Interactive Star Chart

Canopus is the second brightest star in the night sky, and nobody has ever seen it from London. Cloud is not the reason, nor city glare. The star simply never rises there, while over Sydney it climbs 71 degrees high.

That is the whole problem a star chart exists to solve. The stars keep almost exactly the same places relative to one another for many lifetimes; NASA’s education pages say they only seem to move because we watch them from a moving planet. What changes is the watcher. Stand somewhere else and a different half of the sphere is overhead. Wait an hour and the half you have has turned. A chart that is right for you, tonight, at ten, is therefore not a picture of the sky. It is the answer to a small calculation with three inputs: latitude, longitude and time.

Latitude decides which stars you ever get

Start with the one point of the sky that does not turn. The U.S. Naval Observatory’s formula for altitude takes a star’s declination, the observer’s latitude and an hour angle. Put in a declination of 90 degrees, the celestial pole, and the hour angle drops out: the pole stands as many degrees above the horizon as you are from the equator. Polaris sits within a degree of that point, which makes its height a quick reading of your latitude.

Everything else circles the pole once a day, and two consequences follow. Stars close enough to the raised pole never touch the horizon. Stars equally close to the hidden pole never clear it. From latitude 51.5 north, Canopus, at declination 52.7 south, reaches its highest point 14 degrees below the horizon, so it is not a matter of waiting for the right season. The outside view of the tool draws both limits as circles on the sphere, with the flat plate of your horizon slicing between them.

At the equator those circles shrink to nothing and every star rises. At a pole they meet, and nothing rises or sets at all.

The clock the stars keep

The Earth turns once against the stars in 23 hours 56 minutes and 4.1 seconds, as Las Cumbres Observatory puts it, about four minutes short of the day on your watch. The shortfall exists because the Earth also moves along its orbit, so it must turn a little extra to bring the Sun back to the same place.

Four minutes sounds like nothing. It is the reason the sky has seasons. A star that rises at ten tonight rises at four minutes to ten tomorrow, two hours earlier in a month, and at the same time again only after a full year. NASA’s description is that the stars appear each night slightly west of where they were the night before. Orion belongs to January evenings for no deeper reason than that.

Astronomers keep a second clock that runs at the stars’ pace. The Naval Observatory defines sidereal time as the hour angle of the vernal equinox, the zero point of right ascension, and gives a short formula for it in which each ordinary hour counts as 1.0027379 sidereal hours. Add your longitude, fifteen degrees to the hour, and you have local sidereal time. Its use is immediate: a star stands highest, on the meridian, when the local sidereal time equals its right ascension.

Here is a case to try. Choose London in the tool, set the clock to UTC, the date to 4 October 2026 and the time to 22:00. Local sidereal time reads 22h 54m. Fomalhaut, right ascension 22h 58m, is therefore four minutes from the meridian, and the tool finds it 9 degrees up, toward S. Vega crossed more than four hours ago and is 45 degrees up, toward W. Polaris is 52 degrees up, the latitude again. With stars to magnitude 5.0, the count above the horizon is 764.

Now set the next evening, 5 October, at 21:56. The sidereal clock reads 22h 54m once more, and Vega is back at exactly the same height and bearing. Leave the time at 22:00 instead and it has slipped lower.

What the dots on the chart are worth

Star brightness is still measured on a scale that goes back to the ancient Greeks. Stars visible to the naked eye were sorted into six magnitudes, the brightest called first magnitude and the faintest sixth. That is why a smaller number means a brighter star. Ptolemy made the scheme popular, and it is generally believed to have begun with Hipparchus, who died around 120 BC, though his own catalogue is lost and the point cannot be proved. The scale was never turned around; it was made exact.

In 1856 Norman Pogson defined a first-magnitude star as one 100 times as bright as a sixth-magnitude star. Five equal steps multiplying to 100 make each step the fifth root of 100, or about 2.512 times. Made exact, the scale had to run past zero for the brightest stars: in the catalog behind this chart Vega is magnitude 0.03 and Sirius is -1.44. That catalog stops at magnitude 6.5 and holds 8,874 stars, reduced from a compilation built on the European Space Agency’s Hipparcos satellite, which measured 118,218 stars with the highest precision between 1989 and 1993.

Color is a number too. Astronomers measure a star’s magnitude through a blue filter and a green-yellow one and subtract, which gives the color index B-V. Vega is the traditional zero point. The lower the index, the bluer and hotter the star; the higher, the redder and cooler. The Sun is about +0.66, and in this chart’s catalog Spica is -0.24 and Betelgeuse +1.50. The tool tints each star by that index, as an illustration rather than a photograph. The eye does less: once the rod cells take over in the dark, colored spots appear colorless, because only the cone pathways encode color.

Eighty-eight areas, and lines nobody owns

A constellation is officially not a pattern. In 1922 the International Astronomical Union agreed on a list of 88 with three-letter abbreviations, and then asked Eugene Delporte to draw borders. His boundaries, approved in 1928 and published in 1930, run along lines of right ascension and declination, and they tile the entire sky. When an astronomer says a star is in Orion, the IAU explains, the meaning is only that it lies inside that fence. This chart assigns each of its stars to a constellation by testing it against those borders.

The stick figures are another matter. The IAU states that no line figures are formally defined, so the same constellation can be drawn several ways. The figures used here follow the charts the IAU made with Sky & Telescope, whose patterns were influenced by H. A. Rey. Treat them as a memory aid. The stars joined by a line are usually unrelated, at very different distances.

The names are older still. Johann Bayer’s atlas of 1603 labeled the stars of each constellation with Greek letters in roughly descending brightness, followed by the possessive form of the constellation’s Latin name, which is how Vega is also Alpha Lyrae. Roughly is the right word: Pollux is brighter than Castor, yet it is Beta Geminorum.

Wanderers, a drifting grid, and what the chart leaves out

Five bright points refuse to stay put. The word planet comes from the Greek for wanderer, the IAU notes, because planets were first defined as objects that move against the fixed stars. A chart cannot store them; it has to compute them. This one uses the approximate orbital elements published by the Jet Propulsion Laboratory, which are fitted to the years 1800 to 2050 and are not valid outside them, and it treats the Moon with a compact series described by Paul Schlyter. The Moon needs one more step, since it is close enough that its position from your doorstep can differ by more than a degree from its position seen from the center of the Earth.

Even the grid drifts. The Earth’s axis wobbles like a top, a cycle NASA gives as about 25,771.5 years, and the equinox slides along with it at roughly 50 arcseconds a year by the Naval Observatory’s formulas. Catalog positions are quoted for the year 2000, so the chart turns them to the chosen date before drawing anything.

How good is the result? Each body was compared with the Naval Observatory’s data services and with JPL’s Horizons system at hundreds of moments. Stars agree within 0.04 degree this year, the Sun within 0.02, the Moon within 0.1. The planets are looser. Saturn, the worst, was off by as much as 0.23 degree, so the chart promises no better than 0.25 degree; JPL’s own table allows these elements 600 arcseconds of error for that planet. One real weakness grows with distance from the year 2000: the stars’ own slow motions are left out, so the fastest bright star is 0.2 degree out of place by 1805.

Some things are not attempted at all. The air bends light, raising anything on the horizon by about 34 arcminutes, about the width of the Moon, and the chart ignores it, so risings come a few minutes earlier than drawn. There are no hills, no haze, no city glow. The word Night appears when the Sun is 18 degrees down, the Naval Observatory’s line for the end of astronomical twilight, and says nothing about your weather.

The printed chart is the part to take outside. It puts the zenith at the center and the horizon on the rim, in a projection where distance from the center is true, so a star halfway out is halfway up. East is on the left, which looks wrong on a desk and right overhead. Hold it over your head, look up at it, and turn it until its northern edge faces north, which is how NASA’s star finder instructions line up any map of the sky overhead.

Read it by a red light. The eye takes roughly 20 to 30 minutes to adapt fully from bright light to darkness, and the rod cells that do the seeing at night are insensitive to long wavelengths, which is why working by red light has become common practice wherever dark-adapted eyes are needed, from cockpits to submarines.

Then look up. In four minutes the sky will have turned one degree past the paper, and the calculation starts again.