Monday 2 August 2027
Sources and methods
Where the data come from, how the predictions are computed, how they were checked and what their limits are.
Computation method
- Local circumstances are computed in the browser and at build time from NASA's polynomial Besselian elements, with the method of the Explanatory Supplement to the Astronomical Almanac (1992) and of J. Meeus (Elements of Solar Eclipses, 1989). The instants of contact are found by root finding with a precision better than 0.01 s; the maximum by Newton iteration.
- ΔT = TT − UT is set to 69.2 s (measured value about 69.1 s in 2025, expected to remain within about ±1 s in 2027). NASA used 76.0 s. A 1 s error on ΔT shifts the path by about 0.4 km in longitude and changes local times by less than 1 s.
- Coordinates are geodetic WGS 84 (EPSG:4326). The observer is placed at sea level. The Sun’s altitude is geometric (without refraction, which is below 1′ at these altitudes). Times are given in UTC and in Tunisian time (UTC+1, no daylight saving time).
- The path of totality, the central line and the isolines on the map are computed from the same elements, meridian by meridian (every 0.02° of longitude over Tunisia), so the map, the calculator and the tables are consistent with one another.
- Magnitude is the fraction of the Sun’s diameter covered by the Moon at maximum; obscuration is the fraction of the Sun’s area covered. For a partial eclipse the displayed values are truncated, never rounded up, so that a very deep partial eclipse can never appear as 100 %.
Validation 1 — comparison with NASA’s path table
For each line of NASA’s table over Tunisia and its surroundings, the engine (with NASA’s ΔT of 76 s) computes the duration on NASA’s central-line point and the latitude of the limits on NASA’s meridians. Largest differences: 0.16 s on the duration and 0.22 km on the limits (NASA’s coordinates are rounded to 0.1′ ≈ 0.18 km).
| UT | Central line | Duration NASA | Duration engine | Δ north limit (km) | Δ south limit (km) |
|---|---|---|---|---|---|
| 09:00 | 35.477° N, 3.818° E | 322.4 s | 322.3 s | -0.22 | -0.15 |
| 09:02 | 35.360° N, 5.105° E | 326.5 s | 326.4 s | -0.15 | -0.08 |
| 09:04 | 35.225° N, 6.350° E | 330.5 s | 330.4 s | -0.10 | -0.02 |
| 09:06 | 35.073° N, 7.555° E | 334.3 s | 334.2 s | -0.15 | -0.08 |
| 09:08 | 34.905° N, 8.725° E | 338.0 s | 337.9 s | -0.02 | -0.14 |
| 09:10 | 34.723° N, 9.858° E | 341.5 s | 341.4 s | -0.07 | -0.08 |
| 09:12 | 34.527° N, 10.960° E | 344.8 s | 344.7 s | 0.07 | -0.02 |
| 09:14 | 34.318° N, 12.032° E | 348.0 s | 347.9 s | -0.00 | -0.04 |
| 09:16 | 34.097° N, 13.075° E | 351.1 s | 351.0 s | 0.11 | 0.06 |
| 09:18 | 33.863° N, 14.090° E | 354.0 s | 353.9 s | 0.17 | 0.02 |
| 09:20 | 33.618° N, 15.080° E | 356.8 s | 356.6 s | 0.17 | 0.03 |
Validation 2 — comparison with an independent library
Astronomy Engine (D. Cross) uses its own ephemerides. With the same ΔT (69.2 s), the differences below are engine − Astronomy Engine, in seconds. The partial-phase contacts and the maximum agree within about 2–3 s. For totality, the two engines agree on the northern side of the path (Kairouan, Mahdia: same durations), but Astronomy Engine places the southern limit about 2–4 km further south: its totality is longer in the south (Gabès, Djerba, Tozeur) and it classifies as total two places that are just outside NASA’s path (Kébili 3 km and Matmata 2 km outside). This is consistent with its larger lunar radius combined with small differences in the lunar ephemeris. It illustrates the real uncertainty near the limits: this site flags every location within 5 km of a limit. These differences are documented rather than hidden.
| Town | Engine | Astronomy Engine | Limit (km) | C1 | C2 | Max | C3 | C4 | Totality engine | Totality AE |
|---|---|---|---|---|---|---|---|---|---|---|
| Sfax | Total eclipse | Total eclipse | 106.8 | -2.0 | +1.3 | -0.5 | -2.1 | +1.4 | 5 min 39 s | 5 min 43 s |
| Sidi Bouzid | Total eclipse | Total eclipse | 97.3 | -2.0 | +1.1 | +0.4 | -2.0 | +1.4 | 5 min 31 s | 5 min 34 s |
| Gafsa | Total eclipse | Total eclipse | 73.7 | -2.0 | +2.9 | -0.4 | -3.8 | +1.3 | 5 min 09 s | 5 min 16 s |
| Gabès | Total eclipse | Total eclipse | 38.1 | -2.0 | +4.6 | -0.5 | -5.5 | +1.3 | 4 min 06 s | 4 min 16 s |
| Houmt Souk (Djerba) | Total eclipse | Total eclipse | 53.2 | -2.0 | +3.7 | -0.4 | -4.6 | +1.3 | 4 min 42 s | 4 min 50 s |
| El Jem | Total eclipse | Total eclipse | 45.4 | -2.1 | +0.3 | -0.6 | -1.1 | +1.4 | 4 min 22 s | 4 min 24 s |
| Kairouan | Total eclipse | Total eclipse | 15.2 | -2.1 | -0.4 | -0.4 | -0.4 | +1.4 | 2 min 41 s | 2 min 41 s |
| Mahdia | Total eclipse | Total eclipse | 15.8 | -2.1 | -0.4 | -0.4 | -0.4 | +1.4 | 2 min 45 s | 2 min 45 s |
| Tozeur | Total eclipse | Total eclipse | 6.3 | -1.9 | +11.6 | -0.5 | -12.6 | +1.3 | 1 min 46 s | 2 min 10 s |
| Kebili | Partial eclipse | Total eclipse | −3.1 | -2.5 | — | -0.4 | — | +1.3 | — | 0 s |
| Matmata | Partial eclipse | Total eclipse | −1.9 | -1.9 | — | -0.4 | — | +1.3 | — | 49 s |
| Sousse | Partial eclipse | Partial eclipse | −11.5 | -2.1 | — | -0.4 | — | +1.4 | — | — |
| Monastir | Partial eclipse | Partial eclipse | −9.9 | -2.1 | — | -0.4 | — | +1.4 | — | — |
| Tunis | Partial eclipse | Partial eclipse | −113.0 | -2.1 | — | -0.4 | — | +1.5 | — | — |
| Bizerte | Partial eclipse | Partial eclipse | −158.3 | -2.2 | — | -0.4 | — | +1.5 | — | — |
| Tataouine | Partial eclipse | Partial eclipse | −60.6 | -1.9 | — | -0.4 | — | +1.3 | — | — |
Validation 3 — published values
Durations published for Tunisian towns by an independent eclipse guide (nationaleclipse.com, method not documented): Sfax 5 min 39 s, Sidi Bouzid 5 min 31 s, Gafsa about 5 min 10 s, Gabès about 4 min 08 s, El Jem 4 min 25 s, Kairouan about 2 min 45 s. The values of this site differ by 0 to 4 s, the largest differences being near the limits.
Known limitations
- Lunar limb profile: mountains and valleys on the edge of the Moon shift the limits by up to 1–2 km and change durations near the limits by several seconds. Within about 5 km of a limit, results are flagged.
- Elevation: computations at sea level. At 500 m altitude, the limits move by a few hundred metres; the effect on times is below one second.
- Towns are represented by a single point (town centre).
- ΔT for 2027 is an estimate (±1 s).
- The climate statistics come from a model reanalysis (ERA5) and do not replace observations or forecasts.
Source register
| Organisation | Document | Use | Limits | Licence | Consulted |
|---|---|---|---|---|---|
| NASA GSFC — F. Espenak | Besselian Elements — Total Solar Eclipse of 2027 August 02 | Polynomial Besselian elements (x, y, d, l1, l2, μ, tan f1, tan f2), basis of every computation on the site. | VSOP87/ELP2000-82 ephemerides; ΔT = 76.0 s extrapolated; smooth lunar limb (limits accurate to about 1–2 km). | "Eclipse Predictions by Fred Espenak, NASA's GSFC" (free reproduction with acknowledgement) | 25 October 2026 |
| NASA GSFC — F. Espenak | Path of the Total Solar Eclipse of 2027 Aug 02 | Reference table (limits, central line, durations every 2 minutes) used to validate the engine. | Coordinates rounded to 0.1′; same ΔT and lunar model as the elements. | NASA — acknowledgement required | 25 October 2026 |
| Astronomy Engine — D. Cross | astronomy-engine 2.1.19 (npm) | Independent cross-check of contact times, obscuration and Sun position; positions of planets during totality. | Uses a larger mean lunar radius: totality a few seconds longer, slightly wider path. | MIT | 25 October 2026 |
| geoBoundaries (William & Mary) — OpenStreetMap | geoBoundaries gbOpen TUN ADM1 (commit 9469f09) | Boundaries of the 24 governorates (simplified with mapshaper, 6 % of vertices kept). | Year represented: 2017; generalised geometry, not suitable for legal or cadastral use. | ODbL 1.0 — © OpenStreetMap contributors | 25 October 2026 |
| Natural Earth | Admin 0 – Countries 1:10m | Outline of neighbouring countries on the static map of the home page. | Strongly simplified for display only. | Public domain | 25 October 2026 |
| GeoNames | TN.zip and alternatenames/TN.zip | Coordinates and population of 62 towns; Arabic and French names reviewed by hand. | One point per town (town centre); a town can extend several kilometres around it. | CC BY 4.0 | 25 October 2026 |
| Copernicus C3S / ECMWF — Open-Meteo | ERA5 hourly reanalysis via the Open-Meteo Historical Weather API | Cloud cover and temperature at 09:00 UTC, 26 July–9 August, 2000–2024, for each destination. | Model reanalysis on a ~25–30 km grid; climate statistics, not observations at the site. | CC BY 4.0 | 25 October 2026 |
| Eclipsophile — J. Anderson | Total Solar Eclipse 2027 August 2 (weather prospects) | Qualitative comparison with satellite cloud climatology (MODIS) and aerosol information. | Regional figures; no site-level numbers reproduced here. | © Eclipsophile — cited, not reproduced | 25 October 2026 |
| NASA Science / American Astronomical Society | Eclipse safety; How to view a solar eclipse safely | Basis of the safety recommendations (ISO 12312-2, filters in front of optics, totality only). | — | Cited | 25 October 2026 |
| UNESCO World Heritage Centre | World Heritage List — Tunisia | Heritage sites mentioned in the destination sheets. | — | Cited | 25 October 2026 |
| OpenFreeMap | Vector tiles and "dark" style | Background map of the interactive map (no API key). | Free community service without guaranteed availability; the eclipse layers keep working without it. | © OpenMapTiles, © OpenStreetMap contributors (ODbL) | 25 October 2026 |