Taking pictures of exoplanets from the ground is difficult, verging on impossible. Still, astronomers hope to see from the ground a few really giant planets located at moderate distances from their sun.

"From space it looks quite possible but we haven't built the devices to do it yet," says Neville J. Woolf, professor of astronomy at the University of Arizona's Steward Observatory in Tucson. Woolf and his team have been investigating various ways to observe extra-solar planets from Earth and space for the past 14 years, looking to the day when Earth-like planets around other stars are finally discovered.

The search for Earth-like planets is restricted to only a fraction of the Milky Way with current technology. Astronomers are able to look in detail only at the sun's closest neighbors.

"Suppose we send out an instrument to look for Earth-like planets and we give it 70 years of observing time. It will take it about a day to observe a particular star system to find whether it has an Earth-like planet.

"At this rate we might survey perhaps 25,000 stars, a tiny bit of the Milky Way Galaxy, which contains about 10 billion stars. After 70 years, only 2.5 millionth of a percent of our galaxy would have been examined -- and the Milky Way is only one of about 100 billion galaxies," says Woolf.

"There could be 'billions and billions and billions' of Earth-like planets out there, and we would have missed them all. Our only hope is that Earths are extremely common. If so, then looking at the planets around the nearest stars, we will find some," he added.

In the past, astronomers assumed that planets similar in size to Earth were extremely common and Jupiter-like objects were rare. But starting in 1995, discoveries revealed planetary systems very different from the solar system, with giant planets in close-in orbits unlike any body in the solar system. Astronomers began to wonder how rare planetary systems like ours were.

The giant planets in orbits larger than Mercury's had very eccentric orbits so that the Earth-like planets of those systems would have been ejected early on. Woolf however notes that of the four planets with largest orbits (still less than half the size of Jupiter's orbit) two have modest eccentricity. This may be a sign that systems with giant planets in longer period orbits are more like the solar system.

Woolf considers the search for other Earths worthwhile though: "I believe that they are quite common. Every third or fourth star that we look at could have such a planet, and if so, there is a decent chance that we will find signs of life after a reasonable effort."

He adds, "Current planet discoveries are like the tip of an iceberg. For every one of those planets that we have discovered so far there are likely to be perhaps a dozen lower-mass objects. We have not yet found them because we do not yet have instruments that are sensitive enough."

Woolf and Roger Angel, UA Regents' Professor of astronomy and director of the UA Mirror Laboratory, are involved in the Jet Propulsion Laboratory's "Terrestrial Planet Finder" (TPF) project, a space observatory that NASA plans to launch in 2012 as part of Origins Program.

According to Woolf and Angel, most can be learned about the planets by observing in the middle infrared part of the spectrum, the radiation emitted by any object at approximately room temperature. Planets are easier to detect in the infrared because stars are not nearly as bright as they are in visible light.

"We are still finding that the easiest method to see the planets involves the process of 'nulling'," Woolf says. He and his collaborators have been testing this innovative technique for about 5 years.

"The star is usually about 10 million times brighter than its planet. But we can cause the light waves from the star to interfere with themselves, making the star nearly invisible while radiation from the planet comes through. It is possible either to make direct images of a planetary system with this technique or to reconstruct an image by mathematical data processing, Woolf says.

Another imaging technique appropriate for shorter wavelengths is called coronography. This technique requires that the telescope mirror surfaces be extraordinarily smooth to reduce the scattered starlight. Second, the bright rings around a star image must be made fainter by a process called apodization. Third, an Earth-like planet is itself so faint that a very large telescope is needed to analyze its spectrum.

"As a result, telescope systems necessary to observe the visible (wavelengths) from planets are not much different in size from nulling telescopes, and we learn less fundamental information about a planet from its visible spectrum. Studying the planet's infrared glow will allow us to measure its size and how warm it is as well as test the planet's atmosphere for the presence of oxygen in the form of ozone and water," he says.

Woolf and his collaborators have already begun the first explorations in the infrared with the Multi-Mirror Telescope situated on Mt. Hopkins, (before the telescope was converted to a 6.5-meter telescope). In 2004 they will start using the twin-8.4-meter-mirrored Large Binocular Telescope being built on Mount Graham.

Both observatories are located near Tucson, Ariz. These instruments will be able to image the dust around stars as well as Jupiter-like planets but will require the use of 'adaptive optics' technique to correct for blurring effects of the Earth's atmosphere.

"There is a good chance to find Earth-like planets and even find life on these planets if it's there. The main reason for thinking that life is not difficult to get started on an Earth-like planet is, that on Earth life began so soon after the period of catastrophic collisions 4 billion years ago.

"It is hard to imagine that there was any really difficult process on the way. I think that there is continuity from chemicals cyclically combining and breaking up in volcanically warmed places of early Earth, to the formation of self-replicating chemicals, to the development of life and ourselves," Woolf says.