The image of the Scandinavian Ramadan — twenty hours without water, a sunset that barely arrives — is fixed in a lot of people's minds, and for good reason: it was accurate, and it produced a serious body of scholarly discussion about what to do at latitudes the classical rules were not written for.
It is also, right now, wrong. The Hijri year is about eleven days shorter than the solar one, so Ramadan walks backwards through the seasons, and it has walked out of the northern summer. We computed the numbers to show where it has got to.
The measurements
Fast length from Fajr to Maghrib, Muslim World League angles, with the middle-of-the-night high-latitude rule applied where the angle is not reached. Taken at a representative date in each Ramadan.
In Tromsø in the June rows the calculation returns no value at all, because the sun does not set: there is no Maghrib, and so no ordinary end to the fast. Those cells are marked rather than filled with a substitute.
| Ramadan | Tromsø | Stockholm | London | Istanbul | Mecca | Jakarta |
|---|---|---|---|---|---|---|
| 2017 (Jun) | no sunset | 21.2 | 20.3 | 17.2 | 14.8 | 13.0 |
| 2026 (Feb) | 12.5 | 12.5 | 12.6 | 12.7 | 12.9 | 13.4 |
| 2027 (Feb) | 10.9 | 11.6 | 11.9 | 12.2 | 12.7 | 13.5 |
| 2030 (Jan) | 8.0 | 10.0 | 10.8 | 11.4 | 12.4 | 13.6 |
| 2047 (Jun) | no sunset | 21.0 | 20.1 | 17.0 | 14.8 | 13.0 |
What the numbers say
The ordering has inverted. In 2017 the northern fast was the longest by a margin of eight hours over Jakarta; in 2030 it is the shortest by a margin of five and a half hours in the other direction. Tromsø in 2030 fasts eight hours — shorter than any fast anywhere in the equatorial belt, and shorter than a working day.
2026 is the crossover year, where the whole set converges. Around the equinox every latitude gets roughly twelve hours of day and twelve of night, so a Ramadan that falls near an equinox produces near-identical fasts from Tromsø to Jakarta. That is a genuinely unusual configuration and it will not recur for decades.
Jakarta barely moves across the entire table: 13.0, 13.4, 13.5, 13.6, 13.0. Near the equator, day length hardly varies by season, so the drift of Ramadan through the calendar changes almost nothing. The whole phenomenon is a high-latitude one.
And the twenty-hour fast returns around 2047 — the numbers for that year are almost identical to 2017, as the cycle completes.
Tromsø, and the case with no arithmetic answer
The blank entries in the table are the important ones. In Tromsø in June the sun does not set, and above the Arctic Circle there are stretches of winter in which it does not rise. In those conditions the definition of the fast — dawn to sunset — has no referent. This is not an edge case in the calculation; it is the absence of the thing being calculated.
Three approaches appear in the contemporary discussion, and all three have serious scholars behind them. The first is to follow the nearest location at which a normal day and night occur. The second is to follow Mecca's timings, which has the appeal of a fixed reference. The third is to divide the twenty-four hours by a proportion taken from a period when times were normal, or to use the midpoint of the night as a substitute dawn — which is what the calculation above does.
The practical point for anyone north of about 48 degrees is that an app is choosing among these on your behalf, usually silently. It is worth finding the high-latitude rule setting and knowing which one you are on, because the choices differ by hours rather than minutes. We measured where that cap starts to bind, and what it does to the calculation methods, in why prayer time apps disagree.
Why this is worth knowing now
Chiefly because the received wisdom is stale. Advice written in the mid-2010s about surviving a twenty-hour fast, and the fiqh discussions that accompanied it, described a real situation that has since reversed — and someone moving to Oslo today and reading that material would be preparing for a difficulty that will not arrive in their lifetime at that intensity.
It is also a reminder of what the lunar calendar does. Ramadan is not attached to a season; it passes through all of them over roughly thirty-three years, so that a community experiences the long fasts and the short ones in turn. That rotation is a structural feature rather than an inconvenience, and it means no region carries the hardest version permanently.
For the practical side of the month — what breaks the fast, and the sunnah around the two meals — see what breaks the fast and suhoor and iftar. For when the month begins, which is its own dispute, see why Ramadan and Eid start on different days.
Common questions
- How long is the Ramadan fast in northern Europe now?
- Shorter than at the equator. In 2027 we compute Tromsø at 10.9 hours and Stockholm at 11.6, against Jakarta's 13.5; by the 2030 Ramadan, Tromsø is down to 8.0 hours. Ramadan has drifted out of the northern summer.
- When do the twenty-hour fasts come back?
- Around 2047, when Ramadan returns to June. Our figures for that year — Stockholm 21.0 hours, London 20.1 — are within a fraction of the 2017 values.
- How do you fast where the sun never sets?
- The ordinary definition has no referent there, so a substitute rule is used. The main approaches are following the nearest location with a normal day and night, following Mecca's timings, or dividing the day by a proportion — commonly the midpoint of the night as a substitute dawn. All three have scholarly support, and your app has silently picked one.
- Why do all cities have almost the same fast length in 2026?
- Because that Ramadan falls near the spring equinox, when every latitude gets close to twelve hours of daylight. Our figures span only 12.5 to 13.4 hours from Tromsø to Jakarta — a convergence that will not recur for decades.