DRAFT 1.0

7 May 2001

Contribution to the ALMA Future Correlator Science Case

Alain Baudry

(with contributions from S. Bontemps, J Braine, C. Cecacrelli)

 

 

 

 

 

The Future Correlator will provide two major improvements to ALMA with respect

to the first generation correlator:

a- a large bandwidth with the highest possible spectral resolution,

b- a highly flexible use of the 64 telescopes (sub-arraying).

In this document we wish to describe some potential main scientific drivers for such improvements. The following list is biased towards star formation topics and some extragalactic examples. We certainly

do not wish to be exhaustive.

The first section presents some general scientific cases and we add a second section with working examples to make our case clearer. The third section is on redshift determination.

 

1. General scientific cases

---------------------------

1.1 Search for macromolecules and/or pre-biotic molecules

Given the great impact to the possible origin of terrestrial life that the discovery of aminoacids would have, glycine and sugars have been long searched for in molecular clouds. The targets have preferentially been the protostars because they are the densest and warmest sites in the clouds and consequently also the brightest molecular line emitters. The search with single dish telescopes has been so far unsuccesful and

reached a dead lock point, as the line confusion limit has been reached. On the other hand, observational and theoretical considerations suggest that very probably the most complex molecules are formed on the grain surfaces via active grain chemistry. Those molecules are then injected into the gas phase in the so-called Hot Cores of protostars, compact warm regions where the grain mantles evaporate. Note that if formed on the grains those molecules may (at least partially) survive up to the phase of disk formation. If grain mantles are desorbed then at the surfaces of disks, for example because of X-ray irradiation from the central source, we may observe the macromolecules in the gas phase even towards disks.

The implications for such searches are so far reaching that they should be primary goals of the ALMA project. The high spatial resolution of ALMA will allow to filter out the line emission from the cloud and to screen the spectra of the Hot Cores/disks. Since such large molecules have plenty of lines in relatively large bands (for example the glycine-II has about 60 transitions in a 16GHz band around 100GHz) their detection rely on the detection of the largest possible number of transitions to positively identify their presence.

Note that at the same time a high spectral resolution is required (0.1-0.2 km/s) in order to reduce the line confusion problem and make the identification more robust. Therefore, a large bandwidth associated with a high spectral resolution are necessary to carry out the searches of macro and/or pre-biotic molecules.

1.2 Line surveys

The necessity of a large bandwidth associated with a large spectral resolution is also requested to obtain unbiased spectral surveys. It is well known that gas cooling in molecular clouds, with or without newly born stars, is primarily due to atomic and/or molecular line emission in the Far Infra Red to sub-mm wavelengths range. Relatively heavy molecules, like for example formaldehyde or methanol, have their low lying lines at millimetre to sub-millimetre wavelengths. Depending on the gas temperature and density either the low lying

or high lying energy lines of the same molecule can be excited, so that multi-frequency studies of a same molecule are used to constrain theseparameters together with the column density of the species. Simultaneous observations by different chemical species provide the chemical composition of the gas. Thus, knowing which atom/molecule emits at what wavelength is of paramount importance for studying the gas thermal balance and the physical, dynamical and chemical structure of the gas in molecular clouds. (Unbiased surveys of spectral line emission in the millimeteter bands accessible from the ground have been carried out for a number of sources to study how the physical and chemical conditions change in different environments.)

The ALMA spatial resolution and sensitivity will allow to undertake such studies on smaller scales, namely on compact regions, like for example protostellar disks. So far, just a few molecules have been detected towards the disks, because unbiased surveys are out of reach of the current instrumentation.

Nevertheless, this is the kind of study which will allow to track the history of the protostellar disks, specifically the evolution of the gaseous component. Important questions awaiting for an answer will likely come from unbiased line surveys towards disks. we may cite:

why, how and when the gas condenses; what is the real influence of the protostellar harsh environment (e.g. X-rays) on this evolution;

what is the chemical composition of the protostellar disks, or in other words, the initial conditions of the planet formation problem.

Previous experience has shown that the highest possible spectral resolution is necessary to minimize the merging of lines and that the largest bandwidth is necessary to reduce the observing time

as well as to reduce "spectra" overlapping questions.

 

1.3 Overall structure of molecular clouds harboring protostars

It is now clear that the molecular clouds are strongly influenced by the presence of embedded newly born stars and/or protostars. At the same time the formation of a new star depends on the parental molecular clouds. Studies of the interplay between protostars and parental clouds are thus extremely important for understanding the process of star formation and the evolution of molecular clouds at the same time. These studies require simultaneous observations of different tracers: lines probing the cold gas of the cloud; lines probing the denser condensations; lines probing the warm envelopes of protostars and lines probing the disks; lines probing the interaction of the outflows emanating from protostars with the surroundings; and finally observations of the dust continuum (in a large frequency band).

The great flexibility of the proposed Future Correlator will allow us to design specific observations for specific cases, making possible to study simultaneously the dust and gas components with the highest possible

efficiency.

For example, sub-arrays can be used to observe the continuum with a moderate spatial resolution on one hand, and lines probing the protostellar Hot Cores/disks with a higher spatial resolution, and finally the outflows with a third different and appropriate spatial resolution; all this simultaneously. Although these studies will also be possible with the present Baseline Correlator they will take more observing time with probably less efficiency.

If the project indeed needs so many lines, of course the project is better executed with the Future Correlator.

ALMA sensitivities must be matched to the line strengths, as the observing time is constant. In some cases two settings/two integration times might be better suited to the science—C17O and CS seem somewhat mismatched below, for example.

2. Working examples

-------------------

2.1 Structure of a protostellar envelope/disk at 500 pc

Protostellar size: ~ 5000 AU (~ 10 arcsec at 500 pc)

--> one then expects 2 or 3 protostars in the field of view.

Resolution down to 15 AU (0.03 arcsec) --> it fully probes the infalling envelope and even reaches the accretion disk zone.

Sensitivity (1h) down to about 1 Mjupiter in continuum(8 GHz).

In the warm part of the protostellar envelope some important coolants are observable in the 330-360 GHz band: the CH3OH and SO2 bands are especially interesting because they provide forests of lines to derive accurate rotational diagrams. The lines are narrow and a typical spectral resolution of ~0.1 km/s is required to resolve the infall velocity field.

A typical 1 GHz wide spectral coverage is required to image these forests of lines. With double side band mixer tuning, one can cover both the 338GHz CH3OH band and the 357GHz SO2 band, together with other interesting lines such as C17O(3-2), C34S(7-6), CS(7-6), or HCO+(4-3):

LSB (8 GHz): 336 - 344 GHz

C17O(3-2), C34S(7-6), CH3OH band, CS(7-6)

USB (8 GHz) : 352 - 360 GHz

HCO+(4-3), SO2 band

With the baseline correlator, one can achieve a rather good spectral resolution (0.22 km/s) using the 8 digitizers to cover 8 x 250 MHz (and not distinguishing between the two polarizations). However, it is then impossible to image at the same time the additional 4 interesting lines and to image the continuum.

On the other hand, better spectral resolution and all what we scientifically wish is obtained at the same time with the Future Correlator.

One could for instance program the following configuration: with 4 digitizers we can achieve a resolution of about 0.1 km/s resolution across 4 x 500 MHz and with the 4 remaining digitizers and associated basebands we can image 3 lines and the continuum in one 2 GHz baseband; or image 2 other lines and 2 continuum bands in order to derive a dust spectral index image.

More windows always provides more lines at greater expense. Note that with 2SB mixer tuning one reaches the same lines but with better sensitivity.

2.2 Investigation of the Structure of a (Giant) Molecular Cloud at 5 kpc

Pre-stellar/protostellar structure size:

~ 5000 AU (1 arcsec at 5kpc)

Sensitivity down to about 0.01 Msun in dust continuum

Field of view (30 arcsec) --> 0.8 pc

Sensitivity in ~ 10 hours and 8 GHz bandwidth a few 1000Msun

 

Objective: to survey in one observing session the density and velocity structure of the cloud, including low density medium (CO lines), star forming dense cores (CS, H2CO), and pre-stellar/protostellar condensations (continuum, N2H+), together with the molecular outflow streams (wide band coverage of several CO and CS lines including (2-1)/(1-0) line ratio for 12CO and C17O; and (5-4)/2-1) ratio for C34S and CS).

Spectral resolution requirement: velocity dispersion of the molecular cloud structure is typically 1 to 2 km/s.

The resolution required is of the order of 0.1 km/s. Since the outflow streams can reach up to a few 10 km/s, the full coverage should reach ~ 200 km/s which then requires at least 2000 resolution elements in CO and CS. In addition, to get rid of the possible confusion with other molecular clouds on the line of sight, a coverage of at least 50 km/s (500 resolution elements) is required for the density tracers (CO, CS, H2CO, N2H+).

We use an 8GHz double side band mixer tuning to cover (for instance) the following ranges and lines:

100GHz/LSB: 92 - 100 GHz

N2H+(1-0), C34S(2-1), CS(2-1)

100GHz/USB: 108 - 116 GHz

C18O(1-0), 13CO(1-0), C17O(1-0), 12CO(1-0)

230GHz/LSB: 223 - 231 GHz

C17O(2-1), H2CO, 12CO(2-1)

230GHz/USB: 239 - 247 GHz

C34S(5-4), CS(5-4)

We use the Future Correlator and its sub-arraying capability in the following manner (the two sub-arrays are configured so that they give similar synthesized beamwidths around 100 and 230 GHz):

100 GHz band (extended sub-array 1)

LSB: 3 x 125 MHz (3 basebands each narrowed down to 125 MHz) + 1 x

2GHz continuum band

USB: 1 x 500 MHz (to cover both C18O and 13CO with moderate

resolution)+ 2 x 125 MHz + 1 x 2GHZ continuum

230GHz band (sub-array 2)

LSB: 3 x 250 MHz + 1 x 2GHz continuum

USB: 2 x 250 MHz + 2 x 2GHz continuum

Both bands are thus imaged simultaneously with the same synthesized beam and line resolution /coverages are 0.09/360 km/s in both bands.

The baseline correlator would do this same experiment with a 2SB receiver design—the frequency coverage is the same but of course the 2SB receivers admit no noise from the other sideband and provide better sensitivity.

For the baseline correlator, the 8 IF windows would have to be deployed in single polarization mode to cover the same setup with 4 IF polarization pairs:

100 GHz band (extended sub-array 1)

LSB: 2 x 125 MHz (2 basebands each narrowed down to 125 MHz [N2H+(1-0), C34S(2-1)]) + 1 x 2GHz continuum band

USB: 1 x 500 MHz (to cover both C18O and 13CO with moderate resolution)

230GHz band (sub-array 2)

LSB: 3 x 250 MHz [C17O(2-1), H2CO K=1 J=3-2, 12CO(2-1)] + 1 x 2GHz continuum

USB: 3 x 250 MHz [C34S(5-4), CS(5-4), CH3OH band] + 1 x 2GHz continuum

Or with dual polarization windows:

100 GHz band (extended sub-array 1)

LSB: 3 x 125 MHz (3 basebands each narrowed down to 125 MHz) + 1 x 2GHz continuum band

USB: 1 x 500 MHz (to cover both C18O and 13CO with moderate

resolution)+ 2 x 125 MHz + 1 x 2GHZ continuum

230GHz band (sub-array 2)

LSB: 2 x 250 MHz [H2CO K=1 J=3-2, 12CO(2-1)] + 1 x 2GHz continuum

USB: 1 x 250 MHz [CH3OH band]

 

3. Redshift determination of distant dusty sources

--------------------------------------------------

ALMA will be a major instrument in cosmology and perhaps the only capable of determining precise redshifts for dusty distant galaxies. At high redshifts, the dominant population of galaxies may well be the dusty ones because the star formation is present over a smaller area but at a much higher level than in today's universe. At redshifts greater than about 3, the CO lines represent the main means of accurately determining redshifts and we are certain that in dusty star-forming objects these lines will be strong. The CO (rotational) lines are separated by 115 GHz in the rest frame so at a redshift of, say, 5, they will be separated by 115 GHz / (1 + z) = 19 GHz. With only a few frequency settings, ALMA will quickly be able

to detect two lines, thus precisely determining the redshift. Such searches will use the whole 8 GHz bandwidth and both polarizations for maximum sensitivity. The searches will also be done at fairly low frequency (typically 3 -- 4 mm) because at a redshift of 5 the corresponding rest frequency is about 80 GHz * (1 + z) = 480 GHz. This is between the CO(4--3) and CO(5--4) transitions which are respectively at energy levels of 55 and 83 K above the ground level, requiring warm dense gas for excitation as they are far above the temperature of the microwave background Tbg = 2.73 * (1 + z).

Searches near, say, 230 GHz would measure the CO(12--11) line at a redshift of 5 and such high transitions (430 K above ground) are very difficult to excite, such that these lines are expected to be quite weak and not appropriate for redshift determination.

Based on Table 10.3 of the ALMA project book, the Baseline Correlator will have a resolution of about 60 km/s at 80 GHz covering the whole 8 GHz width in both polarizations and NOT keeping cross-products.

The proposed enhanced correlator should have a resolution of about 2 km/s under the same conditions. The difference is potentially most important because galaxies close to face-on, and small systems in general, have line widths of about 100 km/s and sometimes significantly less, resulting in dilution in 60 km/s channels whereas 2 km/s is well adapted.

A galaxy with a line width of 100 km/s may only be "detected" in one channel at 60 km/s resolution. With a resolution of 2 km/s (and adding some smoothing eventually) the line profile would be clearly identifiable as that of a galaxy.

For normal galaxies close to face-on, and with no disturbance (this would be an unusual galaxy at z=5!) the baseline correlator resolution in redshift machine mode is marginal. However most distant galaxies appear disturbed. For disturbed galaxies, e.g. Arp 220, the intrinsic linewidth is near 70 km/s, well-matched to the baseline correlator single channel resolution. Note that all confirmed high-z galaxies have linewidths measuring near 400 km/s or higher and that relatively few galaxies are face on. Because of the small ALMA beam, redshift surveys may be best accommodated by total power telescopes—the GBT, LMT and CSO. Each of these telescopes has planned redshift machine instrumentation; for each of them the resolution adopted is on order of 30 MHz, the same as the baseline correlator for ALMA. At the end of this decade new bolometers will have produced numerous interesting mm-only continuum sources but the bulk of redshift determination for these will probably be carried out on monolithic telescopes.

At Band 1, however, the resolution of the baseline correlator becomes marginal particularly for small undisturbed systems at high z. For absorption searches against continuum, the future/enhanced correlator also provides an advantage (and this is a mode advantageous for an interferometer).

Other lines, particularly the 158 micron CII line and possibly lines such as the CI (492, 806 GHz), NII (205, 122 microns), and OI (145, 63 microns) if strong enough, are not as useful as the CO lines because they are isolated. Once a single CO line is found one may search for these rather than another CO line, of course.