21 days in 365 · measured across five cities

Why two pañcāṅgas disagree about today's tithi

On 21 days of 2026, two Indian cities are on different tithis at the same moment - and both are right. It is not ayanamsa, which cancels out exactly. It is where you stand.

Not because one of them has the wrong ephemeris. In 2026, on 21 days out of 365, two Indian cities are genuinely on different tithis at the same moment — and both panchāṅgas are right. The tithi depends on where you are standing.

First, what it is not

The usual explanation is ayanāṁśa — the same thing that makes two apps give different Moon signs. It has nothing to do with tithi.

A tithi is the angular distance between the Moon and the Sun, divided by twelve. Ayanāṁśa is subtracted from both longitudes, so it cancels out of the difference exactly. Computed for one instant:

Every longitude differs. The difference between them is identical to five decimal places, tropical included. So whatever is causing your two panchāṅgas to disagree, it is not this.

The actual reason: a tithi is named for sunrise

A tithi does not begin at midnight. It begins whenever the Moon–Sun angle crosses a multiple of twelve degrees — which can be any time of day or night — and it can run anywhere from about nineteen to about twenty-six hours.

The convention is that today's tithi is the one prevailing at sunrise. And sunrise is a local event.

Across India, sunrise spans about 65 minutes — Kolkata sees it that much earlier than Mumbai. If a tithi happens to end inside that 65-minute window, then Kolkata woke up in one tithi and Mumbai woke up in the next. Both panchāṅgas name the tithi correctly for the place they were computed for.

How often that happens

We checked all 365 days of 2026 across five cities — Kolkata, Mumbai, Delhi, Chennai and Ahmedabad — computing each city's own sunrise and the tithi prevailing at it.

21 days. One day in seventeen. And it clusters: 7, 8 and 9 January 2026 run three in a row, with Kolkata and Chennai a full tithi behind Mumbai, Delhi and Ahmedabad each morning.

Those are the days when a festival date is "wrong" in one state and right in another, and nobody has made an error.

The smaller cause: which sunrise you mean

There is a second disagreement underneath the first, and it is about four minutes wide. "Sunrise" is not one definition:

Four minutes is small beside sixty-five, but it is not nothing — on a day when a tithi ends at 06:48 in Delhi, it decides the answer on its own.

We use the upper-limb-with-refraction definition because that is what the Indian pañcāṅga tradition uses, and because it is what DrikPanchang matches. It is set in one constant in our engine, and the comment beside it says why: "flip to True only for astronomical (disc-centre) sunrise, not Drik-style panchāṅga muhūrta."

What to do about it

Use a panchāṅga computed for your location, not for a city you do not live in. On 344 days a year it will not matter. On the other 21 it is the whole answer.

And if a panchāṅga does not ask you where you are, it cannot be right for you except by luck.

Sources

• Ayanāṁśa cancels out of tithi — computed — Moon−Sun = 80.57353° identical under Lahiri, Raman, Krishnamurti and tropical, 19 Aug 2026 06:30 UT
• Cities disagreeing, all of 2026 — 21 of 365 days across Kolkata, Mumbai, Delhi, Chennai, Ahmedabad — each at its own sunrise; 7–9 Jan run consecutively
• Sunrise definitions compared — Delhi 1 Mar 2026 — upper limb + refraction 06:46:17; disc centre + refraction +1.2 min; disc centre, no refraction +4.1 min
• The convention this site uses — backend/v2/astromata_core.py — SUNRISE_DISC_CENTER = False, i.e. upper limb with refraction, the Drik-style pañcāṅga convention

Check it yourself — free

• Today's pañcāṅga for your own place
• Your personal pañcāṅga, day by day
• Ekādaśī dates with the nirṇaya

All articles · VedicSpace — every number computed on the Swiss-Ephemeris engine (Lahiri), every rule named to its classical source.