Are custom star maps accurate?

Measured, not promised

How close to the real sky it is

We compare the chart engine with astropy for thousands of instants between 1900 and 2100, and the check fails if any figure goes over its limit. For scale, 0.15° is under a millimetre at the horizon of a 24x36 poster.

Star positions

ESA Hipparcos catalogue, corrected for precession to your date

Precession matches astropy to under 0.001°

Planets

Computed from their orbits for the exact minute

Within 0.15° of astropy for every date 1900–2100

Moon phase

Sun–Moon angle at that minute

Within 0.1°, about 12 minutes of the Moon’s motion

Local time

Time zone and daylight-saving history looked up from the place

Historical time-zone and daylight-saving rules for that place

Left out on purpose

Nutation, aberration and the stars’ own motion

Under 0.01° for 95 % of stars; 0.16° for the fastest

What we leave out, and why: nutation and aberration each shift the sky by under 0.006°. Stars also drift across the sky on their own, but slowly: over 1900–2100, 95 % of the stars on the chart move less than 0.01°. The fastest, 61 Cygni, moves 0.16° at the far ends of that range, about a millimetre on a 24x36 print. Atmospheric refraction, which lifts stars near the horizon by about half a degree, depends on that night’s weather, so like every printed star chart we show geometric positions.

Questions

Star map accuracy: questions

Are custom star maps accurate?

A properly computed one is. Ours places each star from the ESA Hipparcos catalogue, corrects it for the slow wobble of Earth’s axis to your date, and turns it into your local sky using the exact time and place. The planets land within 0.15° of astropy, the software professional astronomers use, which is under a millimetre on a 24x36 print.

Is a star map the real sky or an illustration?

It should be the real sky, computed for one place and one minute. Some decorative prints use a generic star pattern that is the same for every order. A computed map changes when you change the date, time or city, and the constellations sit where an observer standing there would have seen them.

Why does the time of day matter?

The sky turns about 15° every hour as Earth rotates, so a chart for 9 pm and one for midnight on the same date show visibly different stars rising and setting. If you don’t know the time, choose a meaningful evening hour; the stars overhead will still be the ones of that night and season.

How many stars are on the map?

Every star down to magnitude 5.5, which is close to what the eye picks out under a dark country sky. That is 2,851 stars across the whole sky; about half are above the horizon at any moment, so a chart shows roughly that many. Fainter stars are left out because nobody could have seen them without binoculars.

Can I check my star map myself?

Yes. Open the free Stellarium web app, set the same place, date and time, and look straight up. The constellations near the middle of your chart should be the ones overhead in Stellarium, and the bright planets should match. Near the edge, remember that our chart shows the whole sky down to the horizon in every direction.

Why don’t you correct for atmospheric refraction?

Refraction lifts objects near the horizon by about half a degree, depending on the weather that night, which nobody can know for a past date. A star chart traditionally shows geometric positions, as ours does, so a star on the rim may have appeared slightly higher to someone standing there.