Moon Sighting

All of the Moon Sighting News and Resources You Need

Home Introduction Calendars Slideshow Eid al-Adha Moonsighting FAQ Hilal Pictures Articles Why Confusion Updates Volunteers Links Archive

Moon Basics






Get all of your town's daily sun & moon information:


Eid al-Adha
Moonsighting FAQ
Hilal Pictures
Why Confusion



Deceptive Hilal Images

Eid Mubarak Shawwal Moon

Islamic Calendar- Some Basic Issues:

Omar Afzal
Cornell University USA

[Abstract: Every year, for decades the Muslims all over the world start Ramadan on 3-4 different solar dates. Similarly, they celebrate Eidain on 3 or 4 different days. Several conferences and meetings (the most recent one in Jeddah, Nov. 1998) have done little to ease the chaotic situation. The confusion will continue until the Muslim calendar experts and Ulema agree on the basic principles for computing and implementing a global Islamic calendar. The most sticky points are the ever-shifting international lunar datelines that result in the same lunar month being both 29 and 30 days long, the lunar visibility that extends in a parabola rather than on a north/south longitude, the meridian, and the length of the year. Some of these suggestions contained here were earlier presented in a 1986 IIIT Lunar Calendar Conference (1987) and Majallah Taqiqat-I-Islami:1988)].

1.0 Issues

Let us look at the visibility maps of Ramadan and Eid al-Fitr 1423:

Ramadhan 2002 / 1423 AH Visibility of crescent on evening of 5th Nov 2002:


Map produced by MoonCalc 6.0 by Dr. Monzur Ahmed

The colored areas show areas of the world where the crescent is likely to be visible of the evening of 5th Nov 2002.

Key to shading of the 4 parabolas:

- Parabola A: crescent visible with naked eye
- Parabola B: visible with naked eye if conditions are perfect
- Parabola C: may need optical aid to find crescent
- Parabola D: visible only through telescope

The new crescent moon should be visible by naked eye from parts of North America, South America and Africa on 5th November 2002. The crescent may be visible with optical aid from parts of the Middle East on 5th Nov 2002. In locations further east, the crescent should be visible on 6th Nov 2002.

Hence, pending actual visibility of the crescent moon, Ramadhan should start on Wednesday 6th Nov 2002 (USA, S. America, Africa, ?Middle East) or Thursday 7th Nov 2002 (Europe, rest of world).


Eid ul Fitr 2002/ 1423 AH

Visibility of crescent on evening of 5thDec 2002 (left map) and 4th Dec 2002 (right map)

Map produced by MoonCalc 6.0 by Dr. Monzur Ahmed

The colored areas show areas of the world where the crescent is likely to be visible of the evenings of 5th Dec 2002 (left map) and 4th Dec 2002 (right map).

For Ramadan please note:

a) The crescent moon is seen in a parabola. Ramadan 1423 parabola (telescopic visibility) on Nov. 5 extends west/north-south westward from Indonesia whereas naked-eye visibility starts from western Africa.

b) The ‘visibility’ takes at least two days (Nov. 5 and 6) to cover most of the world. The visibility cuts through political boundaries (half of Indonesia, half of India, etc. for telescopic, and USA, etc. for naked-eye sighting).

For Eid al-Fitr map please note the following:

a) The Shawwal crescent (telescopic) visibility begins on Dec. 4 from western regions of South America, and naked-eye sighting in southern Pacific ocean.

Now let us test the criterions used by the Muslims to determine Ramadan and Eidain.

a) Ru’yah (Hilal)

Based on the “visibility” the Muslims begin fasting on Nov. 6 and 7 (Two solar dates). They celebrate Eid on Dec. 6 and 7 (Japan, Northern China, Russia, northern Canada, Alaska, etc.)

Some Muslims (living from western Indonesia, if telescopic visibility is the norm, or from western Africa, if naked eye sighting is the norm) will fast 30 days, whereas the rest will fast only 29 days (as the lunar dateline moved westwards from Ramadan to Shawwal). The problem will not go away even if we straighten the parabola at the eastern-most bulge to the longitude.

At least we are starting the lunar day/date and the month from the sunset, as is the practice.

b) Shahadah (Eye-witness)

Those who want to decide the beginning of the month a “Shahadah” must know that for Ramadan the Shahadah came from Nigeria on Nov. 4. Libya also started Ramadan on Nov. 4.

If we have to decide by the Shahadah/Khabar then Muslims in Nigeria claimed seeing Shawwal crescent on Dec. 3 to have Eid on Dec. 4.

Please remember that there is a simple test for accepting a “Shahdah”. If the Muslims located west of the place from where the Shahadah was received did not see a crescent moon the same evening then the claim has to be rejected. Often, Muslims are told that Saudi witnesses have SEEN the crescent moon (as they did on Dec. 4 for Shawwal this year) and no crescent moon is seen from Saudi Arabia to Hawaii 10-12 hours later.

Shahadah has to be tested for authenticity. The A’rabi Hadith is no excuse to accept bogus claims.

c) Khabar (News of sighting)

Some insist that the news of moonsighting from anywhere (mainly from Saudi Arabia/Egypt, Yemen, etc.) are enough to start the lunar months.

Khabar like Shadah can easily be tested for accuracy. The news may be correct but the those who accepted the Shahadah ignored the ‘authenticity’ criterion if the sighting was not confirmed from places west of them.

d) Conjunction (New Moon)

If the “First sunset after conjunction” is used as a criterion, then Ramadan had to start on Nov. 5 from eastern US (excluding St. John, Canada). From New Zealand to St. John, Canada Ramadan was to start on Nov. 6 (Two solar dates) For Eid, Muslims from the longitude of Brunei to Hawaii would pray on Dec. 5, and the rest on Dec. 6.

As the conjunction (New Moon) occurs at all times of day and night, the modification of the “first sunset after Conjunction” will result in extending or reducing the days of the month arbitrarily.

e) Moon set after the sunset (Egyptian/Saudi criterion)

If we use the Saudi/Egyptian criterion of the “Moonset after the sunset” then Ramadan starts from St. Johns, Canada westward on Nov. 5, and from New Zealand to Europe and Africa on Nov. 6. (Also keep in mind that the old “lunar month” is not complete in St. Johns, Canada, but the moon is setting eight minutes AFTER the sunset.)

Eid al-Fitr date from Brunei to Hawaii was Dec. 5 and eastward of Brunei to New Zealand on Dec. 6 (Two solar dates, and the problem of moonset after the sunset, and yet the old lunar month not completed).

f) Moon’s Altitude/Angle of Separation from the Sun

Any criterion based on the altitude and angle will have the same problems for the calendar.

For example, if we calculate the moon’s altitude of 5 degrees at the sunset (Turkish), Ramadan was to start on Nov. 6 as the moon’s altitude was less than 5 degrees on Nov. 4 all over the world.

Eid al-Fitr in Americas was on Dec. 5, and in Australia, Asia, Africa and Europe on Dec. 6, as the moon’s attitude was above 5 degrees only in north and south America on Dec. 4.

g) Moon’s Age

Moon’s age and time lag is also often presented as a criterion for starting an Islamic month.

For example, if we take 15 hours as the age when the new lunar month should begin, then Ramadan was to start on Nov. 6 from middle part of India, and from New Zealand to the rest of India on Nov. 7. Similarly, Eid was on Dec. 5 from Texas to Hawaii and on Dec. 6 from New Zealand to Florida.

It is obvious that none of these criterions solve the problem because:

1. Most of them do not meet the ‘Hilal’ criterion of the Qur’an and the Sunnah.
2. There is no fixed lunar dateline.

An ever-changing dateline makes every month 29 and 30 days long

2.1 Criterion for a global Islamic Calendar

Islamic calendars in use by various Muslim countries and communities are imperfect, cumbersome, and extremely inefficient. They do not adhere to any one criterion, including the ‘lunar visibility’ - often claimed to be the basis for determining important Islamic religious dates. Lunar visibility from the surface of the earth is difficult to calculate. Besides, a strictly visibility-based global lunar calendar will cut through national and regional boundaries and show each month simultaneously of 29 and 30 days long in different regions. What the Muslims have to do is simplify and rationalize their criterion if they wish to unify the Islamic dates globally and eliminate the confusion.

Let us start from basic facts:

1. The earth is not flat. It is a globe (1).
2. The sun and the moon take 24 hours to be visible over the globe.

But their visibility differs considerably.

* The sun is visible over half of the globe all the time, but the moon disappears on the 27th night. After 1 or 2 nights, it appears again as a crescent after the sunset.

* The sun is visible at a north-south longitude, but the crescent’s visibility starts at a point, and extends west, north/south-west in a parabola. See the maps.

*The place of crescent’s first visibility changes every month. Not only it moves west, but north and south as well. In March the crescent may first become visible as far north as south of Alaska, Mongolia, or Nova Scotia. By August it moves as far south as Madagascar, New Zealand, etc.

3. The moon has no light of its own. It is like a dim mirror when it reflects the sun’s rays.

The moon’s surface is uneven. The mountains and craters block the sunlight from reflecting.

4. The crescent formation starts after it has moved some 7+ degrees away from the sun (2)

5. The moon completes its synodic monthly cycle (New Moon to New Moon phase) in 29 days 12 hours 44 minutes and 2.9 seconds that results in some months being of 29 and others 30 days long.

6. The New Moon is completely dark ( 0 ). It is not visible from the surface of the earth.

A NASA photograph of the New Moon in 1971 from 100+ miles above the surface of the earth proved non-existence of the moon's crescent at the conjunction. Any claim of sighting before the New Moon phase or immediately after it must be rejected.

7. The New Moon phase can be very precisely calculated.

A calendar day/date starts from a fixed point of time. We cannot make an Islamic calendar on the basis of the New Moons: They occur at all times of the day and night. An Islamic day / date starts at sunset. The solar date starts from midnight, at the international dateline. The two cannot be identical

8. The New Moon becomes visible as a crescent usually 18-25 hours later if its angle from the sun and its altitude above the horizon are high enough for the surface of the moon to reflect the sun’s rays. It is not necessary that a crescent will always become visible after the moon's age is 15+ hours. Sometimes you cannot see a 60 hours old moon in South Africa, Germany, England, Alaska or southern Chile, etc.

3. 1 What the Quran says?

The Muslims in N. America, like the Muslims all over the world are seen starting Ramadan and celebrating Eidain on 2, or even 3 different dates. Every Muslim organization and all Muslim countries claim that they strictly follow the Quran and Hadith to determine the beginning of Ramadan and for Eidain dates. They quote the same Ayat and Ahadith.

We know that there is only one moon for our earth. Once its crescent is visible after the New Moon phase it becomes visible over most of the globe within 24 hours. Then why is this confusion and chaos? Let us try to find some answers.

We blame the crescent-sighting for the mess. But the real culprits are human errors, prejudice, lack of understanding, and wrong interpretation of certain texts.

The most quoted aayat of the Qur’an about the moon-sighting controversy are:

...They (the crescents) are the determinants of time ... and for the Hajj. (2:189).

..(He) fixed stages for it (the moon) that you might know the number of years and the count (of time) ... (10:5)

...The moon, We have fixed for it positions (phases) till it goes back to the (thin) shape of an old dry date-stalk. (36:39)

The sun and the moon are on exactly computed courses. (55:5)

The usual interpretation of theseAyat favors the crescents (Ahillah) as determining the beginning of an Islamic lunar month.

3.2 How the Sunnah guides ?

The Messenger (S) determined the beginning of an Islamic month by observing the earliest crescent or after completing thirty days of the month in case the horizon was hazy or cloudy. The most famous of these traditions is:

“Do not fast unless you see it (the crescent), and then do not stop fasting until you see it (again)...” (Bukhari, Muslim and others)

3.3 How the Ummah decided?

The Khulafa al-Rashidun, the companions and early Muslims followed the Sunnah of determining the beginning and end of Ramadan by observing the crescent. The famous Hadith of 'Kuraib' makes it quite clear. If a crescent was not seen on a clear horizon then the news from another place, account of the witnesses, or even completing thirty days of Ramadan was not enough to begin the next month. Fuqaha are also unanimous on the point that the crescents determine the beginning of an Islamic month.

As the complicated calculations needed for accurately predicting the earliest visibility of the crescent were not available till 1980s the Fuqaha tried to resolve the questions of how and what of the 'crescent's visibility' according to whatever information was available to them. How many witnesses are sufficient, whether the news from another place should be taken into account, what if 30 days are complete and a crescent is not visible, should the predictions of the star-gazers (Munajjimeen) be given any weight, etc. were discussed at length. Conflicting rulings hardened into Fiqh positions, each group trying to support its statement with as much as it could muster.

The problems started surfacing when the distances were reduced and communications easy.

4.1 Basic Calendar Concepts

a. An Islamic day begins from after sunset and ends with the sunset next day.

b. An Islamic lunar month begins from the earliest visible waxing crescent moon and ends with the next earliest visible crescent, except that it cannot be more than 30 days.

c. A global Islamic calendar date must always begin at a set point on earth, from a set point of time.

Local Islamic calendars were easy and straightforward to make. The average lunar month was approximately 29 ½ days, thus making months alternately 29 and 30 days long. The mean length of a lunar month is shorter by 0.0306 days (44 minutes 3.8 seconds) than the actual length, the calendar month ends a little earlier than the synodic month. So the beginning of the lunar calendar month gradually begins to lag behind the New Moon. This difference accumulates to about 11 days 18 minutes 43 seconds in 30 years.

For civil needs, an easy solution was to make the last month of the year –Dhul-Hijja, 30 days instead of 29 in the 2nd, 5th, 7th, 10th, 16th, 18th, 21st, 24th, 26th, and 29th years of a 30 year cycle. This gives an average month of 29.5305555…days and average year of 354.3666…=354 and 11/30 days.

The Ottomans used this formula to fix their civil calendar. However, for religious purposes, the Muslims in general adhered to the ‘visibility’ criterion, often fasting or celebrating Eidain on two days in towns hardly a few miles apart. As the means of communications reduced the distances among the Muslim countries, the confusion about these dates intensified. Various Muslim countries adopted standards in the name of unifying these dates within their domains (3).

4.2 Causes of Confusion

The scientific advances in the lunar visibility theory made significant advances between 1977-1990 (4). The credit goes to both Dr. S. Kamal Abdali and Khalid Shaukat for their methodical rigor and meticulous efforts. Yallop, Doggett, Imad and Schaefer also contributed. With their help the Committee has maintained significant accuracy in predicting moon's earliest visibility in many part of the world. Unfortunately, the official calendar-makers in various Muslim countries closed their eyes, and continued harping on the "unity of Islamic dates", instead of searching for solutions of complicated Islamic lunar calendar problems (5).

5.1 Some Suggestions

We cannot adopt the conjunction (NM) as the criterion for obvious reasons. Most Muslims would also like to keep some sort of moon’s earliest visibility as the criterion for global Islamic dates.

We suggest:

1. The solar international dateline (IDL) should also be adopted as the lunar international dateline (LIDL). It appears most practical for obvious reasons.

2. Makka Mukarrama as the focal point for determining the crescent’s visibility.

3. If the conjunction occurs at, or before Makka then the date following the New Moon date should be the first date of the Islamic lunar month from IDL at sunset.

Example: On April 5, 2000 the conjunction occurs approximately 13.5 hours before IDL. The Islamic month will begin on April 6 from IDL.

4. If the conjunction occurs after Makka then the first date of the Islamic month should be two dates following the New Moon date.

Example: On July 1, 2000 the conjunction occurs approximately 12 ½ hours after the IDL. The Islamic month therefore will begin on July 3 at IDL.

This solution may not fulfill the strict ‘lunar visibility’ conditions for all places on the globe, and may require some minor adjustments yet it appears the most practical. Other variations like “ moon’s age=12+ hours at Makka” may be appealing to some but they create more complications, especially for areas east of Makka.

Without a fixed Lunar International Dateline at IDL the lunar month will always be 29 and 30 days long simultaneously, making a unified calendar an impossibility..


1. (‘Wahdat-i Matla’ position assumes that the earth is flat, and that once a crescent is seen anywhere it is seen instantly over the whole world. It is a fact that a place only a few miles east of the initial lunar sighting could not see it on the same date. The moon will be visible there next day after 24 hours when it will appear again on its horizon. Ikhtilaf-i Matali' position is equally invalid. It assumes that every village on the surface of the earth will see a crescent on a different date. Most of the world will see a crescent within 24 hours.)

2. Other factors, especially the glare on the lower horizon further hinders moon's visibility. Generally the Moon should be 10 degrees above the horizon at sunset to be visible in the next 12-25 minutes. Remember the moon is coming down and will be only 4-5 degrees above the ground by the time you see it after 29 days. A crescent after 30 days will be higher and brighter. Many mistake it for being two days old.

3. At present very few Muslim countries use the visible crescent for determining Ramadan and Eidain dates. Their own criteria of determining the beginning of an Islamic month violate clear guidelines of the Quran, Sunnah and Fiqh (that tell of a 'visible crescent' (Hilal). Some commonly used criteria for fixing the first date of an Islamic month are:

a. The day / date of the New Moon (Conjunction).

Saudi Taqweem Umm al-Qura is based on the New Moon (not the visible crescent) dates.

(The New Moon is not a crescent (Hilal) of the Quran and Sunnah. It is completely dark, and it occurs at all times of day or night. Saudis do not convert the New Moon dates given in almanacs into the local dates in S. Arabia. Their calculated first date of an Islamic month, based on the New Moon may be two days/dates ahead of the actual sighting in S. Arabia. Gulf states generally follow Saudi Arabia's lead.)

b. The date following the New Moon (Conjunction) date.

(Malaysia, Indonesia, Algeria, Tunis, etc. prepare their calendars on this principle.)

The crescent will not be always visible on the assumed date.)

c. The Moon is setting 1 to 5 minutes after sunset.

(Egypt follows this criterion. Hence Ramadan and Eid al-Fitr dates in Egypt sometimes differ from S. Arabia. )

(The moon’s surface is not flat. A crescent cannot be visible the instant the moon becomes New. If the purpose is to wait till the New Moon becomes a crescent then waiting 1-25 minutes do not qualify for a visible moon. In many months the moon may set 1-5 minutes before the sun in most of Egypt, but one minute after the sun in an small area. The moon may also set after the sun before the New Moon phase. What to do in these and similar cases? )

d. The moon’s altitude was 5 degrees or more.

(Turkish Islamic dates are fixed by this criterion except for Eid al-Adha when Saudi fixed Hajj date is used to determine Eid al-Adha (Bairam Kurban).

d. The Moon’s age is 15 hours or more.

(The moon is not always visible when its age is 15 hours. Often a 35 hours old moon is not seen. But many accept claims of frivolous sightings.)

e. The moonset is 30-50 minutes after sunset.

(A moon setting an hour after sunset is sometimes not visible in a country like England or South Africa. The age of the moon or the time lag between the sunset and the moonset are least important factors for its visibility.)

4) The Moon's visibility from earth is no longer important for astronomers and observatories. When a Muslim asks them about Islamic dates he is asking a wrong person a wrong question. He should ask: When would the moon be visible in his town? He will be amazed by their reply. Most of the astronomers have no idea what you are asking about. They look to any calendar and see the NEW MOON date marked on it. They guess a crescent should be visible soon after the moon is NEW (which is not true) and tell you so.

Some astronomical experts have taken a keen interest in moon's earliest visibility since 1986 after attending an IIIT seminar in Virginia. They are doing excellent work on various aspects of moon's visibility. But a mad race is also going on for spotting the youngest moon among amateurs who are trying to break the record of the earliest visibility. In recent years several claims were made which later proved to be bogus, or very marginal. Remember that a crescent will NOT ALWAYS BE VISIBLE at the record-breaking point of (age, time-lag, angle or altitude). Many other factors determine a moon's visibility.

A global lunar Islamic calendar involves lots of complicated issues which the Muslim Ulema and calendar experts have yet to cipher out. Many Muslim 'astronomical' (?) experts confuse others by taking positions like: a) the New Moon is Hilal (S. Arabia); 2) The New Moon is visible (Egypt); 3) the New Moon will become a crescent somewhere in the world on the same solar date (Turkey), or something else which they believe is a possible solution for complicated lunar calendar problems. Every effort to bring the Ulema and the Muslim calendar experts together has so far failed mostly because of the lack of communication, and the biases of the sponsors of international conferences. Often they were organized to pass resolutions of ‘unity of Islamic dates’ rather than for finding solutions to complicated problems.

Moon Observation Data 1995-2005

The conclusions are obvious:

a) Saudi Taqweem follows the New Moon date blindly.

b) The witnesses who claim seeing a crescent there on the New Moon dates are not correct. There is no other explanation for a consistent behavior of Saudi eye-witnesses. Once in a while one may mistake some other shiny object for a crescent (though it is improbable on mostly clear skies in S. Arabia). But they are doing it for 25+ years since 1971 when CFCO started keeping records. If the members of Saudi Majlis al-Qada al-A'la and Egyptian Religious Council have eyes then from their palaces they can easily see that there is no crescent on the horizon on the dates they accept 'Udul' witnesses. An analysis of Saudi Taqweem dates for 1973-2019 proves that their witnesses can see a crescent six times every year before the conjunction and in the other six months within 1-10 hours of the New Moon phase. Obviously, these dates do not match the visible crescent which occurs there one or two days later.

Selected Bibliography

1. Abdali, S.K. 1979 “On the crescent’s visibility” in Al-Ittihad (v.6:no.1/2)
2. Abdlai, S.K. 1985 “Notes on the data for crescent observation” (Unpublished)
3. Abuzaakouk, A. 198? “Legal discussions about the sighting of the new moon of Ramadan” (Unpublished)
4. Afzal, Omar 1982 “Witnesses or calculations?” in Tahrik New York (v.4:no.1)
5. Afzal, Omar 1987 “Crescent observation reports 1982-1987” in New Trend, Maryland (v.7:no.4)
6. Afzal, Omar 1987 “Lunar Islamic Calendar : Issues and answers” in the Proceedings of the Lunar Calendar Conference 1986 IIIT Herndon, VA pp.3:1-3:23
7. Afzal, Omar 1989 “Qaradawi on Moonsighting : a rejoinder (IIIT:Unpublished)
8. Afzal, Omar 1991 “Determining the date of Eid al-Adha” The Message Intl. New York (v.15:no 1)
9.Afzal, Omar 1992 “Ruyate Hilal ke masael ka ek naya tajziya” in Majallah Tahqiqat-I Islami, Aligarh, India (v.11:no.1 March 1992)
10. Afzal, Omar 1993 Ruyat-I Hilal, Idara Ma’arif Islami, Karachi Pakistan
11. Ahmad, Imad 1990 A Uniform Islamic calendar for the western hemisphere. Imad-ad-Dean, Bethesda, Maryland
12. Ashbrook, J. 1984 Astronomical scrapbook, Sky Pub. Co. Cambridge. M.A.
13. Dershowitz N. & E.M. Reingold 1997 Calendrical Calculations. Cambridge University Press, New York.
14. Doggett, L. E. 1994 “Lunar Crescent Visibility” ICARUS (v.107: p388-403)
15. Ilyas, M 1995 Islamic Calendar 1407-1421 OIC, IIIT, USM and IICP Penang, Malaysia
16. Ilyas, M. 1997 Astronomy of Islamic Calendar A.S. Nooradeen, Kuala Lumpur, Malaysia
17. Qasum et. al 1997 Ithbat al-Shuhur al-Hilaliyah wa mushkilat al-Tawqit al-Islami. Dar al-Taliyah, Beirut.
18. RGO Astronomical Information Sheets 1987-2000
19. Schaefer et. al 1993 “Records of young Moo sightings” Q.J.R. Astr. Soc. (v.34:pp.53-56)
20. Scheafer, B. E. 1993 “Astronom and the limts of vision” Vistas of Astronomy (v.36,pp. 311-361)
21. Taqweem min 1409 hatta 1440 1408 Madina al-Malik Abd-Aziz lil-Ulum wa al-Taqniya Riyadh