Summary

Reasons for a Heterogeneous Array

Large Collecting Areas

Short Spacings

Flexibility

The committee favored the concept of a heterogeneous array with 20 x 8 m dishes plus 50 x 15 m antennas. Both types of dish should be designed to work to a wavelength of 350 microns, with an r.m.s. surface accuracy of 25 microns, and both types of dish should be designed to have an offset pointing (to calibrators) of 0.5 arcsec.

The strongest scientific reasons for having a heterogeneous array lie in the mainstream mm/submm science: molecular spectroscopy and dust continuum observations of galactic and extragalactic sources. There are other reasons from solar system research and the study of extended emission from other thermal and nonthermal sources. These are covered in the MMA scientific documents and previous LSA documents. We felt the instrument should first be justified by the mainstream science, and we concentrated on this.

There is an exciting new universe being opened up by the powerful optical telescopes now operating (HST, Keck) or coming on line (e.g., ESO VLT, Gemini, Subaru, Carnegie, etc). We should keep in mind that in addition to these large new telescopes, the years in which the MMA/LSA will do research will be the post-Hubble years, the period of the Next Generation Space Telescope. The NGST will be able to detect early-universe galaxies at z > 8. We should build an instrument that can detect CO, C I, C II, O II, and N II in such objects, revealed by all these powerful new optical telescopes. This requires an increase in collecting area comparable to that of the VLA, of the order of 10,000 m ².

The goal of both the MMA and LSA projects has been to reach an angular resolution of 0.1" at 1.3 mm. With a collecting area of ~1000 m ², the best current spectral-line maps at 1.3 mm from the IRAM interferometer have a beam of 0.6" and a noise of 1 K in 0.25 km s^-1 channels (needed for star formation studies), and 0.1 K in 25 km s^-1 channels (needed for extragalactic studies). If we reduce the beam size by a factor of 6, we have to increase the sensitivity 36 times to keep the same noise levels. For spectral lines, we cannot get any sensitivity increase from larger-bandwidth receivers. We hope to gain a factor of 4 from some improvement in detectors and with excellent weather at the new site. However, if we want to really use the increased resolution for molecular spectroscopy, then there is no way around the need for an increase in collecting area by a factor of 10 --- the sources we look at are just too weak. The number of sources we can look at does not go linearly with sensitivity. There are definite threshold effects, above which we cannot detect whole classes of sources. The protostellar disks are a particular example.

In the LSA project, we had estimated the money needed to obtain a 10,000 m² collecting area with an array of 15 m antennas. The total project cost was ~300 M$. The June 1997 NRAO-- European agreement is for a larger investment than we were originally proposing. We should not now spend more money for less collecting area, especially since we are making savings in other areas like logistics, infrastructure, common software, etc. Based on initial cost estimates for the different antenna types in the meetings of the antenna group, it appears possible to build 20 x 8 m and 50 x 15 m dishes within the revised budget framework now being discussed (400 M$).

Since the new instrument will be built in the southern hemisphere, where the central part of the Galaxy goes straight overhead, there will be a renaissance in galactic research. Most of the star-forming regions have multiple molecular-cloud cores and extended molecular line and dust emission all around. A central field of research will involve the study of molecules and dust formation in extended circumstellar envelopes. Good images of all these regions in our Galaxy require short interferometer spacings down to about 12 meters. The complimentarity between 8 m and 15 m antennas is ideal to fill in the short spacings.

Many beautiful millimeter-astronomy images of extended molecular-line and dust features in other galaxies, were presented at the IAU Meeting in Kyoto last month. It is obvious from these data that in extragalactic sources there is much extended, low-brightness emission that requires short-spacing information, on interferometer baseline scales of about 12 meters. This includes emission in the main CO lines, their isotopes, in HCN in nearby galaxies, and for a future mm/submm array, in the neutral carbon lines redshifted from 492 GHz and 809 GHz. Even high-redshift objects are extended in both dust continuum emission and CO lines on scales of arcseconds, and require short-spacing observations. This is true of the detected high-z objects like IRAS 10214+4724, the Cloverleaf quasar, BR1202-07, etc, and will also be true of unlensed galaxies, just because at high redshifts, 1" ~ a few kpc, the typical size in which CO and dust is observed in galaxies. The angular scales on which short spacing information is needed may be inferred from Table 1, which lists the (gaussian) source sizes for which 50% and 100% of the flux is lost. The Table gives two examples, a minimum spacing of 12 m (appropriate for 8 m dishes) and a minimum spacing of 24 m (appropriate for 15 m dishes). The dishes may be moved a bit closer than the minimum spacings given in Table 1, if the telescope mounts are especially compact. Furthermore, the angular size scales can be made a bit larger if mosaicing is used. However, no matter what we do, the short spacing requirement will remain particularly acute at submm wavelengths, where interferometers will miss flux on scales of a few arcseconds. To solve the short spacing problem at high frequencies, one needs some small antennas. This is especially true with the new MMA/LSA requirement to work to a wavelength of 350 microns. The problems can be alleviated by having some 8 m antennas in the array. To get this short spacing information, however, it is not be necessary to have 40 small dishes -- ten to twenty 8-meter dishes will amply suffice.

We think a hybrid array that has both 8-meter antennas and 15-meter antennas is the most flexible solution, that allows the array to cover a wide range of science. We can map extended sources, we can map compact sources, we can do millimeter and submillimeter, and for most projects, we will not have to add in ``missing short spacings'' from a single dish.

There will be great flexibility in the correlation of sub-arrays. One may correlate the whole array, the 8 m dishes only, the 8 m dishes on one project, the 15 m dishes on another project, or only use the 8 x 8 m correlations plus the 8 x 15 m correlations. The 8 x 8 and the 8 x 15 m correlations by themselves will give snapshots with 1190 baselines --- three times the number of baselines of the VLA !

1. Good for mapping EXTENDED objects. Of these extended objects, the two highest priorities will be the molecular spectroscopy and dust continuum in the galactic and extragalactic research. Both of these domains require short-spacing information. The southern-hemisphere location gives the best possible viewing point for studies of our Galaxy, and we should be sure we have the short-spacing information for extended galactic sources. This will be guaranteed if we have some 8 m dishes.

2. Can have minimum spacings in the range of 10 to 12 meters, allowing the angular scale coverages listed in Table 1, for all types of project.

3. For imaging projects that mainly need the 15-meter dish sub-array, the 8-meter dishes could nevertheless provide complimentary information by facilitating the recuperation of the missing flux at short spacings.

4. For the dust continuum studies at 350 microns, we expect the atmosphere to be unstable, even on the excellent site proposed. It may not be possible to recuperate the necessary missing flux information from total-power maps with a single dish. The only solution is to do interferometry with small antennas, that is, 8 m antennas.

--- For these purposes, 20 x 8-meter dishes would suffice. It is not necessary to have 40 small dishes.

1. Good for mapping COMPACT objects.

2. Plenty of collecting area: 50 x 15 m = 9,000 m ².

3. Gives more calibrators, i.e., a wider choice of calibrators.

4. Gives good signal-to-noise on calibrators.

5. Gives good sensitivity for the sources being studied, especially on the long baselines.

6. For some projects at longer wavelengths, the 15-meter dishes, operating in total power mode, can provide the short spacings for the 8-meter dishes. For these projects, the large dishes may eliminate the close packing constraint.

7. For galactic sources with several bright cores within the field of view of the 8-meter dishes, the 15-meter dishes can give complementary long-baseline information by mosaicing within the 8 m dish primary beam. This can also be done quickly for bright-source projects that need only snapshots.

8. The 15 m dishes have a larger primary beam when correlated with the 8 m dishes. Because the correlation of the 8 m dishes with the 15 m dishes is a correlation of their voltage patterns rather than their power patterns, the primary beam for these baselines is larger than the primary beam for the 15 m x 5 m correlations (see Table 2).

9. Given the two constraints of getting a collecting area of 10,000 m ² , and having the possibility of short spacings as well, the use of 15 m dishes minimizes the number of antennas. A crucial consideration for people who must run, maintain, and upgrade interferometer arrays is to keep the total number of antennas small. This is very important for the feasibility and cost of upgrades, since in practice, mm and sub-mm receivers and correlators always should be replaced with next-generation versions, to profit from progress in new detectors, semiconductors, superconductors, faster electronics, etc. This will be impossible if the number of antennas is too large.

10. Large antennas help you to point. The pointing requirement must be specified as a fraction of a beamwidth, e.g., 1/30th of a beam at 300 GHz, 1/10th of a beam at 900 GHz. However, the signal-to-noise on a pointing measurement with an array goes as D ² n ^0.5 , where D is the dish diameter and n is the number of dishes. Hence to reach a similar precision, it is more advantageous to have fewer large dishes than to have many small antennas.