Posted on Monday 9 October 2006

The Islamic calendar is the only one in use in the world today that can be truly described as a lunar calendar. That is, every month is synchronized to the cycle of the moon and not the solar year.

Others that purport to be lunar-based such as the Hebrew and Chinese ones, while certainly being based on lunar months also contain mechanisms designed to synchronize them with the solar year. Hence they are more properly described as lunisolar calenders.

The Islamic calender on the other hand is never synchronized with the solar year because from its inception, special measures were taken to actually prevent this correction process from taking place. The intention behind this was to remove the need for expert knowledge when it came to determining the times and dates of religious observance, something that in the past had always been associated with the work of priests. This was a conscious decision by the founders of Islam as a way to prevent the rise of a priestly class or Church establishment within Islam, something to set it apart from other religions¹.

In the 9th year after the Hejira (Muhammad's migration to Medina in 622 and the starting date of the Islamic calendar), the Prophet forbade the insertion of leap months which used to occur in the traditional Arabic calender:

The postponing (of a Sacred Month) is indeed an addition to disbelief: thereby the disbelievers are led astray, for they make it lawful one year and forbid it another year in order to adjust the number of months forbidden by Allah, and make such forbidden ones lawful. The evil of their deeds seems pleasing to them. And Allah guides not the people, who disbelieve. This Ayah prohibits such practice.

(Al-Taubah 9, Ayah 37)

Solar calendars, it could be argued, are by the nature of their complexity, the product of either wealthy church establishments or of powerful states and their maintenance has over the centuries taxed the efforts of a great many priests, astronomers and mathematicians. In a way they oblige obedience in worshipers to some authority (i.e. governemtn or church) which could therefore be seen as mediating their relationship with their deity.

The movement of the moon, on the other hand (and by the same argument), has always been owned by the common people who have never needed anyone to tell them what phase the moon was in at any moment. This is seen by the adherents of Islam as a demonstration of its greatest strengths: its simplicity, universality and its fundamentally anti-elitist character.

Ibn 'Umar (Radiya-Allahu 'anhuma) reported that the Messenger said: "We are an illiterate nation. We do not use astronomical writing or computation [in our fasting]. A month is so and so and so (and he pointed with his hands three times, folding the thumb on the third time, meaning twenty nine days) or so and so and so (and he pointed with his hands three times, meaning thirty days).
(Bukhari, Muslim, Abu Dawud and An-Nassa'i)

The Islamic year, as with other calendar systems, is the completion of 12 months. A month is based on the cycle of the moon and can be no shorter than 29 days and no longer than 30. This means that the Islamic year is about 11 days shorter than the solar year and, as a consequence, the lunar months drift backwards through the solar year without paying any heed whatsoever to the season. Before the Islamic calender came into effect, the ancient Arabic month of Ramadan had originally referred to the hottest month of year (from ramda meaning "hot stones"). Now it can fall at any time in the solar year and in any season. It takes lunar 33 years before the Islamic calendar realigns itself with the solar year once again.

While this slippage is predictable and can be calculated many years in advance, another feature of the Islamic calendar is far less deterministic: the exact day on which a month begins. Technically, an Islamic month commences at the time of the first sighting (by two Muslim witnesses) of the crescent moon (the Hilal) which occurs after a new moon. This becomes especially important at the start of the holy month of Ramadan because it marks the beginning of a month of fasting and of other rituals.

“Whoever witnesses the crescent of the month, he must fast the month.”
(Qur'an, 2:185)

“If you see the Hilal, then fast, and if you see it (at the end of Ramadan), then break your fast, and if the sky is overcast, then fast thirty days.” (Muslim/Bukhari)

“A month is either 29 or 30 days long. So if you see the Hilal, then fast, and if you see it again (at the end of Ramadan) then break your fast. And when the sky is overcast, then complete the count (of thirty days).”
(Nasai)

Based on such a simple criteria, you'd think that there would be no difficulty in ensuring automatic unanimity amongst observers of the start of a lunar month. In practice, however, it turns out to be an extraordinarily difficult astronomical problem because unlike other celestial events this one must also take into account the question of human visibility. It's not enough to simply refer to calculations in order to say that something or other has happened, it must also have been observable by someone.

A new moon is the conjunction of the earth, moon and sun, this means that their centres all lie in exactly the same line. Under these conditions the moon becomes invisible when viewed from the earth because the side of the moon that is being illuminated by sunlight does not face the earth. As the moon moves in its orbit away from conjunction, some of sun's light hitting the moon starts to be reflected toward the earth illuminating an edge of the moon which then appears as a crescent. As the month wears on the proportion of moon visibly illuminated increases until the 14th day, otherwise known as the full moon. From then until the end of the month the moon is on the wane with the crescent receding to the other edge of the moon until it finally disappears altogether. This is yet another new moon.

This is a New Moon. It just so happens to be in front of the Sun. Not all new moons pass in front of the sun. Usually the moon passes above or below the sun when it is new or "zero days old". This is an eclipse photo from June 10, 2002 8:49 pm EDT through a 6" f/8 telescope. [Moon Photo Page]

The problem with first spotting the Hilal is that at the time of the conjunction, the moon is located at a point in the sky that is very close to the sun, usually it is just above or below it and very occasionally directly in front of it, as in the case of an eclipse. As it moves out of conjunction, the tiny amount of light coming from the sliver of the new crescent is completely swamped by the glare of daylight and making it invisible to observers on the earth. The earliest time that the crescent can be observed is after sunset and even then only once the twilight has decreased to a point where the crescent light is brighter than it. By the time the moon is finally visible, it is already likely to more than twelve hours "old", that is it would have been at least twelve hours since it left conjunction with the earth and sun. The world record for observing the youngest crescent is 15.4 hours with naked eye, 12.7 hours with binoculars, and 12.2 hours with a telescope.

This is a one day old moon. More precisely, it's 1 day 15 hours and 35 minutes. That's how much time has passed from the moment of New Moon till this picture was taken. One day moons can be difficult to spot. They are very close to the horizon and visible for just a few minutes after sunset. This moon was 39 hours old. This picture was taken with a Nikon Coolpix 885 digital camera held up to the eyepiece of a Celestron C-90 Astro telescope on July 11th 2002 at 10:03 pm EDT in Holt, Michigan. [Moon Photo Page]

Because the moon is still very close to the sun, there's hardly any time to observe the Hilal before it promptly sets as well. Under even the very best conditions the crescent might be visible for up to an hour after sunset but usually this time is much shorter, often only a few minutes. If the moon is younger than 12 hours it is unlikely to be visible at all.

Atmospheric conditions and weather patterns can have a major impact on the observation, of course, but there are many other factors to consider as well: the latitude and longitude of the observer, the position of moon in its orbit (because of moon's orbit is tilted with respect the plane of the earth's orbit around the sun making crescent more or less obvious) and even the lunar mountains can have an effect by casting shadows onto the tips of the crescent! The longitude of the observer matters in the a relative sense that if observers to the East have failed to sight the Hilal, there is still a chance that observers further to the West will manage it. That's because the moon will have had more time to move out of conjunction before sun sets in that part of the world.

Latitude is also important because at the time of the vernal equinox (March in the Northern Hemisphere, September in the Southern) the sun's path is steeply perpendicular to the horizon, i.e. the sun comes straight down when it sets rather than moving down diagonally. This steepness of descent means that the crescent moon will be higher in the sky just after sunset, thereby improving its visibility.

In cases where there has been a failure to sight the Hilal then the old month (Sha'ban in the case of Ramadan) is not considered to have ended yet and it automatically runs to 30 days, potentially creating disagreement amongst observers. The only thing that is certain is that the previous month is never shorter than 29 days and that the Hilal cannot ever be observed before the new moon.

Who claims that he viewed the thin waxing crescent before the New Moon instant, is like the one who claims that he saw the Sun before dawn, or he saw the baby before delivery. (proverb)

This is the back door in which calculation and computation reenter the picture. Web sites like Moonsighting.com run a service not unlike a weather bureau in which they forecast Hilal visibility information based on calculations made by Dr. Monzur Ahmed's MoonCalc program.

Here is the forecast that was issued for the Ramadan of 2003

Ramadan: The moon's conjunction is on Saturday October 25, 2003 at 12:50 UT i.e., 8:50 am Eastern Daylight Time - 5:50 am Pacific Daylight Time. On Oct 25, the moon is going to be less than 13 hours old on West coast of USA, and impossible to see. On Oct 26, the moon can be seen in most of the world, except Japan, Northern Asia, and most of Europe. In England, it may not be seen even on October 26, although it is going to be 28 hours old, because it will be only 3 degrees above horizon at sunset and the glow of sunlight will make it very difficult to see.

The first day of Ramadan will be on October 27, 2003 for Australia, Africa and Americas, and on October 28, 2003 for Japan, Northern Asia, and most of Europe. England is in a situation that moon may or may not be seen on October 26.

This was then followed up by an announcement made about the starting date by the Islamic Shura Council of North America

The Islamic Shura Council of North America and Fiqh Council of North America, of which ISNA is a member, agreed that there were no confirmed moon-sighting reports on the evening of Saturday, October 25, 2003, the 29th of Sha'ban. We will complete 30 days of Sha'ban. Therefore, Monday, October 27 will be the first day of Ramadan in North America. Ramadan Karim

To get a sense of the difficulty of spotting the new crescent moon, you could try your hand at simulating it using Helmer Aslaksen's very nice (and very simple to use) Java applet entitled "What does the waxing or waning Moon look like in different parts of the world in the course of the year?".

First set the month to the one you want to observe (let's say October) and then set the day of the month to 1. Choose your location in the world from the map and then turn the speed down to really slow (the default speed is way too fast). You will probably also want to "freeze" the day so that you can watch it in a loop over and over again because the Hilal happens so quickly.

The program makes some fairly drastic simplifications, especially that the moon and sun orbits are on the same plane but there's still plenty of complexity here to contend with even with such a simple model.

Note 1 : Calendars and Priesthoods - It is no coincidence that the Catholic Church led the way with the reform of the Julian calendar.

The institution of Gregorian calendar in 1582 was made necessary in order to stop seasonal drift in the date of Easter. This date was originally agreed by the First Council at Nicaea to be calculated as falling on the Sunday following the first full moon that follows the Northern Spring equinox. The problem was that with the Julian calendar, the equinox was drifting forward from year to year, even by the time of the First Council in 325 (only 371 years after the Julian calendar had been established) the equinox drifted 4 days from its traditional date of the 21st of March.

By 1582, this drift had grown to ten days which was an intolerable situation and had to be remedied and despite religious schism, even protestant Europe agreed (very gradually) accepting the reforms although without accepting Catholic religious primacy.