Steven Dutch, Professor Emeritus, Natural and Applied Sciences, Universityof Wisconsin - Green Bay

It is no accident that we call enormous numbers "astronomical". Astronomersdeal with quantities so vast, distance, mass, and energy, that they tax our imaginationsand mathematical skills. But such vast quantities also provide some of the most excitingideas the human mind can encounter.

We can determine the distances to objects in the Solar System, and to nearby stars, by *Triangulation*.If we observe an object from two stations a known distance apart, we can find the distanceto the object. For objects in the Solar System, we can measure the position of a theobject from different sides of the Earth. Observations of Venus crossing or *transiting*the face of the Sun, or asteroids passing very close to the Earth, were once used. Once weknow the actual distance of any object, Kepler's Third Law allows us to determine thedistances of every object in the Solar System. [Nowadays we can actually bounce radarsignals off nearby planets and measure the distance more accurately this way]

The stars are so far away that they look the same all over the Earth. The farthestapart we can ever make observations at present is from opposite sides of the Earth'sorbit, on opposite sides of an ellipse 186 million miles in diameter. When we photographthe stars six months apart, some of them shift very slightly because of the Earth'smotion. You can demonstrate this shift, or *parallax*, easily. Hold up your fingerat arm's length and alternately open and close each eye. Your finger will appear to moverelative to objects in the background. For the stars, the *parallactic shift* isextremely tiny but it can be measured. The distances to the stars turn out to be so greatthat we need a new unit of distance: the

light year

. The light year is not a unit of time; it is the distance light travels in a year, orabout 6 trillion miles (10 trillion km)

Measuring parallax involves measuring very tiny angles. A degree can be divided into 60min*utes* and each minute into 60 *seconds*. A quarter spans one degree at adistance of 1.5 meters. The Sun and Moon span about half a degree in the sky. A quarterspans one minute at a distance 60 times greater, or 90 meters. Venus at its nearest andJupiter both span about a minute of arc in the sky. A quarter spans one second at adistance of 5 kilometers. The moons of Jupiter appear about one second of arc across.Astronomers routinely measure the positions of objects to about 0.1 second of arc and theHubble Space Telescope, above the distortions produced by Earth's atmosphere, can seedetails that span a few hundredths of a second of arc.

Astronomers also use another distance term, the *parsec*. A parsec is thedistance at which a star would have a parallax of one second of arc, or 3.26 light years.Parsecs are convenient because the distance in parsecs is simply one over the parallax,making it easy to convert the two units.

Hipparcos of Nicaea lived in the second century B.C. and left us the first starcatalog. On August 8, 1989, a satellite named in his honor, the **Hi**gh **P**recision**Par**allax **Co**llecting **S**atellite, was launched by the European SpaceAgency. Despite an engine malfunction that left the satellite in a less desirable orbitthan planned, the spacecraft gathered data on stellar positions and magnitudes until March1993. The data were processed and finally released in catalog form in the summer of 1997.The data are of unprecedented accuracy and have completely revolutionized thedetermination of stellar distances.

Hipparcos worked by observing stars through two telescopes aimed 58 degrees apart. Thelight from the two telescopes was merged into a detector with a fine grid of wires. As thesatellite rotated, different stars passed through the field of view of each telescope andblinked on and off as the stars passed across the grid of wires. These observationsallowed extremely accurate relative positions of the stars to be determined. The relativepositions of all the stars could then be combined into an extremely accurate catalog ofstar positions across the entire sky.

For 118,000 selected stars, Hipparcos measured their parallax accurate to .001 secondof arc. That's the apparent diameter of a quarter at a distance of 5000 kilometers, orputting a quarter in New York and viewing it from San Francisco. It's also the amount thehair on a person a meter away appears to grow in one second. A secondary mission namedTycho measured another million stars to an accuracy of "only" 0.01 second.

Distances in older star catalogs look so authoritative that many people think thedistances to the stars are known very exactly, but this is not true. For parallaxmeasurements made from Earth:

- The nearest star is some 4.2 light years away, and that distance was known to within an accuracy of better than 0.1 light year.
- For a star measured as 20 light years away, the distance estimate is accurate to within a light year or so.
- For stars 50 light years away, the margin of error is perhaps 5 to 10 light years.
- Beyond 100 light years, the parallax is so tiny it cannot be measured accurately, and astronomers resort to more indirect methods to find the distances to the stars. We will examine some of these methods later.

The Hipparcos catalog of star distances has made older distance estimates obsolete andall new astronomical publications will use the Hipparcos figures. Even more accurateobservational systems are in the planning stages. However, until there is a distancemeasuring spacecraft on continuous duty, capable of being aimed at new targets on demand,there will still be earth-based parallax measurements and these will be subject to thesame limitations as the older measurements

The following table shows how Hipparcos has improved our knowledge of stellardistances:

Accuracy Level | Earth-Based Data | Hipparcos Data | ||
---|---|---|---|---|

1 percent | 50 stars | 10 light years | 400 stars | 30 light years |

5 percent | 100 stars | 20 light years | 7,000 stars | 150 light years |

10 percent | 1000 stars | 50 light years | 28,000 stars | 300 light years |

The earth-based data are approximate. Much of the improvement in Hipparcos is due to itsbeing completely automatic, whereas earth-based data were combined by laboriousmeasurements of individual stars. Thus, earth-based data were incomplete and of variableaccuracy.

Films like Star Wars create the impression that travelling from star tostar is only a bit more complex than driving down to the corner store for a loaf of bread.In fact, distances in the Universe are so vast it is hard to comprehend them. Everydayspeeds will not begin to cover the distances in the Universe. For example, walking nonstopat a brisk pace, a person could cover about 75 miles (120 km) in a 24-hour day. If he orshe could keep up the pace indefinitely, it would take about 40 days to cross the3000-mile (5000 km) width of the United States, and about 11 months to travel the 25,000(40,000 km) miles around the Earth. It would take about 8 1/2 years to travel the 235,000miles (380,000 km) to the Moon, and 95 years to cover the distance to Venus at itsclosest.

In an automobile travelling a steady 55 miles (90 km) per hour, it takes about 2 1/2days to cross the United States and about 19 days to circle the Earth. It will still takesix months to travel the distance to the Moon. Venus, however, is 100 times farther awayat its closest, and would be a 54-year trip. We are not even out of the inner Solar Systemyet. At 55 miles an hour, just under half a million miles (800,000 km) a year, the Sun istwo centuries away. If we had started out from the Sun at that speed in 1 A.D. we wouldnow have covered about 960 million miles (1.54 billion km) and would be just beyondSaturn, only a third of the way out of the Solar System.

In a jetliner, at 550 miles (900 km) an hour, we cover distance 10 times as fast as ina car. The United States is 5-1/2 hours wide, and two days will take us around the world.The Moon is three weeks away, Venus over 5 years, the Sun 20. Pluto, the outermost planet,is still 750 years away. The nearest star is still so far away as to be unimaginable: 52million years.

At 18,000 miles (29,000 km) an hour, the Space Shuttle crosses the Unites States in 10minutes and circles the Earth in an hour and a half. In 13 hours it covers the distance tothe Moon, but even so would take two months to cover the distance to Venus and seven tocover the distance to the Sun. Pluto is still over two years away. The nearest star isstill 160,000 years away.

Light, at 186,300 miles (298,000 km) per second, could circle the Earth seven times inone second and reach the Moon in 1-1/2 seconds. The Sun is about 8-1/2 minutes away at thespeed of light. It takes light about 5-1/2 hours to reach Pluto from the Sun, but 4.3years to reach the next star. % Scale Models of the Universe

It is impossible to make a physical model that shows Man, the planets, and the stars onthe same scale. If we make the Earth a quarter (one inch or 2.5 cm in diameter), the Moonbecomes a pea 29 inches (74 cm) away. The Sun is 9 feet (2.7 m) across and 1000 feet (300m) away. Pluto is 7 miles (11 km) away, and the nearest star is still off the real Earth:49,000 miles (79,000 km) away.

If we let the Sun be a quarter, the Earth is a speck 1/100 inch (1/40 cm) in diameterand ten feet (3 m) away. Pluto is more than a football field away: 350 feet (110 m). Andat last we can begin to show stars in our scale model. The nearest one is about 500 miles(900 km) away, and it, too, is the size of a quarter, with a pea-sized companion starabout 200 feet (60 m) away. Placing a single coin in each State capital covers the U.S.with coins more densely than space is filled with stars.

Most of the stars we see with the naked eye at night are within a few hundred lightyears. A few are as far away as 2000 light years, only about 1/50 the diameter of ourGalaxy. We can only use parallax to determine the distances of stars within about 50 lightyears, a realm that bears about the same relationship to our Galaxy as a period bears tothe size of this page. How can astronomers measure the distances of distant stars?

There are many ways to explain how we know the arrangement of the universe, but perhapsthe clearest analogy is the City Lights Analogy. Imagine being on the roof of a tallbuilding with no way to leave. At night you can see lights in all directions. How can youmap your "universe"? For nearby lights, you can use triangulation. By observingfrom different locations on the roof, and measuring changes in the relative positions ofthe lights, you can determine how far away the lights are. Once you know how far away thelights are, you can determine their true brightness. In nearby towns, the lights are alltoo far away to measure their distances by triangulation, but you can recognize lights ofthe same types as those you see nearby. Since you know how bright these nearby lights are,you can calculate how far off the lights in the town are. For distant towns, you cannoteven see individual lights, but you know how far away the nearby towns are, and how muchtotal light the towns emit, so you can use this information to estimate how far thedistant towns are. At even greater distances, only clusters of towns are visible, andfinally only great urban complexes, but at each stage you can use what you already know todiscaover facts about the more distant universe. We build our picture of the Universe inmuch the same way.

Within the range we can measure parallax accurately, there are thousands of stars ofmany spectral types. With the apparent brightness and distances of these stars known, itis possible to determine their absolute magnitudes. Distant stars of the same spectraltypes will probably have the same absolute magnitudes, enabling us to estimate thedistance to the unknown star. For isolated stars, this is virtually the only method ofdetermining distance. When we say that the bright star Deneb is 1600 light years away, forexample, the distance is an estimate based on its brightness and spectral type. Thedistance could easily be 25% larger or smaller. Deneb is still too far away for evenHipparcos to measure accurately.

Another technique for finding the distances of faraway stars involves star clusters. Onphotographs of clusters taken many years apart, the stars move. Some stars will move athigher speeds than others, but the velocity will show a characteristic statisticalpattern. We can also measure the Doppler shift of stars in the cluster to find out howfast the stars are moving toward or away from us. The star cluster has no connection toEarth at all, so the velocity pattern of stars along our line of sight should be the sameas that across our line of sight. Radial velocity can be determined in kilometers persecond. By assuming the range of velocities across our line of sight is the same, we candetermine what angle in the sky corresponds to a known distance and thus determine thedistance to the cluster.

The cluster method works out to a thousand light years or so. Fortunately, within thatdistance are stars that give us a yardstick to distant galaxies. These are the *CepheidVariables*. Cepheids are named for a star in the constellation Cepheus, the first starof this type discovered, but the most famous Cepheid, and also the nearest, is Polaris,the Pole Star. Polaris is about 300 light years away and varies in brightness too slightlyto be obvious to the unaided eye. Before Hipparcos, the distance to Cepheids had to bedetermined indirectly by the cluster method, but now the distances to several have beendetermined directly.

Cepheid variables are yellow-white giant stars that pulsate and vary in brightness, butin a regular way. The brighter the absolute magnitude of a Cepheid, the faster itpulsates. It is easy to measure the period of a Cepheid variable, and armed with thisinformation, we can determine the absolute brightness of the star. Instead of usingdistance and apparent magnitude to find absolute magnitude, we compare the absolute andapparent brightness of the star to determine its distance.

Return to Course Syllabus

Return to Course Notes Index

Return to Professor Dutch's Home Page

*Created 26 March 1998, Last Update 2 April 1998*