# When will the sky go dark at noon over your town?

In 1961 and in 1999, the Sun went out over Romania in broad daylight. Pick the nearest town and see all of its eclipses from 1900 to 2100: which were total, how much Sun was left each time, and what comes next.

By [Marius Comper](https://mariuscomper.uk/en/) · 21 September 2026 · every time and percentage is computed from NASA's Canon; the method is under [sources](#surse)

## Pick the nearest town

The list holds all 42 county capitals. If you live in a village or a small town, pick the nearest one: a few dozen kilometres away the sky looks the same, except right on a shadow's edge.

Next eclipse: 2 August 2027 (partial, 12:32 EEST).

## Do you remember 11 August 1999?

What did your town see that day? Guess before you look.

## When does your town enter the shadow next?

The next totality over your town comes…

## Every eclipse over your town

- **28 May 1900** partial: 53% — 17:57 UTC+1:44 (total, saros 126)
- **11 November 1901** partial: 57% — 07:39 UTC+1:44 (annular, saros 141)
- **31 October 1902** partial: 2% — 08:31 UTC+1:44 (partial, saros 151)
- **30 August 1905** partial: 60% — 15:27 UTC+1:44 (total, saros 143)
- **17 April 1912** partial: 53% — 14:26 UTC+1:44 (hybrid, saros 137)
- **21 August 1914** partial: 86% — 14:36 UTC+1:44 (total, saros 124)
- **23 January 1917** partial: 53% — 08:44 UTC+1:44 (partial, saros 149)
- **8 April 1921** partial: 44% — 10:50 UTC+1:44 (annular, saros 118)
- **28 March 1922** partial: 36% — 16:27 UTC+1:44 (annular, saros 128)
- **29 June 1927** partial: 54% — 06:50 UTC+1:44 (total, saros 145)
- **12 November 1928** partial: 24% — 10:43 UTC+1:44 (partial, saros 122)
- **21 August 1933** partial: 48% — 07:03 EEST (annular, saros 134)
- **19 June 1936** partial: 86% — 06:58 EEST (total, saros 126)
- **10 September 1942** partial: 20% — 18:32 EET (partial, saros 153)
- **9 July 1945** partial: 56% — 16:38 EET (total, saros 145)
- **28 April 1949** partial: 4% — 09:08 EET (partial, saros 147)
- **25 February 1952** partial: 30% — 11:41 EET (total, saros 139)
- **30 June 1954** partial: 79% — 15:19 EET (total, saros 126)
- **2 December 1956** partial: 35% — 09:01 EET (partial, saros 151)
- **2 October 1959** partial: 13% — 14:46 EET (total, saros 143)
- **15 February 1961** total: 100% — 09:54 EET (total, saros 120)
- **20 May 1966** partial: 83% — 11:45 EET (annular, saros 137)
- **22 September 1968** partial: 41% — 13:04 EET (total, saros 124)
- **25 February 1971** partial: 20% — 11:59 EET (partial, saros 149)
- **30 June 1973** partial: 0% — 13:53 EET (total, saros 136)
- **11 May 1975** partial: 13% — 08:11 EET (partial, saros 118)
- **29 April 1976** partial: 68% — 12:52 EET (annular, saros 128)
- **15 December 1982** partial: 38% — 10:51 EET (partial, saros 122)
- **4 December 1983** partial: 1% — 15:07 EET (annular, saros 132)
- **12 October 1996** partial: 45% — 17:52 EEST (partial, saros 153)
- **11 August 1999** total: 100% — 14:07 EEST (total, saros 145)
- **31 May 2003** partial: 66% — 06:09 EEST (annular, saros 147)
- **3 October 2005** partial: 36% — 12:31 EEST (annular, saros 134)
- **29 March 2006** partial: 71% — 13:57 EEST (total, saros 139)
- **1 August 2008** partial: 9% — 13:07 EEST (total, saros 126)
- **15 January 2010** partial: 3% — 08:14 EET (annular, saros 141)
- **4 January 2011** partial: 67% — 10:35 EET (partial, saros 151)
- **20 March 2015** partial: 43% — 11:55 EET (total, saros 120)
- **21 June 2020** partial: 4% — 08:39 EEST (annular, saros 137)
- **25 October 2022** partial: 38% — 13:38 EEST (partial, saros 124)
- **2 August 2027** partial: 51% — 12:32 EEST (total, saros 136)
- **1 June 2030** partial: 83% — 08:09 EEST (annular, saros 128)
- **20 March 2034** partial: 24% — 12:54 EET (total, saros 130)
- **16 January 2037** partial: 44% — 11:41 EET (partial, saros 122)
- **2 July 2038** partial: 5% — 17:38 EEST (annular, saros 137)
- **11 June 2048** partial: 66% — 17:10 EEST (annular, saros 128)
- **14 November 2050** partial: 52% — 16:28 EET (partial, saros 153)
- **12 September 2053** partial: 49% — 11:41 EEST (total, saros 145)
- **5 November 2059** partial: 56% — 09:53 EET (annular, saros 134)
- **30 April 2060** partial: 63% — 13:40 EEST (total, saros 139)
- **5 February 2065** partial: 45% — 12:22 EET (partial, saros 151)
- **12 September 2072** partial: 6% — 11:05 EEST (total, saros 155)
- **13 July 2075** partial: 79% — 07:35 EEST (annular, saros 147)
- **26 November 2076** partial: 52% — 13:50 EET (partial, saros 124)
- **3 September 2081** partial: 99% — 11:00 EEST (total, saros 136)
- **21 April 2088** partial: 87% — 13:59 EEST (total, saros 130)
- **18 February 2091** partial: 22% — 12:03 EET (partial, saros 122)
- **23 July 2093** partial: 87% — 16:24 EEST (annular, saros 147)

## Four shadows over Romania

The only eclipses that put, or will put, the shadow on Romania between 1900 and 2100: two totals in the past, one total and one annular to come. Pick a town on the map.

- Total, 15 February 1961
- Total, 11 August 1999
- Total, 3 September 2081 (predicted)
- Annular, 23 July 2093 (predicted)

Nine towns stand right on a shadow's edge: there, the Moon's mountains can move the line by a few kilometres. And the 2081 and 2093 tracks are predictions: they assume Earth's spin keeps following the curve measured so far.

## Why the Moon can cover the Sun

The Sun is about 400 times wider than the Moon and stands about 390 times further away. The two numbers match almost exactly, so the Moon's disc covers the Sun's disc almost exactly.

The match is not perfect, which is why eclipses differ. When the Moon is nearer Earth, its disc outgrows the Sun's and totality follows; when it is further, a ring of Sun stays visible around it: an annular eclipse. In 1999 the ratio was 1.014, a margin of one and a half percent.

Eclipses return in families. One Saros lasts 18 years and 11 days: after that the Moon and the Sun line up the same way, but Earth has turned another 120 degrees, so the shadow falls on another meridian. The 1999 eclipse and the American one of 2017 are sisters, of Saros 145; the 2027 eclipse over Egypt and the 2081 eclipse over Oltenia are sisters too, of Saros 136, three generations apart.

### How to watch safely

Only totality is safe for the naked eye, while the Moon covers the Sun's disc completely. A partial or annular eclipse must never be watched without proper equipment, and sunglasses are not proper equipment. [The full rules, at NASA](https://eclipse.gsfc.nasa.gov/SEhelp/safety2.html)

## Eight true-or-false questions

- **Bucharest saw two total eclipses in the 20th century.** True. On 15 February 1961, for 129 seconds, and on 11 August 1999, for 142 seconds. None since, and none to 2100. ([NASA Canon + this page's computation](https://eclipse.gsfc.nasa.gov/eclipse_besselian_from_mysqldump2.csv))
- **In 1999 the Moon looked bigger than the Sun.** True. The ratio of apparent diameters was 1.014: the Moon covered the Sun with a one-and-a-half percent margin. ([NASA, 1999 path table](https://eclipse.gsfc.nasa.gov/SEpath/SEpath1951/SE1999Aug11Tpath.html))
- **The August 2026 total eclipse was visible from Romania too.** False. At maximum the Sun was already below the horizon everywhere in the country: the eclipse set with it. Nothing was visible from Romania. ([NASA Canon + this page's computation](https://eclipse.gsfc.nasa.gov/eclipse_besselian_from_mysqldump2.csv))
- **A partial eclipse can be watched through sunglasses.** False. It is never safe to look at a partial or annular eclipse without proper equipment. Only totality is safe for the naked eye. ([B. R. Chou, NASA Eye Safety](https://eclipse.gsfc.nasa.gov/SEhelp/safety2.html))
- **Eclipses repeat the same, in the same place, every 18 years.** False. After one Saros of 18 years and 11 days, the path falls about 120 degrees further west. The same place sees it again only after three Saroses: 54 years and 34 days. ([F. Espenak, Eclipses and the Saros, NASA](https://eclipse.gsfc.nasa.gov/SEsaros/SEsaros.html))
- **The 1999 eclipse peaked over Romania.** True. Greatest eclipse fell at 45.076° N, 24.298° E, in the Southern Carpathians, about 6 km from Râmnicu Vâlcea. The longest duration, 2 minutes 23 seconds, was measured there. ([NASA Canon + this page's computation](https://eclipse.gsfc.nasa.gov/eclipse_besselian_from_mysqldump2.csv))
- **Romania's next totality comes on 3 September 2081.** True. The shadow crosses the south: Drobeta-Turnu Severin gets 256 seconds, Craiova, 254 seconds. Bucharest misses it by about 20 km. ([NASA Canon + this page's computation](https://eclipse.gsfc.nasa.gov/SEpath/SEpath2051/SE2081Sep03Tpath.html))
- **Buzău caught no totality in 200 years, though it twice came close.** True. In 1961 it stood at 0.9993 of totality, in 1999 at 0.9904: each time the shadow's edge passed a few kilometres away. ([NASA Canon + this page's computation](https://eclipse.gsfc.nasa.gov/eclipse_besselian_from_mysqldump2.csv))

## Where the numbers come from

The list starts from NASA's Canon of solar eclipses: 454 eclipses between 1900 and 2100, each with date, time, type and Saros series. For every eclipse and every town, a program found the minute the Moon came nearest the Sun in that town's sky, using the JPL DE440 ephemeris and Romanian time with its historical rules. Until July 1931 the clock followed the Sun over Bucharest (UTC+1:44); since then, Eastern European Time, with summer time in some years.

The sums were checked in two places. For Bucharest in 1961 they give the same times the press of the day printed, to the second; for 1999 they give 142 seconds of totality, against 142.5 seconds in NASA's table for the centre line, which passes 3 km from the city centre.

What the sums leave out: the mountains on the Moon's edge. So in nine cases right on a shadow's edge the page says “on the edge” instead of deciding between total and partial. The 2081 and 2093 tracks are predictions too: they assume Earth's spin keeps following the curve measured so far.

The towns are the 42 county capitals, with GeoNames coordinates. Differences from nearby villages run to minutes and a few percent, except on the edges. Only eclipses where the Moon covers at least 1% of the Sun are listed; smaller grazes cannot be seen and are left out.

- Espenak, Meeus. Five Millennium Canon of Solar Eclipses: -1999 to +3000. NASA/GSFC (Besselian table, CSV format). https://eclipse.gsfc.nasa.gov/eclipse_besselian_from_mysqldump2.csv — the list of 454 eclipses in 1900–2100: date, time, type, Saros, magnitude
- Espenak. Five Millennium Catalog of Solar Eclipses. NASA/GSFC. https://eclipse.gsfc.nasa.gov/SEcat5/SEcatalog.html — the Canon's front page and table key
- Espenak. Key to Catalog of Solar Eclipses. NASA/GSFC. https://eclipse.gsfc.nasa.gov/SEcat5/SEcatkey.html — magnitude means the fraction of the Sun's diameter covered
- Espenak. Path of the Total Solar Eclipse of 1999 Aug 11. NASA/GSFC. https://eclipse.gsfc.nasa.gov/SEpath/SEpath1951/SE1999Aug11Tpath.html — shadow limits and centre line, minute by minute; the 02m23s duration
- Espenak. Path of the Total Solar Eclipse of 1961 Feb 15. NASA/GSFC. https://eclipse.gsfc.nasa.gov/SEpath/SEpath1951/SE1961Feb15Tpath.html — shadow limits over southern Romania; the 02m45s duration
- Espenak. Path of the Total Solar Eclipse of 2081 Sep 03. NASA/GSFC. https://eclipse.gsfc.nasa.gov/SEpath/SEpath2051/SE2081Sep03Tpath.html — the predicted path over Oltenia and Muntenia
- Espenak. Path of the Annular Solar Eclipse of 2093 Jul 23. NASA/GSFC. https://eclipse.gsfc.nasa.gov/SEpath/SEpath2051/SE2093Jul23Apath.html — the predicted path over northern and eastern Romania
- Espenak. Eclipses and the Saros. NASA/GSFC. https://eclipse.gsfc.nasa.gov/SEsaros/SEsaros.html — the 18-year-11-day Saros; the 120-degree westward shift
- Chou. Eye Safety During Solar Eclipses. NASA/GSFC (after NASA RP 1383). https://eclipse.gsfc.nasa.gov/SEhelp/safety2.html — only totality is safe for the naked eye
- Williams. Moon Fact Sheet. NASA NSSDC. https://nssdc.gsfc.nasa.gov/planetary/factsheet/moonfact.html — the Moon's radius, 1,737.4 km; semimajor axis, 384,400 km
- Williams. Sun Fact Sheet. NASA NSSDC. https://nssdc.gsfc.nasa.gov/planetary/factsheet/sunfact.html — the Sun's radius, 695,700 km; mean distance, 149.6 million km
- Rhodes. Skyfield: ephemerides and astronomical positions in Python. https://rhodesmill.org/skyfield/ — the library used to compute each town's times and coverage
- Park et al. JPL Planetary and Lunar Ephemeris DE440. NASA/JPL SSD. https://ssd.jpl.nasa.gov/planets/eph_export.html — the Sun and Moon ephemeris used in the computation
- Espenak, Meeus. Polynomial Expressions for Delta T. NASA/GSFC. https://eclipse.gsfc.nasa.gov/SEcat5/deltatpoly.html — the dynamical-minus-universal time difference across the range
- Espenak. Uncertainty in Delta T. NASA/GSFC. https://eclipse.gsfc.nasa.gov/SEcat5/uncertainty.html — why far-future paths are predictions, not measurements
- GeoNames. Geographical database (CC BY 4.0). https://www.geonames.org/ — the coordinates of the 42 county capitals
- IANA. Time Zone Database. https://www.iana.org/time-zones — the Europe/Bucharest zone and daylight-saving rules
- Marius Comper. Eclipsa din 1961 la București: două minute sub nori. https://mariuscomper.uk/eclipsa-1961/ — the sibling page: shared 1961 figures are identical here

Towns: Alba Iulia, Alexandria, Arad, Bacău, Baia Mare, Bistrița, Botoșani, Brașov, Brăila, Bucharest, Buftea, Buzău, Cluj-Napoca, Constanța, Craiova, Călărași, Deva, Drobeta-Turnu Severin, Focșani, Galați, Giurgiu, Iași, Miercurea-Ciuc, Oradea, Piatra Neamț, Pitești, Ploiești, Reșița, Râmnicu Vâlcea, Satu Mare, Sfântu Gheorghe, Sibiu, Slatina, Slobozia, Suceava, Timișoara, Tulcea, Târgoviște, Târgu Jiu, Târgu Mureș, Vaslui, Zalău.
