Twilight vs Light Pollution

Zenith sky brightness (V-band / SQM, mag/arcsec²) — how residual summer twilight compares with your site's artificial skyglow.

Location

London, United Kingdom — 51.51°N, 0.13°W — Europe/London

Night of

Evening date; the night runs into the next morning. Times shown in the location's local clock.

Your site's light pollution

Site sky: 21.05 mag/arcsec² ≈ Bortle 4
Model settings
Negative offset = brighter twilight. Sea-level / hazy sites run ≈ −0.25 to −0.3 mag vs the Paranal-calibrated curve (Patat et al. 2006). Natural background 21.75 is the fit's asymptote; true value varies 21.7–22.0 with the solar cycle.

Tonight's verdict

Sun at solar midnight
-17.3°
at 01:05
Twilight-only sky at midnight
21.73 mag/as²
≈ a Bortle 2 sky
Crossover sun altitude
-14.3°
twilight = your LP below this
LP-limited window
23:31 – 02:41
3.2 h
Astronomical darkness
none
sun < −18°
Moon
9% waxing
below horizon at midnight · sets 22:20
Best combined sky tonight
21.04 mag/as²
≈ Bortle 4 at 01:05
Your light pollution is the limit for 3.2 h (23:31 – 02:41), while the Sun sits below -14.3°. Outside that window residual twilight dominates. Best sky ≈ 21.04 mag/as².

Night curve — 2026-07-16 → morning, London, United Kingdom

Twilight + natural sky Light pollution + natural sky Moonlight + natural sky Combined LP-limited (“effectively dark”) Astronomical darkness

Season overview — 2026, London, United Kingdom

Twilight + natural Moonlight + natural Light pollution + natural Combined “Effectively dark” dates Astronomical darkness
One point per night. Combined and moon are taken at the darkest combined moment of the night (the moment dodges the Moon when it can — moon humps show what's unavoidable). Twilight is the night's twilight floor (solar midnight), so on bright-moon nights the white line can sit above both floors. Green strip: “effectively dark” dates — the twilight floor drops below your LP. Blue strip: astronomical darkness occurs. Hover for details; click to jump to a night.
Model notes & sources

Twilight curve. Zenith V/SQM brightness vs Sun altitude a: for 0° ≥ a ≥ −12°, m = 6.7489 − 1.057·a; for −12° ≥ a ≥ −17.3°, m = −0.0744a² − 2.5768a − 0.5845 (blended near −12°, capped at the 21.73 vertex). Empirical fit by Han Kleijn (hnsky.org) to SQM measurements cross-calibrated with ESO-Paranal V-band twilight photometry. Agrees with the Patat, Ugolnikov & Postylyakov (2006, A&A 455, 385) polynomial to ≲0.1 mag over −12°…−15° (e.g. both give ≈21.3 at −15°).

Combining components. All quantities are converted to linear flux f = 10−0.4m. Twilight contribution = f(mtwi) − f(21.75); LP contribution = f(SQMsite) − f(natural), clamped ≥ 0. Combined sky = natural + LP + twilight, converted back to mag/arcsec². The crossover Sun altitude solves twilight contribution = LP contribution; since both share the same natural baseline, this is simply where the twilight-only sky equals your site's SQM.

Bortle ↔ SQM (approx., Bortle 2001 / Wikipedia): 1: 21.76–22.0 · 2: 21.6–21.75 · 3: 21.3–21.6 · 4: 20.8–21.3 · 4.5: 20.3–20.8 · 5: 19.25–20.3 · 6: 18.5–19.25 · 7: 18.0–18.5 · 8–9: <18.

Sun position: NOAA solar algorithm (declination + equation of time), accurate to ~0.1°. No refraction term (negligible for depression angles).

Moonlight: Krisciunas & Schaefer (1991, PASP 103, 1033) model evaluated at the zenith: Moon magnitude m = −12.73 + 0.026|α| + 4×10⁻⁹α⁴ (α = phase angle), scattering f(ρ) = 105.36(1.06 + cos²ρ) + 106.15−ρ/40 with ρ = Moon zenith distance, optical path X(Z) = (1 − 0.96 sin²Z)−1/2, V-band extinction k = 0.20 mag/airmass; nanolambert result converted via B = 34.08 e20.7233−0.92104V. Checks: full Moon at 60° alt → zenith ≈ 17.9, quarter Moon ≈ 20.4 mag/arcsec². Model accuracy is 8–23% (K&S); contribution is zeroed once the Moon is ≳1° below the horizon. Moon position from a truncated Schlyter/Meeus ephemeris (~0.2°, plenty for brightness). Moonlight adds to the flux sum like the other components.

Assumptions / caveats. Zenith values only — twilight is substantially brighter toward the sunward (northern, in summer) horizon, and moonlight toward the Moon's sky quadrant, so targets near those directions fare worse. Clear, aerosol-typical sky assumed; night-to-night scatter is ±0.2–0.4 mag (Patat 2006). Deep-twilight brightness scales with atmospheric pressure: sea-level sites ≈ 0.27 mag brighter than 2600 m Paranal — use the twilight offset. SQM (broad, ~V+B) vs Johnson V differences are within these tolerances.