Updated My New Moon Calculations

Posted by Byron Pendason on August 22, 2025 CE, in Heathenry, Anglo-saxon calendar, Reconstruction

Earlier this month, I asked a British friend1 who starts his lunar months with visual observation of the new crescent moon for reference dates of his new crescent sightings. I was wanting to check the accuracy of a new methodology for calculating new moons. Unfortunately, that method was garbage according to the reference dates I was given. Just for shits and giggles, I decided to check another method that I had stumbled upon, and to my surprise, it matched all the reference dates exactly. So,as of 2025-08-112 (the 17th of Æfterra Liða) my calendar page uses this new algorithm. In an effort for full transparency, I wanted to discuss it a little bit here.

A Quick Review About Anglo-Saxon Months

We know from Bede that the Anglo-Saxons used lunar months, and he tells us that lunar months went from the light of the new moon to the next new moon. Although in modern times the new moon is the conjunction when the moon is completely dark and unable to be seen, this is not what ancient cultures who used lunisolar or solar calendars meant when they used the term (with a few exceptions, such as the Chinese). What they meant was the first visible crescent after the conjunction, usually visible 17 to 36 hours after the conjunction. This is why a lot of lunar or lunisolar calendar observers today prefer the term new crescent or new crescent moon over simply new moon, to avoid confusion.

Unlike the conjunction which can have its precise moment calculated with 100% accuracy to within a few seconds, the new crescent is notoriously difficult to calculate, especially if you’re speaking of a very small and specific area. Many factors can effect visibility, such as weather, altitude or elevation, the time of day, etc. So it’s virtually impossible to calculate new crescents with any kind of certainty.

For this reason, most lunar and lunisolar calendars have rules to try to mitigate the significance of those factors. For example, if the new moon is spotted anytime during the night up until sunrise, it counts as the start of the month. And if a new crescent hasn’t been seen on the thirtieth, the next day is considered the first of the month. I think both of these are reasonable rules, and given how widespread and common they are I think it’s likely the Anglo-Saxons had these, or at least similar, rules.

Some lunisolar calendars have even ceased to go by moon sightings. The Hebrew Calendar (used by Jews across the world) has been a calculated calendar for the last 1500 years or so. Not only do they determine when to add leap months and leap days using the Metonic Cycle, but the number of days for each month are fixed, some months only varying month length by what year in the Metonic Cycle they are in.

Stumbling Upon The New Algorithm3

With my interest in lunisolar calendars, I like to try to keep up to date on news involving lunar or lunisolar calendars. It appears in recent decades that there’s been a huge controversy in the Islamic world about whether to go only by sighting the new crescent or whether to accept astronomical calculations. This isn’t anything new for calendars that use lunar months. Even though the Hebrew Calendar is calculated now, there are still Jews (known as Karaites) who reject that calendar in place of one with actual sightings of the moon, with leap months determined by the spring equinox and the ripening of barley in Israel.

This conflict seems to really be going strong among Muslims. Traditionally, the crescent moon was required to be seen for a lunar month to begin. Back in 2006, however, The Fiqh Council of North America (and though I can’t find anything from them about it, I’ve read that the Islamic Society of North America joined them in the ruling) made a ruling that in order to promote unity among Muslims, they were adopting a way to calculate new moons so that all Muslims can follow the same calendar. There are Muslims in North America, however, who don’t trust this, and take sighting reports every month and determine the new month from them. Amazingly, they seem to line up pretty well, in spite of the conflict that would lead one to believe there were wild discrepancies. As it turns out, it’s the act of sighting the new moon itself, and not the calculations or alleged inaccuracies, that’s at issue here.

It was this news about North American Muslims beginning to use calculations that intrigued me, and I really became curious how they were calculating it. And the simplicity of the new algorithm stunned me with the high level of accuracy it seemed to have.

The Algorithm to Calculate the Crescent Moon

The algorithm is really simple. You take the time of the conjunction in UTC, and add one day if it’s before noon. If the time of the conjunction is past noon, add two days. This is the estimated new crescent date.

This makes sense, if you think about it like this: assuming we’re going with the rule that the crescent moon can be seen anytime during the night, a conjunction before noon would give an approximate age of the moon of at least 18 to 19 hours at the end of the night at sunrise, and the visibility of the new crescent is generally visible at about 17 to 18 hours. Conjunction past noon would likely not give the moon enough time to be visible by sunset the following day.

Comparing the dates this algorithm gives with actual moon sightings (such as CrescentWatch.org) seems to provide a very high degree of accuracy. The only discrepancy for this year was the first new moon, which I have calculated at January 1, 2025 but CrescentWatch.org has listed as January 2. Looking at the conjunction time for that moon gives us December 30, 2024 at 22:27, so I suspect an additional rule to this algorithm that adds three days if the conjunction is past astronomical sunset (since that’s the next day anyway by Anglo-Saxon reckoning) would improve the accuracy. I’m holding off on implementing that rule though until I can sift through more new crescent sightings to find more discrepancies to test it out.

The only other source for discrepancies that I can foresee would be conjunctions that are within a half hour or so of noon. Being so close to the threshold, they could probably go either way. I cannot think of a way to account for this, but the conjunction being within the hour window of noon should theoretically only happen 4.1% of the time, and considering a new lunar month only happens 12-13 times a year, it should be rather rare. It is important to note, however, that due to local conditions potentially being big factors in new crescent visibility, it’s impossible to come up with an algorithm that will be 100% accurate for everyone. We can either come up with something that’s close enough and is accurate most the time and is only off by a day the rest of the time, or continue to use visible sightings as our methodology and be unable to calculate the calendar ahead of time very accurately.

Objections

Should We Be Trusting Non-Germanic Sources For Our Calculations?

I can see this objection particularly from pantheon purists. This is going to be a short but probably not so sweet response as I find myself less and less able to tolerate their nonsense as time goes on. Maybe it’s just because I feel like I’m getting old so I’m manifesting the stereotype of the grumpy old man, but who knows? If there is anyone who uses my calendar who objects to my use of non-Germanic sources, I will feel no great loss if they quit using it just because my algorithm originally came from Muslims trying to solve the same problem I have been.

Woden (known as Oðin to the Norse) travelled the worlds seeking wisdom, wandering far and wide in his search. He didn’t particularly care where the wisdom came from, he even took wisdom from his enemies when he found it useful. Likewise, when it comes to that which is useful, in this case a calendar algorithm, I don’t care where the source is from. If it works, it works!

Due to their location, the Islamic world was at a crossroads of several cultures, including Hellenic, Babylonian, Chinese, and Hindu civilizations. They made great contributions to mathematics due to being able to take from all these different cultures and put together the pieces to understand the universe better. Particularly when it comes to astronomy. To reject their contributions just because they’re not Germanic is foolish. It reeks of the genetic fallacy. It’s an undeniable historical fact that the Muslims had a lot of contact with the Norse. We have Norse artifacts with “Allah” written on it, some Norse Heathens revered certain Islamic figures like Ali, and they even adopted some mathematical concepts from the Muslims. In short, I don’t think Germanic polytheists learning from Muslims is all that unprecedented.

I’m not saying we should all say the Shahada and become Muslims, I just don’t think we should be rejecting useful knowledge just because it came from Islamic sources.

We Should Only Be Using Visual Observation For the Beginning of Lunisolar Months

I actually have a lot of respect for people of this opinion. The dedication to go out and and visually check for the new crescent each month is admirable. I do think that it is impractical for a great many of us to do so, however. It becomes necessary to live out in the country outside the city limits and the light pollution that comes with it. If you live in the city and it’s not practical to go out into the country every month to manually look for the new crescent, you need an alternative. One option is to rely upon new crescent sighting groups. But at that point, does it really matter if the new crescent is spotted or calculated?

There are also benefits to knowing in advance when your month is going to start. You can plan better, maybe take vacation from work for example. One reason that Muslims in America have started switching to a calculated new crescent is because many jobs require 2 or more weeks advanced notice for taking time off work, and its easier to get your holidays legally recognized if you can tell them when your holiday is rather than saying “Well, its going to be one of these two days, depending on when the new crescent can be seen.”

Most lunisolar and lunar calendars that have survived into the modern day have evolved to using a calculated calendar. I mentioned earlier how the majority of Jews have been using a calculated calendar for almost a millennium and a half, and how many Muslims are transitioning to using a calendar that can be planned out years in advance. Same thing is happening with Hindus, especially those outside India. The Chinese have also been calculating their calendar for quite a while now (though admittedly it’s a lot easier for them since the conjunction begins their months and can be calculated in advance with great precision). The transition from purely observational calendar to one that can be calculated seems to be a natural development for lunisolar calendars, and I think the Anglo-Saxon Heathens would have been no different if their religion had survived into the modern day.

Conclusion

I have been trying for a long time to get my calendar app to be as accurate as possible. The problem that I’ve consistently ran into is accurately calculating the beginning of the month because there are so many variables that can factor into the visibility of the new crescent. I’ve used several different methods over the years that I’ve been using my reconstructed Anglo-Saxon calendar. My newest algorithm appears to have at least an 87.5% accuracy when compared to new crescent sightings, and the remaining 1 out of the 8 lunar months so far this year appears to be remedied with a simple adjustment for conjunctions close to midnight. Mine Wyrtruman is committed to having the highest accuracy possible, and in pursuit of that goal I present my latest update to my calendar page.

9e3tt qthat require a location. I have to use some time zone as my “international dateline”, and the one that contains the homeland of the people for whose calendar I am trying to reconstruct only makes sense to me. This also makes calculations easier since most astronomical data seems to be given in UTC.

  1. I use UTC for all my calculations for the new moon, and I use Tamworth, England, for any calculations ↩

  2. I had intended for this blog post to come out much earlier in the month, but a bout of sickness caused me to have to delay it. ↩

  3. This section is all about how two other religions, outside of Heathenry, have dealt with this question about calculating the beginning of months, and the conflicts caused by it, so feel free to skip it if that doesn’t interest you. ↩