NASA SPENT $2.1 BILLION to escape from poor atmospheric seeing; that's what it cost to put the Hubble Space Telescope above the atmosphere. Backyard observers on a smaller budget, however, need not despair of improving their fuzzy, shimmering views. You can avoid the worst effects of atmospheric turbulence by understanding its nature and learning a few tricks.

Viewed at high power from the bottom of our ocean of air, a star is a living thing. It jumps, quivers, and ripples tirelessly, or swells into a ball of steady fuzz. Rare is the night (at most sites) when any telescope, no matter how large its aperture or perfect its optics, can resolve details finer than 1 arcsecond. More typical at ordinary locations is 2- or 3-arcsecond seeing, or worse.

It's not hard to understand why. The usual definition of a "good" telescope is one that keeps all parts of a light wave entering it nicely squared up to within ¼ wavelength accuracy by the time the wave comes to focus. But that same light wave, in traversing just six feet of air inside a telescope tube, is retarded by about 800 wavelengths compared to where it would be if the telescope contained a vacuum. Clearly the air is an important optical element, and it had better affect every part of a light wave equally. If the refractive power of the air down one part of the telescope tube differs from the rest by more than just one part in 3,200, the ¼-wave tolerance will be breached. Such a change results from a temperature difference of just 0.1° Celsius.

Add the miles of air that the light wave traverses before it even gets to the telescope, and it's a wonder that we can see any detail beyond the atmosphere at all.

The air's light-bending power, or refractive index, depends on its density and therefore its temperature. Wherever air masses with different temperatures meet, the boundary layer between them breaks up into swirling ripples and eddies that act as weak lenses. You can see this where hot air from a fire or a sunbaked road mixes with cooler air; the heat waves are astronomers' poor seeing writ large. Our windy, weather-ridden atmosphere is almost always full of slight temperature irregularities, and when you look through a telescope you see their effect magnified.

Look at Marc Coco's picture of the setting Sun (taken through an 8-inch telescope) at the top of this page; it illustrates atmospheric dispersion, distortion, absorption, reddening, and refraction all at once. The Sun's limb is distorted into horizontal ridges by layers of different-temperature air. A fragmentary "green fringe" floats on top, the result of dispersion placing the blue and green images of an astronomical object slightly higher than the yellow and red. The Sun is flattened to an oval shape by the greater atmospheric refraction at the horizon, where the thick air also absorbs and reddens the light more.

Much of the problem with atmospheric seeing arises surprisingly close to the telescope, where you can take control of it to reduce it.

Inside the Scope

Seeing problems are often at their worst a fraction of an inch from the objective lens or mirror. If the objective is not at air temperature, it will surround itself with a wavy, irregular, slowly shifting envelope of air slightly warmer or colder than the ambient night. So will every other telescope part. Therefore, give the telescope time to come to equilibrium with its surroundings. Amateurs soon learn that the view sharpens within about a half hour after bringing a telescope outdoors. The full cool-down time for a large, heavy instrument may be much longer. It pays to set up early.

Usually the telescope is too warm, especially if it is stored indoors to prevent destructive dampness from condensing inside it during weather changes. But sometimes the opposite happens. Whenever a telescope begins to collect dew or frost, you know that it has grown colder than the air via radiational cooling. In this case gentle heat not only prevents dew but also keeps the scope closer to the air temperature -- and thus may sharpen its resolution.

"Tube currents" of warm and cool air in a telescope are real performance killers. Reflectors are notorious for tube currents, but closed-tube Schmidt-Cassegrains and refractors can get them too. Any open-ended tube, amateurs these days tend to agree, should be ventilated as well as possible. Suspending a fan behind a reflector's mirror has become a popular way to speed cooling and blow out mixed-temperature air.

It's easy to check whether tube currents trouble your images. Turn a very bright star far out of focus until it's a big, uniform disk of light. Tube currents will show as thin lines of light and shadow slowly looping and curling across the bright disk.

Near the Scope

Some seeing problems arise just a few feet in front of the telescope. Obviously, try to keep your breath and body heat out of the light path. This is one reason to put a cloth shroud around an open-framework tube.

A telescope's immediate surroundings should have low heat capacity so they don't store up the warmth of the day. Grass and shrubbery are better than pavement. The flatter and more uniform the greenery the better. Heated buildings are disasters of poor seeing, especially if you find yourself looking over a chimney.

If you build an observatory, make it of thin materials that cool quickly: plywood or sheet metal, not masonry. Paint it white or a very light color to reflect solar heat, and ventilate it very well. A thick rug belongs on the floor. A roll-off roof that opens the whole room to the sky provides quicker cooling and better seeing than a dome with its chimneylike slit. If you insist on a dome, it's a good idea to install a large fan in one wall to suck air down through the slit past the telescope, just as professional observatories do. It's widely considered a poor idea to attach an observatory to a heated house unless you resign yourself to low-power work. At least put it on the upwind side.

Much poor seeing hugs the ground, so an elevated observing platform is a good idea if you can manage it. A scope is likely to show the stars and planets more sharply if you can get it up just a few feet closer to them.

High-Altitude Seeing

Now we come to the unavoidable heart of the problem. There's not much you can do about the air thousands of feet up. But you may be able to predict when and where it will be smoothest.

Telescope users recognize two types of seeing: "slow" and "fast." Slow seeing makes stars and planets wiggle and wobble; fast seeing turns them into hazy balls that hardly move. You can look right through slow seeing to see sharp details as they dance around, because the eye does a wonderful job of following a moving object. But fast seeing outraces the eye's response time.

An old piece of amateur folklore is that you can judge the seeing with the naked eye by checking how much stars twinkle. This often really does work. Most of the turbulence responsible for twinkling originates fairly near the ground, as does much poor seeing. But high-altitude fast seeing escapes this test. If the star is scintillating faster than your eye's time resolution (about 0.1 second), it will appear to shine steadily even if a telescope shows it as a hazy fuzzball.

Astronomers often talk of "seeing cells," air-eddy lenses ranging in size from millimeters to a few meters wide that swarm through the sky. These eddies originate wherever air masses rub past each other -- either horizontally in winds, vertically by convection, or both. Sometimes, when watching an extended object like the Moon or a planet, you can focus on a horizontal layer of "shear turbulence" a few thousand feet high. The ripples sharpen up when you turn the focuser slightly to the outside of infinity focus (eyepiece farther from the objective). This is the signature of an inversion layer in which a mass of warm air flows across cooler air below. The actual temperature difference may be very slight.

Large or slow-moving eddies cause slow seeing, but they don't stay large forever. No matter what size the eddies are when they originate, they break up into smaller and smaller ones. When these become as small as roughly millimeter-scale, they finally die out and dissipate their energy as heat via the air's fluid friction (viscosity).

This complex situation belies an often-repeated piece of astronomer's lore: that seeing cells are 10 centimeters (4 inches) in size. In fact they come in all sizes. But cells in this middle range do have an important property: they affect a large telescope more seriously than a small one. If you have a 4-inch scope, 4-inch and larger cells passing through its line of sight will make an image shift around while staying relatively intact. The same cells passing in front of a 12-inch aperture will superpose multiple images at once, making a fuzzy mess.

This fact has led to another piece of folklore: that when the seeing is bad, a large telescope shows less detail than a small one. Therefore, supposedly, you can improve the view by stopping down a large aperture with a cardboard mask.

Technically there is a bit of truth in this, as mathematical analysis of seeing has shown, but in practical terms the improvement is slight to nonexistent. I have never seen an improvement by stopping down a telescope when the problem was poor seeing. The most that can usually be said is that on a really rotten night, large- and small-aperture views will be equally poor. Even then, if you constrict the aperture you miss the chance for the momentary high-resolution views that the full aperture will provide if the air briefly steadies.

There are unrelated reasons why you may indeed see more in a stopped-down telescope, most of them bad. Maybe your eye was dazzled by a too-bright planet; in that case an eyepiece filter would solve the problem better. Maybe the aperture stop is masking optical errors in a flawed objective, or maybe it's just allowing a mediocre eyepiece to perform better by increasing the telescope's f/ratio. Poor collimation is also less damaging when the f/ratio is increased. On a reflector or Schmidt-Cassegrain, an off-axis mask does give you the advantage of a clear aperture. A clear aperture, mathematical analyses have shown, is slightly less affected by atmospheric turbulence than an obstructed one.

In Search of Steady Air

The seeing quality depends on the weather, but not by simple rules that apply everywhere. Poor seeing does seem more likely shortly before or after a change in the weather, in partial cloudiness, in wind, and in unseasonable cold. Any weather pattern that brings shearing air masses into your sky is bad news. Good seeing, some observers claim, is most likely when a high-pressure system settles in to bring clear skies for several days running. Keep a seeing-versus-weather log for your locality, and you may discover correlations that will become your key to sharp viewing.

Seasonal patterns are more predictable. The seeing is often mediocre in the cold months over the northern United States and southern Canada, when the high-altitude jet stream flows above these latitudes. The very best seeing often comes on still, muggy summer nights when the air is heavy with humidity and the sky looks unpromisingly milky with haze. Some astronomers claim that a blanket of industrial smog steadies the air as effectively as summer humidity -- or rather that it accompanies the same tranquil air masses that are conducive to fine seeing.

Time of night also plays a role, but again there are few universal rules. Right after sunset the seeing is apt to be excellent, so start your planetary observing as soon as you can find a planet in twilight. The seeing is apt to deteriorate before dusk fades out. Some observers find that their seeing improves after midnight; others say it goes to pieces. This depends largely on local topography; observers in valleys might get worse seeing as the night goes on and cold air pools in the valley. Late dawn may be another excellent time.

For observing the Sun (use an astronomer's solar filter!), the best time is early morning before the Sun heats the landscape. The very worst seeing of the 24-hour cycle comes in the afternoon.

Geography is critical. Smooth, laminar airflow is the ideal sought by observatory siting committees worldwide. The best sites on Earth are mountaintops facing into prevailing winds that have crossed thousands of miles of flat, cool ocean. You don't want to be downwind of a mountain; the airstream breaks up into turbulent swirls after crossing the peak. Nor do you want to be downwind of varied terrain that absorbs solar heat differently from one spot to the next. Flat, uniform plains or gently rolling hills extending far upwind can be almost as good as an ocean for providing laminar airflow. You may learn to predict which wind direction brings you the smoothest air.

One easy countermeasure when observing bright objects like the Moon and planets is to use a color filter. Different colors seem to shimmer out of phase with each other in the seeing (the reason stars twinkle in colors), and in a telescope this contributes to the general fuzzing up. The blue image of a planet may align with the yellow image one instant and separate from it the next. If you isolate just the yellow light, for instance, the planet will often appear to quiet down noticeably, at least in a small aperture.

A color filter is especially useful when you're aiming at altitudes lower than 45° above the horizon. The seeing is always worse at low altitudes because you're looking through more air. In addition, you face more atmospheric dispersion. This is the smearing out of a celestial image into a short spectrum with blue on top and red on the bottom. Even as high as 60° up, the blue component of an image appears 0.9" above the red. The difference is 1.5" at 45°, 2.5" at 30°, and 5" at 15°. Your eye is fairly insensitive to light at the extreme ends of the spectrum, so dispersion really doesn't look quite as bad as this. Still, filtering out all but one color in a swarm of chromatic aberration will sharpen your view. In the summer of 1994 I found a yellow or orange filter invaluable for following the comet-crash spots on Jupiter as the planet sank each evening toward the horizon.

Atmospheric problems get worse the lower you look. A star 15° above the horizon will be enlarged twice as much by atmospheric turbulence as one at the zenith, regardless of whether the seeing is good (defined here as 1" star images overhead) or poor (4"). Atmospheric dispersion elongates a star into a colorful little spectrum; at very low altitudes this overtakes even poor seeing as a cause of blurry images. Click on image for larger view. Courtesy Andrew T. Young.

Mostly, though, beating the seeing is just a matter of patience. Keep watching, and intermittent good moments may surprise you. One reason why experienced observers see more on the planets than beginners is that they simply watch longer, ignoring all but the steadiest moments. Moreover, the seeing can change as radically from minute to minute as it does from second to second. When that perfect minute comes along, the dedicated observer is the one most likely to be there at the eyepiece to catch it.

A Scale of Seeing

Amateurs have long recorded the seeing quality in their observing logbooks on a rather subjective scale of 1 to 10, with 1 hopeless and 10 perfect. People's ideas of what the numbers mean are likely to differ. In the interest of uniformity, here is the scale in its early form as described by Harvard Observatory's William H. Pickering (1858-1938). Pickering used a 5-inch refractor. His comments about diffraction disks and rings will have to be modified for larger or smaller instruments, but they're a starting point:

1. Star image is usually about twice the diameter of the third diffraction ring if the ring could be seen; star image 13" in diameter.

2. Image occasionally twice the diameter of the third ring (13").

3. Image about the same diameter as the third ring (6.7"), and brighter at the center.

4. The central Airy diffraction disk often visible; arcs of diffraction rings sometimes seen on brighter stars.

5. Airy disk always visible; arcs frequently seen on brighter stars.

6. Airy disk always visible; short arcs constantly seen.

7. Disk sometimes sharply defined; diffraction rings seen as long arcs or complete circles.

8. Disk always sharply defined; rings seen as long arcs or complete circles, but always in motion.

9. The inner diffraction ring is stationary. Outer rings momentarily stationary.

10. The complete diffraction pattern is stationary.

On this scale 1 to 3 is considered very bad, 4 to 5 poor, 6 to 7 good, and 8 to 10 excellent.

Alan MacRobert is an associate editor of Sky & Telescope magazine and an avid backyard astronomer.