Remote Sensing of Planetary Atmospheres and Surfaces
This session dealt with how we would observe planets to find whether they are present, whether they are habitable like Earth, whether life has developed on them, and whether they have developed intelligent life. These questions are arranged sequentially such that any positive answer further along the chain probably requires a set of positive answers to all the prior questions too. However, a negative answer at any point does not imply that answers to previous questions are negative.
Spectroscopic studies of other planets in our Solar System have already started. However, these planets are sufficiently close that we cannot exclude the possibility that microbial life has traveled from one to another. On the other hand, interstellar distances are so great that the presence of life on other solar systems implies either independent origination, or some type of panspermia process on a large scale.
Radio Searches
The examination of possibilities for searching other systems broke into two fundamentally different categories. Testing for certain types of intelligent life, which might mimic our own behavior, is currently being done and is the only study within this group that is presently possible. It is also possible at relatively modest cost. The Galileo flyby of Earth demonstrated that even close to Earth, only radio observations of communications signals gave totally unambiguous indications of life here. The remote sensing group at this workshop felt that the results from radio studies (like SETI) were of compelling interest and asked the entire conference to endorse the method and current programs as valid scientific work. However regardless of the outcome of the SETI study, it enters the problem so far down the chain of inference (see first paragraph above), that additional searches addressing the frequency of planets, the frequency of habitable planets, and the fraction of these on which unintelligent life has developed are needed to build the science of searching for life on a sound foundation.
Figure 3. The spectrum of an Earth twin after 6 weeks of integration. (from "A Road Map for the Exploration of Neighboring Planetary Systems," C.A. Beichman, ed., JPL Publication 96-22, 1996).
Optical/Infrared Searches
i) Indirect searches: Optical/Infrared searches for planets can either be direct, which is seeing radiation emitted, absorbed or reflected by a planet, or indirect, in which the gravitational pull of a planet on a star causes the star to perform a miniature orbital motion. This motion can be sensed either by its effect on the speed of the star in the direction of the line of sight, or on the movement of the star in the plane of the sky. Both techniques have already proved successful in finding the first few planets around other stars. The status of the techniques is that if either were applied to the Sun at the performance level now applied to other stars, and from a reasonable distance (more than 5pc), no planets would be detected. The techniques depend on the movement of the star not being confused with processes (starspots, prominences, faculae, granulation etc) which vary the brightness over the surface of a star, and therefore make its light-center an inadequate measure of its center of mass. The evidence available today suggests that these effects will make the detection of extrasolar Earth-sized planets extremely difficult. Whether or not the searches become impossible is disputed.
ii) Direct searches. Searches for planets can be made by attempting to detect the radiation either reflected or emitted from the planet. Earth's emission is most visible at infrared wavelengths where it emits about 1 part in 10 million as much as the Sun.
If the orbit of the planet passes through the line joining the star to the Earth, then once each orbit, the planet will eclipse (occult) the star. Earth eclipses of the Sun cut out about one part in 10,000 of the Sun's light for a total of 10 hours every year. It is also possible to search for these eclipses as direct evidence of the presence of planets. Such searches do not seem practicable on Earth, because of weather and other atmospheric effects, but they are possible from space. In space, the topic of debate is not so much whether the measurements can be made, but the size and cost of the operation to make such observations, and how, with the delays inherent in sending up space telescopes, they fit into a program to observe external planets.
Spectroscopic studies of the atmospheres of planets are in principle possible for either of these processes. In eclipse, only the part of a planet's atmosphere beyond the limb and contributing to the eclipse, affects a star's spectrum. Thus we would be trying to see optical features with a maximum possible depth of about 3 parts in 10 million of starlight. For looking straight at the planet, the depths are 100%, but in 1 part in 10 million of the star's radiation.
Direct versus Indirect Searches
Among the differences between these two methods are the following:
1) It is possible to blot out the starlight while still seeing the planet in the second observation, but not possible to blot out the starlight because we must observe the eclipse for the first kind of observation.
2) Because typically only 1 in every 200 Earth-like planets will show an eclipse, studies of eclipses will typically look at stars (200)1/3 further away, and they will be about 35 times fainter than non-eclipsed counterparts.
The technical problems of either observation are acute. To see planets by their own radiation requires an interferometer to blot out the starlight. This technique has not been used previously and it stretches our current technical abilities. The device must operate in the infrared where both the Solar System dust and the telescope glow. The telescope must be sent out to about 5 Astronomical Units at a site where the dust and telescope glow less because they are colder. The device would consist of four telescopes each of about 1.5 meters diameter and aligned in a linear array 75 meters long.
The eclipse measurements must be made with a supergiant space telescope, probably over 100 meters in diameter, in order to see enough stars. Because only a few ten millionths of the radiation would be affected by planetary atmosphere absorptions, detection to this precision would require CCD detectors with very perfect surfaces. The required storage per resolution element of 50 bits, together with the necessary transfer between the CCD and the computer, including the A-D conversion, would be challenging.
Beyond these technical issues are astronomical unknowns. For seeing planets' radiation, how badly are we bothered by glowing dust in the planet's own system? For the eclipse studies, how stable is the spectrum of a sun-like star which has those atmospheric disturbances previously mentioned? It is in principle possible to overcome external dust by building a larger interferometer with larger mirrors. But it would be good to know if this is needed before sending a spacecraft on this long journey. If a star's spectrum is not sufficiently stable, the eclipse technique fails and does not seem to be curable.
What might be seen?
A first generation interferometer to observe planets would see water and carbon dioxide easily, and if the O3 levels are like the Earth's, O3 will be observable but harder to see. The interferometer will detect a planet's radiation, derive a color temperature for the emitting region of its surface or atmosphere, and determine how far the planet is from its star. To detect the next most significant molecule, methane, will require telescopes 100 times larger in area. Nitrous oxide needs an even larger telescope. An ultra-large telescope to observe eclipses would see the molecular oxygen A band. There are water and carbon dioxide features in the same spectral range, but they are very weak. The limits of performance would likely be set by intrinsic star spectral noise. The distance of the planet from the star could be determined from the planet's period, but there would be no independent evidence of temperature.
Additional Recommendations
The study of planet spectra during occultation is a very new possibility. Although some aspects mentioned here appear to be not very encouraging, only a more detailed study can reveal the true size of the problems and the possible solutions. Investigation of possible occultation strategies by FRESIP-style space observatories is worth continuing.
The potential studies that are needed to make space infrared interferometry into a practical tool include flying an optical interferometer, flying an infrared interferometer to detect dust around other stars and flying a probe of the radiation from the zodiacal dust of our own system. All these possibilities are discussed in the EX-NPS final report by NASA. We support such work.
Last updated Jul-10-1997
Responsible NASA Official:
Dr. Larry Caroff
Webmaster: Symtech Corp