A Scientific Assessment of a New Technology Orbital Telescope (1995)

If the ATD/NTOT passes many of the key technological tests outlined in Chapter 4 and can also exceed its design lifetime by a considerable margin, it should have a significant capability for astronomical research. This being the case, what science projects can be conducted with the ATD/NTOT, but not with any other facility currently in development?

The report of the Astronomy and Astrophysics Survey (Bahcall) Committee ¹ emphasized an understanding of origins—the formation of the universe, galaxies, and stars and planets—as an essential theme for astronomy in the 1990s. Similarly, a more recent report by the Committee on Planetary and Lunar Exploration ² underscored an understanding of the origins of planetary systems and life as one of the principal motivations for research in the planetary sciences in the period from 1995 to 2010.

Given these endorsements, the task group emphasizes in this chapter some projects that are critical to understanding origins. Furthermore, it emphasizes activities that require large amounts of observing time (i.e., surveys). As explained in Chapter 3 and Chapter 4, surveys are perhaps the minimal-cost mode for operating a large telescope, and they should not be done with complex, versatile instruments such as the HST that serve wide communities of users.

The first project discussed here involves the use of wide-field infrared surveys to directly address the origin of galaxies. The second and third projects address the origin of planets and stars; one looks at primitive planetesimals preserved in the Kuiper Disk beyond the orbit of Neptune, while the other focuses on young stars still in the throes of their birth. The fourth is a possible project for addressing the origin and subsequent evolution of planetary atmospheres by observing global change on Pluto and Triton.

The ability to carry out these projects must, of course, be assessed after the ATD/NTOT is in orbit and its performance has been adequately tested. The infrared galaxy survey, though significantly enhanced by the addition of a large-format framing camera, can be carried out fruitfully so long as the baseline ATD/NTOT performs somewhere near its expected level. The Kuiper Disk survey requires the addition of a framing camera. Moreover, the scientific return from both projects is improved dramatically (by at least an order of magnitude) if the ATD/NTOT is flown with the enhanced mirror suggested by the task group. The young-star project can be carried out in part if the telescope performs nominally, but other aspects of this undertaking require that the point-spread function be very stable and that scattered light is minimized—factors that will not be fully understood until after launch. Although important scientifically, this project is one that becomes appropriate for the ATD/NTOT only if the technology performs very well.

The study of global change on Pluto and Triton differs dramatically from the other projects in that it requires

extensive operational control of the spacecraft. Indeed, conducting this project depends not so much on the ATD/NTOT’s performance as on other factors. Prime among these are the extra operational costs. Nevertheless, if the ATD/NTOT is viewed as a demonstration of technological capability, such an innovative use of a spacecraft represents an interesting challenge.

The area in which the baseline ATD/NTOT has the greatest promise for making new astronomical discoveries is in deep infrared surveys of the early universe. Several possible observing campaigns are described below.

Several lines of evidence suggest that the epoch of galaxy formation lies beyond a redshift (z) of about 1. For example, the co-moving number density of both quasars and their absorption-line systems changes dramatically for redshifts greater than approximately 1.5. The nature of galaxies associated with distant radio sources, and the nature of the radio sources themselves, also change significantly for redshifts beyond z ~ 1.

This has been beautifully exemplified in HST images of the radio source 3C 324 at z ~ 1.2 (Figure 5.1). The dramatic examples of interacting and merging galaxies in this image are also seen at lower redshifts, but it is becoming clear that not all galaxies have evolved significantly below z ~ 1. Dwarf galaxies at lower redshifts show evidence of substantial evolution, primarily through bursts of star formation. But more massive galaxies appear to be different; the available data suggest that the bulk of their formation occurred at redshifts greater than about 0.8 to 1. As a class, massive galaxies appear to have undergone little evolution between that time and the present day. Thus the formation or assemblage of these objects must have occurred at a higher redshift.

Unfortunately, extensive studies of such objects with the HST will prove to be very difficult. Because they are expected to have high sensitivity from below 0.4 microns to ~2 microns, the HST’s second-and third-generation imaging instruments, such as NICMOS and HACE, respectively, will clearly provide substantial gains for studying distant galaxies. However, several factors mitigate against the HST for a comprehensive study of galaxies at redshifts of z ~ 1 to 5.

At these redshifts, the optical and near-infrared images correspond largely to wavelengths in the rest-frame ultraviolet. The brightness of a galaxy in the ultraviolet, however, depends dramatically on the amount of star formation and on the quantity and distribution of dust. Surface brightness dimming (through the (1 + z)⁴ factor) also has a major impact at these redshifts. The contrast of distant galaxies against the foreground therefore decreases rapidly at redshifts of z > 1. While quite model dependent, the dilution factor owing to foreground galaxies could be as large as 100:1.

The cumulative effect of all these factors is large and is exacerbated by the HST’s small collecting area. Extremely luminous star-burst galaxies with strong near-ultraviolet and ultraviolet fluxes will likely be identified and studied with the HST, but a more complete census of high-redshift objects will require a telescope with greater infrared sensitivity. Typical “young” galaxies are likely to have broadly peaked energy distributions around 0.4 to 0.6 microns, and so the study of these objects is best carried out in the wavelength range 0.5(1 + z) microns, or ~ 1 to 5 microns for such objects in the redshift range 1 < z < 6.

The sensitivity of most large, ground-based telescopes is limited for wavelengths greater than 2 microns by their own extremely high thermal background. Only telescopes that are both optimized for the infrared and located in Antarctica have a chance of reaching the limits of the atmospheric background, which even in Antarctica is considerable compared to, for example, the zodiacal background seen from space. Adaptive optics will clearly help these telescopes by increasing the contrast between small sources and the background, but the background remains a dominant source of noise, especially for any extended components of the sources.

Cryogenic space telescopes, such as the European Space Agency’s soon-to-be-launched Infrared Space Observatory (ISO) and NASA’s proposed Space Infrared Telescope Facility (SIRTF), will be free from any thermal background at the wavelength regions being considered here, but they will face another problem. At the faint magnitudes required for study of these distant galaxies, source confusion is already a serious concern. At a median

FIGURE 5.1 This deep HST image shows the distant cluster of galaxies around the radio source 3C 324. Very few of the cluster’s members are recognizable as normal spiral galaxies. Most have irregular shapes (inset, top) and appear disrupted, possibly due to mergers and interactions. In contrast, the elliptical galaxies present (inset, middle) are remarkably similar to those seen at lower redshifts. The peculiar radio galaxy 3C 324 (inset, bottom) has a redshift of 1.2. This image, which shows objects as faint as 29th magnitude, is the full field of view of the HST’s Wide-Field/Planetary Camera-2. It required an 18-hour-long integration, made over 32 orbits between May 11 and June 12, 1994. (Courtesy of NASA’s Space Telescope Science Institute.)

redshift of z ~ 0.6, there are almost 10⁶ background galaxies per square degree. Thus, as seen with a resolution of approximately 1 arc sec, background galaxies cover some 10 to 15% of the sky. Surveys attempting to reach significantly higher median redshifts will require sensitivities an order of magnitude higher, at which point small space telescopes (like ISO and SIRTF) or large ground-based telescopes without adaptive optics will become seriously confusion limited.

Given that ground-based telescopes experience very large infrared backgrounds relative to the zodiacal minimum (see Chapter 3), it is clear that the optimum approach for galaxy studies is through the use of a large, cooled space telescope. A 4-meter-class telescope in space, with resolutions of the order of 0.1 arc sec, cooled and instrumented for optimum detection in the 2 to 5+ micron range, would be an immensely powerful tool for comprehensive exploration of the galaxy population in the 1 < z ≤ 6 range. Although the baseline ATD/NTOT mission is not optimized for low-temperature operations, with modest passive cooling its performance in the 2- to 5-micron range is still impressive—better, in fact, than anything else currently under development. Moreover, with the implementation of some of the enhancements described in Chapter 2, the ATD/NTOT’s performance would be close to that of an ideal system for observing galaxies in the early universe.

The simulations shown in Figure 5.2 compare the expected capability of a 4-meter-class, passively cooled space telescope with that from the ground in the K-band at 2 microns and hence into the background minimum at 3 microns (where the baseline operating temperature of 200 K provides background-limited performance at 2 microns—but optimum sensitivity at longer wavelengths requires an enhanced design that uses passive cooling to reach 160 K). The simulations are also shown for two different values of the deceleration parameter q₀ (i.e., Ω₀ =

1 (flat) or #937;₀ = 0.1 (open), assuming that the cosmological constant (Λ) is 0). At such high redshifts the counts clearly are influenced very dramatically by the cosmology. Thus, they are a means of discriminating between different cosmological models, although not a very quantitative one because of highly model-dependent uncertainties from merging, star formation, dust, and luminosity evolution.

It is clear that a deep, multifield survey would be of great utility for exploring the early universe. A few regions at high galactic latitude should be selected (free from obvious clustering at lower redshifts and galactic “cirrus”), and a series of integrations carried out in a strip.

Based on a 5-year science mission in which 25% of the time is spent on the zodiacal-minimum deep (and ultradeep) surveys, some 2800 hours of integrations should be a practical goal for these multifield, multi-waveband observations (assuming an on-target data collecting efficiency of 25%). For year-round access to low-background regions, and for subsequent ground-based follow up (radio maps, further optical and near-infrared imaging and spectroscopy, and so on), a nominal goal would be to observe three fields. Each of these fields is optimally a strip at least several arc minutes long; the task group has chosen a region 2.5 arc min wide by 12 arc min long for the purpose of estimating the limiting magnitude. Since the InSb sensor has a nominal field of view of only 54 arc sec, many pointings will be needed, even for this relatively small strip. One pointing in each of these fields would also be chosen for an ultradeep survey (see next section).

The primary imaging band (with a nominal 30% bandpass) would be centered on the background minimum, as defined by the ATD/NTOT’s temperature and the zodiacal background but likely to be near 2.5 microns. In such a band, the limiting magnitude at a signal-to-noise ratio of 5:1 can be expected to be approximately K = 29 in 10 hours of integration (for an unresolved source—and based on the system parameters listed in Chapter 2). Although galaxies will typically be resolved, there are indications from recent HST images that the source size will be small. Very faint sources appear to have half-light radii of <0.5 arc sec, and probably nearer 0.3 arc sec for typical galaxies at z = 1.0.^3,4 Thus the loss in limiting magnitude will not be large for distant galaxies (a reasonable estimate might be K ~27). To cover the above strip with images would require 42 pointings, each of 10 hours, for three fields for a total of 1260 hours of integration.

While the primary waveband should be centered on the background minimum, observations of additional bands can broadly define the redshifts of the sample objects because of the dramatic changes in luminosity occurring across major spectral features (e.g., the H/K break at 0.4 micron). Several additional bands, probably two on the blue side and one on the red side of the minimum, should be selected. These need not go as deep. If an optical framing camera is added to the ATD/NTOT’s baseline instrument suite, simultaneous broad-band optical (nominally I-band at 0.8 micron) imaging would also be carried out on these fields. Simultaneous use of the framing camera would provide one of the blue bands at no incremental cost in observing time. Since this optical image would be obtained along with every infrared integration, the limiting magnitude would be large (but against a zodiacal background that is substantially larger than that at 2 to 4 microns). If 5 hours per pointing is allocated per band (for each of the two additional infrared bands), or an additional 10 hours per pointing, a further 1260 hours are needed to obtain the additional colors.

It is unlikely that accurate redshifts can be obtained for many of the distant galaxies that would be imaged with ATD/NTOT. Only those with magnitudes brighter than K ~ 21 are likely to be within the reach of 8- to 10-meter-class telescopes with high throughput spectrographs. Thus it is very important to estimate the redshifts from infrared colors as described.

What does the universe look like at the faintest possible levels? The ATD/NTOT could go to fainter magnitudes and look deeper into the universe than any other telescope has ever done. An ultradeep survey of a small region of the sky through a broad filter centered on the background minimum could be used to define the magnitude/number-count relation N(m) to unprecedented levels. The size-distribution/magnitude relation could also be determined, as could the surface density of objects as a function of magnitude. Existing HST data suggest that the sizes and shapes of high-redshift galaxies are dramatically different from those of objects at intermediate redshifts.

One could also envisage carrying out a measurement of the cosmic background light, that is, the unresolved background flux. This would require careful modeling and subtraction of the resolved sources. Given the problems with confusion of foreground galaxies mentioned above, this undertaking will not be practical with conventional ground-based telescopes. Even equipping ground-based telescopes with adaptive optics will not help, since the project requires a very stable and quantifiable point-spread function that cannot be achieved given the variable seeing ground-based telescopes suffer. Furthermore the infrared array must be very stable and well calibrated.

Logistically, the most practical approach to this ultradeep survey would be to center the field in one of the strips used for the deep galaxy survey. Since the InSb fields are so small, it would be very prudent to have a single pointing in each of the fields, so as to evaluate the dispersion in the counts over several independent fields. Within the allocated budget a single integration of 100 hours per strip would be practical. This is 10 times longer than the other integrations discussed so far, thus satisfying the goal of a substantial increment in depth (for galaxies one might expect to get to K ~ 29-30!). Together with the strip survey, the ultradeep survey fills the nominal 2800 hours for the deep zodiacal minimum surveys. If additional time were available it might be valuable to obtain increased integrations in two other infrared filters at that pointing. Again, as noted above, the addition of an optical framing camera would allow I-band images to be obtained contemporaneously with each of the infrared images, thereby building up an ultradeep optical image.

These images would provide a means of establishing, albeit crudely, some idea of the redshift distribution of the detected objects. Unusually large color changes would indicate, as discussed above, that major spectral features such as the Lyman limit, Lyman Alpha, or the H/K break could lie between the filter bandpasses.

Over the next few years it is expected that substantial progress will be made in establishing the geometry of the universe from survey observations of type-1 supernovae and the use of other “standard candles” at intermediate redshifts (e.g., using internal velocities for fundamental plane studies of ellipticals galaxies). However, as history has shown, such tests are likely to be less definitive than expected. Thus tests at higher redshifts will likely be needed to minimize the uncertainty in the determinations of the cosmological parameters. The sensitivity to different cosmologies is just too small at low redshifts to provide a highly accurate determination of the deceleration parameter, q₀. The relative values of Ω_baryon, Ω_matter, andΩ_Λ are therefore likely to remain in dispute for some time to come. Supernovae or certain classes of objects may well provide useful probes of the scale of the universe at high redshifts.

The ATD/NTOT could help begin the exploration for suitable standard candles, even if it is unable to complete the program during its lifetime. It could be a pathfinder that would identify some of the central goals of cosmology that would be addressed by a subsequent, longer-lived, and more capable astronomy mission.

Since comets represent some of the least altered objects left over from the formation of the solar system, the study of comets provides important constraints on solar nebular models. However, to use these constraints, astronomers must understand where in the solar nebula these primitive objects formed and, therefore, where the constraints are pertinent.

Known comets are believed to have originated in one of two reservoirs in the solar system. The first of these is the Oort cloud, a spherical ensemble of comets that surround the Sun at distances of 10³ to 5 × 10⁴ AU. The comets in the Oort cloud probably formed in the Uranus-Neptune zone and were ejected outward by planetary perturbations. Perturbations pumped up the inclinations until a halo of objects resulted. Comets from the Oort cloud are occasionally nudged into a random walk toward the inner solar system by the passage of nearby stars, interstellar clouds, or other disturbances.

However, more recently, it has been noted that the comets with the shortest orbital periods (less than 20 years), the so-called Jupiter-family comets (JFCs), all have inclinations that are near the plane of the ecliptic. Since no known mechanism could preferentially flatten the inclinations of these JFCs when they were perturbed from the

Oort cloud, it has been argued that the source of the JFCs must be a disk of comets whose inner edge is just beyond the orbit of Neptune. ^5,6 This disk is now known as the Kuiper Disk because it was Gerard Kuiper⁷ who pointed out that unless there was more material past the orbit of Neptune, the solar system would have a sharp edge to its mass distribution, which was unlikely. These comets likely formed where they are currently located and have undergone little or no evolution.

Until recently, the Kuiper Disk was only a theoretical construct because there was no observational evidence for its existence. Since 1992, however, more than 20 objects with trans-neptunian orbits have been found. All have V magnitudes of 23 to 25, putting them near the limits of ground-based detection. Assuming a typical cometary albedo of 0.04, these objects all have radii from 50 to 100 km, most at the larger end. These sizes are inconsistent with the sizes of known JFCs, which typically must have radii of 1 to 20 km. Astronomers know of no mechanism for converting 100-km Kuiper Disk objects to 10-km-sized cometary nuclei.

Thus, our knowledge of the JFC parents is still not firm. It is critical to our understanding of conditions in the early solar nebula to provide a link between the Kuiper Disk and the JFCs. Finding the link requires finding Kuiper Disk objects whose sizes are more typical of those of JFCs. However, at 40 AU, the inner edge of the Kuiper Disk, a 10-km object with an albedo of 0.04 would have a V magnitude of 28.5, much too faint for detection by ground-based surveys.

While the HST with its current Wide-Field/Planetary Camera 2 (WFPC2) can reach such faint limiting magnitudes, it requires a combined integration time of 5 hours to reach signal/noise 4 for V = 28.5. With the HST’s low orbit and the resultant Earth occultation periods, it takes approximately 30 hours to obtain the necessary 5 hours of integration time. Even at quadrature, when Earth parallactic motion is zero, an object at 40 AU would move 15 arc sec (or 150 pixels on the wide-field CCDs of the WFPC2) in 30 hours. Additionally, the WFPC2’s field of view is small (around 4.2 square arc min for the three wide-field CCDs), and observing time is highly competitive. The combination of these factors means that although the HST can observe deeply enough to image objects of the size of a large comet at 40 AU, an extensive survey for such objects is not an appropriate use of the HST.

With a 4-meter space telescope, especially one with a Molniya orbit or another orbit that does not suffer from frequent Earth occultation, a definitive survey for the Kuiper Disk could be undertaken. As noted in the previous section, the larger aperture and higher resolution of the ATD/NTOT should allow researchers to observe point sources with I = 29.5 (V ~ 30 for typical colors) (i.e., objects with radii R ~ 6 m at r = 40 AU or with R = 100 km at r= 150 AU) in a 1-hour integration. One could even detect a Pluto-sized object in the inner Oort cloud. Since the Molniya orbit allows multiple 1-hour exposures per orbit (probably five to eight) whereas the HST orbit requires 30 hours to complete a single 5-hour integration, the difference in results is dramatic.

The ATD/NTOT would allow astronomers to observe 10-km objects to 50 AU. Assuming the canonical r⁻³ distribution of bodies with size, researchers should find roughly 1000 objects per square degree, or one in every few fields, even with the very small field of view of a single 2048 × 2048 chip.

It should be noted that the 2048 × 2048 CCD projected under the task group’s proposed enhancements would have pixels of 0.03 arc sec or a field of view of only 1.1 square arc min. Therefore some of the ATD/NTOT ’s speed advantage over the HST is negated by the even smaller field size. However, a 5-hour integration could be achieved in a single orbit, making this telescope much more efficient than the HST for carrying out such a survey.

A very deep survey covering a significant region of the sky would enable a study of the mass distribution of the Kuiper Disk. Current ground-based surveys are restricted to the brightest and largest objects, and only a very small region of the sky can be examined by the HST. For the latter, distances and sizes must be computed based on assumptions about the orbit of candidate objects, since it will be impossible to follow up the original observations and compute orbits. With a more complete survey using the ATD/NTOT, it will be possible to compute actual orbits for objects if follow-up observations are planned.

The group of 20-plus known, large Kuiper Disk objects is a critical confirmation that trans-neptunian objects exist. However, their link with the JFCs is uncertain. A mass distribution is needed to constrain dynamical

integrations of the lifetimes of such objects. A comparison between the observations of JFCs and the numerical integrations of Levison and Duncan ⁸ or Holman and Wisdom ⁹ shows that the current population of Kuiper Disk objects is smaller by a factor of 1000 to 10,000 than a simple extrapolation of the surface density in the outer planetary region would predict. This discrepancy leads to the conclusion that there is a dip in the surface density of the solar system outside the orbit of Neptune. Knowing the size distribution of Kuiper Disk objects can allow unique differentiation among three possible explanations for this feature:

1. There was an edge to the original solar nebula,

2. The Kuiper Disk initially had a much larger mass but evolved to its current mass due to collisions, or

3. The early migration of Neptune (either inward or outward) dynamically cleared the inner portion of the Kuiper Disk.

In order to continue progress in our understanding of the Kuiper Disk, it is not sufficient to continue to discover the objects that are the largest members of the class. Thus, there will come a limit to the effectiveness of ground-based surveys. The small field of view and the nature of the way in which the HST is scheduled do not allow for a detailed survey, although the HST could obtain observations of an interesting, small, limiting size. Indeed, the HST recently obtained suggestive, but as yet unconfirmed, evidence of the existence of a population of small Kuiper Disk objects. With the much longer continuous viewing time of the Molniya orbit, a 4-meter telescope that undertook as a project to complete the surveying of a large piece of sky (models predict 1000 10-km objects in 1 square degree) to I = 29.5 or V ~ 30 would finally achieve a definitive picture of the mass distribution in the Kuiper Disk (Figure 5.3), both according to the size of the object and, for objects like those already discovered, with distance from the Sun. Assuming 16 useful hours per day of observation in the Molniya orbit, such a survey would take roughly 8 months, a reasonable share of a 5-year mission.

Depending on the details of the operations of the telescope, a variety of more or less automatic follow-up programs could be executed on each object discovered in the Kuiper Disk. These observations would be aimed at determining the orbit quickly and determining some of the basic physical properties. The first such follow-up observation would be a sequence of observations from opposite sides of the spacecraft ’s orbit around Earth in order to directly determine the distance to the body (see Chapter 4). This would allow a rapid separation of Centaurs, Neptune librators, and transition objects from true Kuiper Disk objects.

Another important follow-up observation would be to obtain colors of the Kuiper Disks objects and to determine their photometric variability in order to understand the distribution of rotational periods of these objects and even to place constraints on their deviations from a spherical shape. Observations to date have shown that at least one of the known trans-neptunian objects exhibits significant variability in its brightness. All of these observations are straightforward with the ATD/NTOT once the object has been discovered and its motion determined to sufficient accuracy that it can be found again. The task group notes that an I-K color would be reliably determined automatically for objects in the central part of the field, since the InSb detector would presumably be taking data at all times also.

Closely related programs can also be carried out in a survey mode. One obvious choice is to image all known cometary nuclei as close as possible to their perihelia in order to estimate nuclear sizes (assuming an albedo), estimate nuclear rotational periods, and determine nuclear colors. Lamy and Toth ¹⁰ have already demonstrated the great advantage of the HST over ground-based observations for photometrically separating the nucleus from the coma, even when comets are near the Sun and have significant comae.

The potentially superb spatial resolution of the ATD/NTOT at optical wavelengths enables higher spatial resolution than with the HST, and the large collecting area makes it feasible to carry out the observations when comets are further from the Sun and therefore when cometary activity is weaker. The task group notes that the point-spread function of the ATD/NTOT will have a full width at half maximum (FWHM) of roughly 150 km for a JFC near aphelion and about 300 km at Chiron. Although this performance is not sufficient to spatially resolve

the nucleus, it is sufficient to photometrically resolve, and thus determine the nuclear brightness from, any coma that is strong enough to matter.

FIGURE 5.3 A comparison between various searches for Kuiper Disk objects and a numerical model of the size distribution of these trans-neptunian primitive bodies. The number of Kuiper Disk objects per square degree as seen on the sky brighter than a certain V magnitude is shown as a function of that magnitude. The top axis shows the size of the smallest object at 40 AU detectable with these surveys (assuming an albedo of 0.04). The dashes with downward pointing arrows represent upper limits determined by unsuccessful surveys. The success of the model is indicated by noticing that all the unsuccessful searches resulted in points lying above the curve and that the survey (asterisk) that discovered the Kuiper Disk objects 1992 QB1 and 1993 FW lies below the curve. The model also predicts that the ongoing so-called HST/GO and Archival searches should be successful in detecting objects as small as 10 to 20 km across. Use of the ATD/NTOT would open a much larger region of search space for investigation. A 250-day survey to V = 30 (I = 29.5) would cover 1 square degree and thus set three-sigma limits of three per square degree as shown.

Nuclear colors, particularly those extending into the near infrared, are extremely important for understanding the physical nature of cometary nuclei. Colors are particularly important for determining whether there is bare ice on the surface, either everywhere on the surface or in patches such as the active areas observed on Halley ’s comet.

The ATD/NTOT should be capable of measuring even the near-infrared colors for some comets at aphelion and for others when they are closer to the Sun. Observations at other points in their orbits would also be invaluable for understanding the onset of activity as periodic comets approach the Sun. Since the number of known short-period comets is less than 200, this entire program is only a small part of the basic survey of the Kuiper Disk.

As noted above, the key research areas identified in the Bahcall report include the origins of stars and planets.¹¹ The study of star-forming regions is another area that can benefit dramatically from an extensive survey with the ATD/NTOT if it performs as desired.

At the time of the ATD/NTOT’s flight (presumably shortly before the turn of the century), adaptive optics will enable large, ground-based telescopes to often achieve high Strehl ratios in the infrared, although limitations on sky coverage will remain unless laser guide-star systems are routinely in use. Thus, the ATD/NTOT’s resolution, as such, in the infrared is not a dramatic gain for addressing these problems. On the other hand, studies of star-forming regions require extracting every possible bit of resolution from the telescope. Thus if its point-spread function (PSF) is very stable, the ATD/NTOT will have a major advantage over ground-based telescopes whose PSFs are certain to be rendered unstable by variable seeing. Furthermore, even at 2 microns, studies at NASA ’s Infrared Telescope Facility (IRTF) have shown that neither the Keck nor the Gemini telescopes will be able to compete with IRTF with tip-tilt (and thus with ATD/NTOT) until they adopt full, high-order adaptive optics.

The ATD/NTOT’s much bigger advantage will occur at visible wavelengths because adaptive optics will not have been developed to a sufficient extent to allow multimeter-aperture, ground-based telescopes to achieve high Strehl ratios. Thus the PSF of the ATD/NTOT will provide a dramatic increase in resolution so long as the performance expected from an enhanced primary mirror (see Chapter 2) is achieved. This increase in resolution will be particularly advantageous if the PSF is stable over long periods.

Clearly an optical framing camera of astronomical quality will be required to take advantage of these technological gains in the telescope. Ideally, a coronagraph should also be installed to minimize the scattered starlight when looking for or at faint disks around stars. But it appears to this task group that the quality of the primary mirror and the significant diffraction by the secondary supports do not justify a coronagraph. Even without a coronagraph, however, the ATD/NTOT has the potential to make significant incremental advances in studies of star formation.

The following sections describe several important observing campaigns that would take advantage of the unique capability of an enhanced ATD/NTOT with a stable PSF to address three separate phases in the formation of stars and particularly planets.

The task group believes that the ATD/NTOT can address three of the goals for understanding the phase of star formation concerned with embedded young stellar objects. These include:

• Quantifying the frequency of disks in star-forming regions of differing densities,

• Inferring the structure of envelopes surrounding forming stars, and

• Determining the duration of the envelope-infall phase for stars of different mass.

The first goal is to quantify the frequency of disks around stars of masses ranging from 0.1 to 5 solar masses in star-forming regions of differing stellar densities. The effect of stellar density on the formation of disks around protostars is an important open question that dramatically affects the inferred number of planets around other stars. The appropriate technique to investigate this question is to use imaging in the J, H, K, L, and M color bands of regions in which the density of young stellar objects (YSOs) is low (i.e., the mean separation between stars is ~0.1 parsec) and high (i.e., the mean separation between stars is < 0.01 parsec). As noted in Chapter 3, observations in the L and M bandpasses are far better done from space because of the dramatically reduced background emission. A space telescope’s advantages relative to those of a warm, ground-based telescope enable sensitive imaging of deeply embedded YSOs, including those still shrouded by their natal cores. This imaging would allow extremely young, embedded stars (A_v < 50) to be examined and disk frequencies and lifetimes to be evaluated over the entire mass range. The ATD/NTOT’s high angular resolution is also crucial because dense star clusters, which are

forming the full range of stellar masses, are found only at distances greater than some 500 parsecs where the projected separations between objects in a dense cluster are only of the order of 1 arc sec.

The second goal is to quantify the structure of envelopes surrounding forming stars. It should then be possible to infer the range of both angular momenta (characterizing protostellar cores) and centrifugal radii (characterizing the disks surrounding young stars) as a function of stellar mass. This characterization is best done by comparing models of infalling envelopes, computed for a range of initial conditions (e.g., core sound speed, mass, rotation), with JHKL imaging of the scattered-light envelopes surrounding embedded YSOs.

The study would also ideally utilize narrow-band (e.g., [Fe II] at 1.64 microns; Br γ; Br α; and H₂ at 2.2 microns) imaging to locate and characterize jets, which would establish the rotation axis of the system and the interaction between stellar jets and infalling matter. The task group notes that the limited instrumentation in the focal plane might not allow a wide selection of filters. The filters themselves, however, would cost only a nominal amount if the capability were present to install and use them.

These observations would allow imaging of embedded stars out to the distance of Orion and would enable properties of the envelope and disk to be inferred from scattered-light patterns around stars of differing mass. The presence of disks and the magnitude of their centrifugal radii would be inferred from the scattered-light patterns. ¹² The presence of disks with centrifugal radii ranging from 20 to 200 AU should be readily inferred due to the ATD/NTOT’s high angular resolution (FWHM ~ 20 AU for the near infrared at Orion).

The third goal is to determine empirically the duration of the envelope-infall phase around stars of different mass. As with the two previous projects, the appropriate technique is to use JHKL imaging to search for the scattered light patterns produced by dust in rich, embedded clusters. These images can be the same ones used for the second project.

Other interesting observations (e.g., those aimed at diagnosing the presence of accretion-driven jets and winds) would benefit from narrow-band images in the light of H₂ lines. High-angular-resolution observations of jets associated with YSOs spanning a range of masses and relative evolutionary states would enable characterization of jet collimation and wind-envelope interactions, and provide important insight into the role played by winds and jets during the early stages of star formation. The advantages of the ATD/NTOT would be its high angular resolution and the ability to obtain deep-infrared images of heavily obscured YSOs in their earliest evolutionary phases.

Even when a YSO’s optically thick, infalling envelope disappears, it is still surrounded by a massive, optically thick accretion disk. There are two main goals for understanding this so-called accretion-dominated phase of star and planet formation:

• Determining the sizes of disks surrounding stars of differing mass, and

• Understanding wind and jet morphology and collimation as a function of evolutionary state.

Observing scattered-light patterns is a good way to determine the sizes of disks. This is best done with R- and I-band imaging of a sample of optically visible YSOs in nearby (~150 parsecs) star-forming regions. Selecting which stars to examine is best done by looking for signatures of disks in their spectra. For geometrically flat disks, the disk/star contrast ratio may be less than 0.001 and thus extremely difficult to observe. The suitability of the ATD/NTOT for this project would need to be evaluated after the telescope has been thoroughly tested in orbit.

These observations would also be influenced by the presence of scattered-light contributions from dust embedded within an optically thin, remnant infalling envelope. With the correct geometry, this contribution might overwhelm that from the disk. Detecting such envelopes is interesting for understanding star formation but not (directly) for understanding the formation of the solar system.

The second goal is to examine the wind and jet morphology and understand wind collimation as a function of evolutionary state. Optical (R, I) probes (e.g., [S II], which is formed and excited in shocks) would enable imaging on scales of 5 AU.

After the accretion phase is over, YSOs have disks that are populated by “secondary” dust produced by collisions between planetesimals. The study of these YSOs, called weak T Tauri stars, represents potentially the most exciting arena for ATD/NTOT observations. Astronomers would first observe a sample of several hundred weak T Tauri stars in nearby star-forming regions and search for evidence of low-surface-brightness scattered-light disks. Observations in the R and I bands would take full advantage of the ATD/NTOT’s high angular resolution to image scattered-light disks and resolve objects as small as 5 AU. Disk sizes and radial surface densities could be inferred from surface-brightness distributions.

The presence of “dust-free” holes in disks might provide the strongest, although indirect, evidence for the presence of planets of sufficient mass to sweep up the material and create a void between the inner and outer disks. By targeting stars whose spectra (from ISO and ground-based telescopes) suggest large inner holes, astronomers could search for evidence of such (r>5 AU) holes in a disk’s surface-brightness distribution. Narrow-band filters could be used to attempt to carry out mineralogical studies of disk dust, at least for those mineral species containing Fe⁺ and Fe²⁺ that have absorption features in the 0.5- to 5.0-micron spectral range. Because of the low surface brightness of some such disks, the capability to do this project must be tested on orbit when the ATD/NTOT’s actual PSF has been evaluated.

Numerous synoptic problems relating to planetary atmospheres are well suited to the high spatial resolution of the ATD/NTOT, although the advantage of using the ATD/NTOT rather than the HST are not particularly clear. One particular project relating to Pluto and Triton stands out because it is more closely related to origins than the others that immediately come to mind and also represents a much more significant demonstration of technological capability.

Pluto and Triton are of interest because these bodies most likely formed at the same time as the Sun and not when the other planets formed. As such, they provide a sample of objects somewhat larger than those in the Kuiper Disk whose formation was contemporary with the Sun’s. Bodies of this type may have been the building blocks of the outer planets. Their relatively large sizes mean that they are differentiated bodies and are massive enough to retain atmospheres, with surface pressures in the microbar range.

Because of its low gravity, Pluto’s atmosphere is the largest in the solar system relative to the size of the body. As a consequence, Pluto’s outer atmosphere is in hydrodynamic escape. Unlike the jovian planets, the major constituent of Triton ’s and Pluto’s atmospheres, N₂, is likely to be in vapor-pressure equilibrium with surface ice. These atmospheres also contain small amounts of CH₄ and CO. The only other body in the solar system whose atmosphere is known to be in vapor-pressure equilibrium with surface ice is Mars, whose principal atmospheric constituent is CO₂.

Pressure changes of several orders of magnitude have been predicted to occur as Pluto and Triton respond to seasonal changes in insolation. This behavior arises because, at the temperatures involved (~40 K), vapor pressures are extremely sensitive to temperature. The magnitude of these surface-pressure changes depends on the sublimation-condensation properties and extent of subsurface volatile reservoirs.

Learning about how the atmospheric structure of these bodies changes over time scales of months and years is necessary for understanding the complex relationships involved in surface-atmosphere volatile exchange, as well as the processes that control their basic atmospheric structure. For example, one model predicts that the structure of Pluto’s middle and lower atmospheres is controlled by CH₄, whereas another predicts that CO is more important. The more we learn about these atmospheres from remote observations, the more effectively we can plan investigations to be conducted by future spacecraft missions to these bodies.

The structure of Pluto’s atmosphere has been probed at only one time—by a stellar occultation in 1988. The structure of Triton’s atmosphere has been probed twice—during Voyager 2’s encounter in 1989 and by a stellar occultation in 1993.

If the technology demonstration of the ATD/NTOT’s maneuverability (see Chapter 4) proves successful, then it would be possible to carry out a program of 1 to 2 atmospheric probes of these bodies per year, based on their (present) average rate of stellar occultations. This rate could be higher or lower, depending on the orbit of the telescope and the maneuverability demonstrated (i.e., how much, how accurately, and how often) during the testing phase of the ATD/NTOT mission.

The occultation rate is proportional to the fraction of time during an orbit when Triton or Pluto is visible. Also there would be a gain in the number of potential occultations that is proportional to the diameter of the ATD/NTOT’s orbit (projected onto the Earth’s polar axis) relative to the diameter of the Earth. For a Molniya orbit this factor is 2.6.

The ability to undertake this observing program does not depend on any of the enhancements suggested in Chapter 2. Rather, the decision should depend on the availability of resources to pay for the significant extra operational costs entailed in maneuvering the spacecraft to a new orbit or to a different position in its current orbit. It is also unclear to the task group whether or not the spacecraft would be usable for other astronomical tasks while the maneuvering was taking place. If not, the observational efficiency, defined as time collecting scientific data divided by elapsed time, would be extremely low. Low efficiency is not inherently bad, but it represents a value judgment that cannot be made until more is known about the maneuverability of the spacecraft.

1. Astronomy and Astrophysics Survey Committee, National Research Council, The Decade of Discovery in Astronomy and Astrophysics, National Academy Press, Washington, D.C., 1991.

2. Space Studies Board, National Research Council, An Integrated Strategy for the Planetary Sciences: 1995-2010, National Academy of Sciences, Washington, D.C., 1995.

3. Phillips, A.C., Gronwall, C., Koo, D.C., Forbes, D.A., Illingworth, G.D., and Huchra, J.P., Astrophysical Journal Letters, 1995, submitted.

4. Driver, S.P., Windhorst, R.A., Ostrander, E.J., Keel, W.C., Griffiths, R.E., and Ratnatunga, K.U., Astrophysical Journal Letters, 1995, submitted.

5. Fernandez, J.A., Monthly Notices of the Royal Astronomical Society 192:481, 1980.

6. Duncan, M., Quinn, T., and Tremaine, S., Astrophysical Journal Letters 328:69, 1988.

7. Kuiper, G., Astrophysics: A Topical Symposium, J. Hynek, ed., McGraw-Hill, New York, 1951, p. 357.

8. Levison, H.F., and Duncan, M.J., Astronomical Journal 100:1669, 1990.

9. Holman, M.J., and Wisdom, J., Astronomical Journal 105:1987, 1993.

10. Lamy, P.L., and Toth, I., “Direct Detection of a Cometary Nucleus with the Hubble Space Telescope ”, Astronomy and Astrophysics, 293L:43, 1995.

11. Astronomy and Astrophysics Survey Committee, National Research Council, The Decade of Discovery in Astronomy and Astrophysics, National Academy Press, Washington, D.C., 1991.

12. Kenyon, S.J., Calvert, N., and Hartmann, L., “The Embedded Young Stars in the Taurus-Auriga Molecular Cloud I: Models for Spectral Energy Distribution”, Astrophysical Journal 414:676, 1993.