradition of mathematical astronomy that was, in some respects, more advanced in its computational methods.

The pivotal text is the Surya Siddhanta, a Sanskrit treatise on planetary motion whose surviving form dates to roughly the 4th to 5th century CE. It set out trigonometric methods (including an early sine table), procedures for computing the positions of the Sun and Moon, and rules for the moments of conjunction and opposition, exactly the calculations on which crescent prediction depends. Indian astronomers also worked comfortably with the tithi, a lunar day defined by elongation, which made the Moon-Sun separation a natural quantity in their system.

This Indian material reached the early Islamic world in the 8th century, when astronomical works were translated into Arabic at the court of Baghdad. The body of Indian astronomical knowledge that resulted was known to Arabic writers as the Sindhind (from the Sanskrit siddhanta). It was precisely this tradition that al-Khwārizmī drew on a generation later, which is why his great tables carry the name Zīj al-Sindhind. The conceptual chain, Babylonian intervals to Greek geometry to Indian trigonometry to Arabic zīj, is one of the clearest examples of knowledge transmission across civilisations in the history of science.

The Golden Age of Islamic Astronomy (750 to 1400 CE)

The Islamic contribution to crescent observation science is immense, arguably the most productive period in the history of the field. This is not coincidental: the Islamic calendar's dependence on crescent sighting created a powerful, sustained incentive for precise lunar astronomy.

Al-Khwārizmī and the First Visibility Tables

Muhammad ibn Mūsā al-Khwārizmī (c. 780 to 850 CE), the mathematician whose name gave us the word "algorithm", compiled astronomical tables (Zīj al-Sindhind) that included methods for determining whether the crescent would be visible on a given evening. As the title records, his tables rested on the Indian Sindhind parameters described above, combined with Ptolemaic refinements, and considered the Moon's elongation, altitude, and latitude.

Al-Khwārizmī's approach was essentially what we would today call a lookup table, a pre-computed grid that an astronomer could consult by looking up the relevant parameters for a given date and location.

Al-Battānī's Refined Parameters

Al-Battānī (c. 858 to 929 CE), working in Raqqa (present-day Syria), made the most precise astronomical observations of his era. He corrected several of Ptolemy's parameters, including the obliquity of the ecliptic and the precession of the equinoxes. His improved lunar model significantly enhanced the accuracy of conjunction predictions, the necessary first step in any crescent visibility forecast.

Al-Battānī also explicitly discussed the problem of atmospheric refraction near the horizon, noting that celestial objects appear higher than their true geometric position when close to the horizon, an observation that would not be quantitatively modelled for nearly a millennium.

Ibn al-Haytham and the Science of Vision

Ibn al-Haytham (Alhazen, c. 965 to 1040 CE), widely regarded as the father of modern optics, wrote the Kitāb al-Manāẓir (Book of Optics), which fundamentally transformed the understanding of how the eye perceives light. His work established that vision occurs when light enters the eye (not when "visual rays" emanate from it, as the Greeks believed) and developed the theory of image formation by the lens of the eye.

While Ibn al-Haytham did not write specifically about crescent visibility, his framework for understanding the limits of visual perception, minimum detectable contrast, the role of background brightness, and the angular resolution of the eye, is exactly the physics that underlies the Danjon Limit and the visibility zone classifications used today.

Al-Bīrūnī and the Tabular Calendar

Al-Bīrūnī (973 to 1048 CE), the polymath from Khwarezm, made critical contributions to calendar science. His al-Qānūn al-Masʿūdī (Canon of Masʿūd) included a comprehensive treatment of intercalation schemes for the Islamic calendar, the mathematical rules for alternating 29- and 30-day months to keep the calendar in step with the synodic month.

The "Kuwaiti algorithm", the tabular calendar still implemented in Microsoft Windows and countless digital systems, is a direct descendant of al-Bīrūnī's arithmetic calendar. It distributes 11 leap years across each 30-year cycle, adding a leap day in years 2, 5, 7, 10, 13, 16, 18, 21, 24, 26 and 29. (The well-known tabular variant used by the Microsoft and Kuwaiti implementations places the sixth leap year at year 16, rather than the year 15 found in some other tabular schemes.) The arithmetic calendar has been in continuous use, in one variant or another, for nearly 1,000 years. You can see this same tabular method running live, alongside two astronomical engines, in our triple-engine Hijri calendar at /calendar.

Ulugh Beg's Observatory

Ulugh Beg (1394 to 1449 CE), the Timurid sultan and astronomer who built the magnificent observatory at Samarkand, compiled star and planet tables of unprecedented accuracy. His Zīj-i Sulṭānī included refined lunar parameters that improved conjunction predictions.

Ulugh Beg's observatory, with its massive 40-metre sextant, represented the pinnacle of pre-telescopic astronomical precision. His work on the length of the sidereal year was not improved upon until Tycho Brahe's observations more than a century later.

The Telescope Revolution (1609 to 1900)

The invention of the telescope in 1608 and its astronomical application by Galileo Galilei in 1609 transformed every area of astronomy, including crescent observation.

Earliest Telescopic Crescent Sightings

With even a small telescope, observers could detect crescents invisible to the naked eye. Systematic telescopic observations in the 17th and 18th centuries gradually built the empirical dataset that would later allow Danjon and Yallop to construct quantitative visibility models.

André Danjon and the Visibility Limit (1932 and 1936)

By the early 20th century, enough data had accumulated to support a systematic analysis. André Danjon's 1932 paper ('Jeunes et vieilles lunes') first reported the phenomenon of cusp shortening near small elongations. His 1936 paper ('Le Croissant Lunaire') then quantified the lower bound at approximately 7° of elongation, based on 75 crescent-length measurements from across Europe. This was a watershed moment: for the first time, crescent visibility was framed not as a subjective experience but as a physical constraint subject to mathematical analysis.

The Modern Era: Computation and Crowdsourcing (1980, Present)

Yallop (1997) and the q-Value Revolution

Dr Bernard Yallop's 1997 technical note at the HM Nautical Almanac Office introduced the Yallop q-value, a single number that encodes the likelihood of crescent visibility. Calibrated against historical observations, the Yallop criterion gave astronomers and Islamic calendar committees a rigorous, reproducible framework for evaluating whether the crescent could be seen.

For the first time, a sighting committee could compute a q-value for their location and compare it against a standardised zone classification. The subjectivity of "the committee chairman thinks the crescent might be visible" was replaced with "the q-value is +0.14, placing this sighting in Zone B." Crucially, the criterion finally retired the persistent age myth: the bare number of hours since conjunction is a poor predictor of visibility and appears nowhere in Yallop's formula. What decides the outcome is the geometry, ARCV, ARCL (the arc of light, or Moon-Sun elongation), the topocentric crescent width W, and DAZ (the difference in azimuth). Yallop's q is built from ARCV and W, evaluated at his "best time", about four ninths of the lag time after sunset:

q = (ARCV - (11.8371 - 6.3226·W + 0.7319·W² - 0.1018·W³)) / 10

The six visibility zones it defines are mutually exclusive bands, not a cumulative ladder:

You can read a fuller derivation of these criteria in our explainer on the science behind the crescent.

Odeh (2004) and the ICOP Dataset

Mohammad Shawkat Odeh's 2004 refinement drew on 737 observation records, about half contributed through ICOP and the remainder from earlier published datasets, including a substantial proportion of telescopic and CCD sightings. By weighting optical-aid observations explicitly, the Odeh V-value filled a gap in Yallop's predominantly naked-eye model and produced sharper thresholds at the binocular boundary (Odeh, 2004).

The ICOP Network

The Islamic Crescents' Observation Project (ICOP), founded in 1998 by Mohammad Odeh under the International Astronomical Center, created the first global, standardised network for collecting crescent sighting reports. The ICOP archive, accessible within the platform at /archive, holds 1,028 ICOP reports for Ramadan and Shawwal from 1440 to 1446 AH, and ICOP data remain the backbone of modern visibility model validation (Odeh, 2004). The archive is available in extended form to Pro subscribers, who also gain access to cloud-cover overlays that complement the historical sighting records.

Digital Astronomy Engines and Crowdsourced Networks

The development of high-accuracy computational libraries, such as Don Cross's astronomy-engine (implementing VSOP87 for the Sun and ELP2000 for the Moon), made sub-arcsecond positional calculations accessible to any programmer. This democratised crescent visibility computation, allowing web applications and mobile apps to compute Yallop q-values and Odeh V-values for any point on Earth in milliseconds.

The smartphone era added a further dimension: real-time, global crowdsourcing. On the 29th night of each lunar month, thousands of observers worldwide submit geolocated sighting reports, which are plotted against predicted visibility zones in real time, providing a live validation of the models. CCD cameras have simultaneously pushed crescent detection to elongations below the traditional Danjon limit, refining the graduated nature of that boundary and deepening understanding of cusp fragmentation.

Conclusion

The history of crescent observation is a 4,000-year story of humanity trying to answer a deceptively simple question: "Can I see the Moon tonight?"

From Babylonian scribes pressing wedge-shaped symbols into wet clay, through medieval Islamic astronomers computing conjunction tables by candlelight, to modern platforms computing visibility grids in sub-second response times, each era brought new tools to the same fundamental problem. What makes this history unique is that it has never been purely academic: the crescent determines when 1.8 billion people fast, feast, and worship, and that practical urgency has driven continuous innovation for four millennia.

The next chapter is being written by every observer who steps outside on the 29th night and logs what they see. Compute the Yallop q-value and Odeh V-value for your own location on the global visibility map at moonsighting.live, and explore the record for your region in the ICOP archive.

References and Further Reading

• Sachs, A. and Hunger, H. (1988 onwards). Astronomical Diaries and Related Texts from Babylonia. Verlag der Österreichischen Akademie der Wissenschaften, Vienna.
• Yallop, B.D. (1997). "A Method for Predicting the First Sighting of the New Crescent Moon." HM Nautical Almanac Office, NAO Technical Note No. 69.
• Odeh, M.Sh. (2004). "New Criterion for Lunar Crescent Visibility." Experimental Astronomy, 18, 39-64. (Data and tools at https://www.astronomycenter.net.)
• Danjon, A. (1932). "Jeunes et vieilles lunes." L'Astronomie.
• Danjon, A. (1936). "Le Croissant Lunaire." L'Astronomie.
• Fatoohi, L., Stephenson, F.R. and Al-Dargazelli, S.S. (1998). "The Danjon limit of first visibility of the lunar crescent." The Observatory, 118, 65-72.
• Kennedy, E.S. (1956). "A Survey of Islamic Astronomical Tables." Transactions of the American Philosophical Society, 46(2).

Clear skies and happy sighting.