resolution.
Report of the Sun and Stars Working Group
One of the most interesting scientific problems which the MMA can address is that of solar flares. In particular, the so-called gamma-ray/mm-wave flares have been identified as some of the most challenging to existing ideas regarding particle acceleration and energy transport. Electrons and ions are accelerated almost simultaneously to very high energies with electrons attaining energies of 10-100 MeV within 1-2 sec of flare onset. The electrons emit both millimeter waves and continuum gamma-rays of high intensity. Recent work with BIMA [9,11,14,17] suggests that i) flares of all sizes produce MeV electrons on prompt time scales; ii) that the MeV electrons appear to form a distinct population of fast particles from those producing microwaves and HXRs. The first of these results comes as something of a surprise. It implies that gamma-ray/mm-wave flares may represent no more than the extreme tail of the flare distribution; i.e., that all flares are, to some degree, gamma-ray/mm-wave flares.
The MMA will be the *only* goundbased instrument capable of providing high-quality snapshot maps of impulsive nonthermal emission from the evolving distribution of relativistic electrons during flares. In probing emissions from fully relativistic electrons on the Sun, the MMA will provide valuable constraints on the evolution of the electron distribution function in space and time and hence, on acceleration and transport mechanisms. We note that the circular polarization properties of the mm-source will be of value in constraining the magnetic field in the source. Furthermore, since the emitting electrons are essentially relativistic the source is expected to emit a linearly polarized component as well.
Thermal emissions from flares are also of great interest. It is not yet clear whether direct bombardment by an intense flux of nonthermal particles is the dominant source of energy into the chromosphere or whether an ion-acoustic conduction front also plays a role. In order to distinguish between various possibilities, a time resolution of 100 ms is necessary; it should be sufficient to resolve electron time-of-flight from the propagation of a conduction front in many cases.
The solar chromosphere is the thin layer above the temperature minimum region where non-radiative heating first becomes manifest. In previous years detailed semi-empirical models of the chromosphere [4,15] -- largely based on UV/EUV line and IR/submm continuum observations -- have been constructed. It has become apparent in recent years that these models are must be revised to reflect the highly inhomogeneous and dynamic nature of the chromosphere [1]. The study of the solar chromosphere is an area where the MMA can make important contributions. Given that the two sources of continuum opacity (H and H- free-free absorption) are well understood and that the continuum forms in LTE, mm- and submm-wavelength observations provide a convenient, linear thermometer with which to probe all layers of the chromosphere. The importance of the MMA is that it will provide high-angular-resolution maps on short timescales so that the thermal structure of the chromosphere can be mapped as a function of optical depth.
The MMA observations will be useful for understanding problems now
emerging from helioseismic research connected with the structure of
the solar chromosphere. Frequency shifts of solar modes with the
solar cycle appear to be related to thermal variations in the
structure of the outer solar atmosphere (i.e., chromospheric plages).
Submm- and mm-wavelength observations are essential for reliable
thermal modeling of the low chromosphere. The MMA should also be used
to observe thermal variations due to local chromospheric oscillations
concurrently in different frequency bands [12]. These observations
would be useful as a thermal diagnostic of the chromosphere based on
its response to the hydromechanical waves that cause the brightness
variations [8]. Submm observations with 1 arcmin resolution show
thermal variations of roughly 10 K. Statistics done with Doppler
observations suggest that the temperature amplitude increases in
proportion to the inverse root of the resolution. So we might expect
to see thermal variations of order 75 K with 1 resolution.
1.3 Birth and Death of Solar Filaments
Solar prominences and filaments are cool (chromospheric temperatures), dense, filamentary structures which form in the low corona along magnetic neutral lines. They are presumed to form via a thermal instability, but the details remain unclear. By the time they are seen in absorption in spectral lines at visible wavelengths, the formation process is largely complete. Filaments are optically thick to wavelengths longward of a few mm. The filament channel is seen well-before H-alpha filaments form. Given that filamentary structures appear in a number of astrophysical contexts, the ability to study the formation of solar filaments and prominences is of general interest. Once formed, comparison of high spatial resolution observations at mm- and cm-wavelengths can relate the geometry of the filament and its surrounding coronal cavity, leading to a determination of the density--temperature structure of the sheath and coronal cavity. Finally, some filaments are ejected from the Sun in spectacular eruptions. The destabilization and eruption of filaments in eruptive flares is of particular interest.
Dupree [4] first proposed that dielectronic recombination in the solar
corona and transition region leads to a significant overpopulation of
high-n states of certain ions such as O VI. Searches for solar RRLs in
the early 1970s [2] yielded no detections. Corrections to Dupree's
work revised the expected line-to-continuum ratios downward by
significant factors. When pressure and Zeeman broadening are included,
these estimates fell further still. A recent search with the BIMA
array again yielded no detections (Grossman & White, private comm.).
Nevertheless, for ions with Z>4, recombination lines may yield to the
extremely high sensitivity of the MMA. Their detection would represent
direct verification of dielectronic recombination processes in the low
corona and transition region, and would offer the possibility of
measuring the magnetic field strength in those regions through the
Zeeman effect. More probable is the detection of hydrogen radio
recombination lines at lower n. Clark et al. [3] recently reported the
detection of the H19-alpha line in the 350 
m band. If a submm
capability is supported, RRLs may become an important diagnostic of
conditions in the corona and transition region.
It appears that with improved receivers and a superior site
(Chajnantor), a gain by a factor of four or more in the sensitivity
(over that anticipated in 1990) is expected. The number of detectable
stars is proportional to 
, where 
is the limiting
sensitivity. Instead of detecting a few hundred stars, as anticipated
during the last MMA Science Workshop, nearly 5000 stars could be
detected with only 10 min integration each. An hour of integration on
each increases the number to nearly 100,000 stars! The impact of the
MMA will therefore be profound. The MMA will dominate radio studies of
these objects. Studies will no longer be restricted to a handful of
objects in each class. Statistical samples will be available with the
advent of the MMA.
Many observations of continuum emission from stars have now been made at mm- and submm-wavelengths by the JCMT, IRAM, SEST, the NRAO 12m, and other telescopes. Examples include the detection of Be stars [16], the detection of Wolf-Rayet and OB stars [10], the study of symbiotic stars [6,13], and observations of Nova Cygni [7]. These objects have turned out to be as strong as, or stronger than, expected at mm-wavelengths and confirm the projection that the MMA will dominate radio studies of these objects.
Studies of mass loss from hot stars will particularly benefit from the MMA. Mm-wavelengths probe the same region of the wind where UV lines originate. Furthermore, mm-wavelength observations are not corrupted by nonthermal components (as are microwave measurements). More accurate mass loss rates should therefore result. We also point out that the interpretation of mass loss from cool giant stars should also benefit from the MMA, particularly the Mira variables, for which both line and continuum observations of these pulsating stars are required.
The MMA will be an important addition to the study of the earliest stages of outbursts on novae and recurrent novae. Studies of the 1985 outburst on RS Oph indicated the presence of both thermal and nonthermal components at microwavelngths. The MMA will be critical in disentangling the two in future events of a similar nature. Studies of Nova Cygni 1992 [7] at mm-wavelengths indicate that the mm-wavelength emission is sensitive to the details of the mass loss at the earliest stages of the outburst. Again, the MMA will play a key role in probing the earliest stages of nova outbursts.
Recent VLBA and MERLIN observations of novae and symbiotic stars have shown structure that was initially complex and optically thick, but subsequently evolved into optically thin shells and jets, respectively. High resolution mm-wavelength imaging will enable the study of the density and temperature structure of the circumstellar emission from the photosphere outwards, not only in the outer, optically thick layers.
A number of stellar objects display vigorous solar-like activity. dMe stars, active binaries (RS CVns and Algols), and weak-lined T Tauri stars all produce flares which are much more energetic than those on the Sun. As on the Sun, magnetic fields are suspected of playing a key role in the storage and release of flare energy. The spectral maximum of the incoherent gyrosynchrotron emission is expected to be several 10s of GHz on these stars. Sensitive multi-band, time-resolved observations of stellar outbursts will enable constraints to be placed on stellar coronal magnetic field strengths. Furthermore, as on the Sun, observations of mm-wavelength emission from active binaries and weak-lined T Tauri stars will probe the most energetic electrons in the emitting distribution, thereby shedding light on the relevant acceleration processes. We note that accurate measurements of circular polarization will be required here.
X-ray binaries and X-ray transients will also benefit from the MMA. Mm-wavelength spectral evolution measurements are critically needed for direct studies of radiating electrons at acceleration sites. Measurements of the linearly polarized emission are needed. For the X-ray binaries and transients, in every case where high resolution observation have resolved the emitting regions, one does not see the previously expected expanding synchrotron ``bubbles''; rather they exhibit roughly linear jet structures, indicating the potential of a mm array with 10-15 mas resolution for measuring not only the mm portion of evolving spectra, but also imaging the optically thin, non-scattered jet emission in X-ray binaries and X-ray transients.
As already mentioned, the scientific return of observations of both thermal (e.g., novae, symbiotics, cool and hot stars) and nonthermal sources (e.g., XRBs and X-ray transients) would be greatly enhanced if the objects were resolved. In addition to directly imaging the structure of winds and ejecta in the case of thermal sources, or the source structure at site of particle accleration in the case of CVs and XRBs, observations on baselines of 10-30 km could: i) measure the photospheric or chromospheric diameters of several hundred stars, thereby providing a direct measurement of the effective temperature in these layers; ii) observe the presence of starspots and/or stellar plages and determine rotation periods, important inputs to stellar dynamo theories. Therefore we strongly advocate the deployment of outrigger stations for use during 3-km array observations.
A preliminary investigation, where 6 of the 40 antennas are deployed
in a Y configuration to maximum baslines of either 10 or 30 km has
been explored (by R. Hjellming). The angular resolution provided by
the 3, 10, and 30 km arrays would be 
,
, 
, respectively, sufficient to
allow imaging down to the 10-15 mas level. Frequency synthesis may be
employed to fill in the the uv domain more uniformly. We estimate a
reduction in sensitivity of roughly a factor of four on the outrigger
baselines; hence the sensitivity of the long-baseline configuration
will be sufficient to resolve hundreds of objects. Note that outrigger
configurations would be possible on the Chajnantor site.
Antennas: the MMA antennas must be designed to point at or near the
Sun without a significant degradation in performance. We favor the
BIMA approach. The BIMA antenna panels are ``scalloped''. Hence,
while the surface accuracies are 8-10 
m rms -- significantly
better than the current 25 
m rms spec on the MMA antennas -- the
visible/IR part of the spectrum is scattered over an angle of about 5
deg. The BIMA antennas have therefore been used routinely for solar
observations with no degradation in performance. Thermal gradients
are also an important concern. These may be largest when pointing
directly at the Sun. Although a detailed study must be carried out, a
CFRP backstructure may be needed to reduce thermal gradients and their
effects. However, if performance at submm-wavelengths is supported, a
CFRP backstructure may be needed anyway.
Antenna Outriggers: the potential for stellar work by the MMA could be fully realized if at least six outrigger stations were constructed. These could be deployed around the 3 km array in a wye (or other) configuration to provide baselines of 10--30 km. The correlator must provide sufficient delay to accomodate long baselines.
Polarimetry: we wish to perform full Stokes polarimetry with an accuracy of approximately 0.1% for both solar and stellar observations. We advocate supporting circular polarization measurements in at least two bands: the 3 and 9 mm bands, which we anticipate will be the ``workhorse'' bands for flare studies.
Time Resolution: a minimum correlator dump time of 100 ms is required. We do not believe that this requirement is significantly different from the dump times anticipated for on-the-fly mapping experiments. These dump times would generally be employed for continuum work. The data rates are therefore not extreme.
Spectroscopy: two kinds of solar experiments are anticipated which require a spectroscopic capability: i) the measurement of linearly polarized radiation; ii) the observation of radio recombination lines. The former measurement is limited by bandwidth depolarization and internal Faraday rotation. Bandwidth depolarization can be overcome if the observation is made in spectral line mode with a resolution of 1 MHz over 128 channels. The latter effect will then be the dominant one, but linearly polarized flux at the few-percent level may occassionally be observable during flares. Radio recombination lines may be detectable at submm-wavelengths. They are expected to be broad and shallow and will therefore require a large spectrometer bandwidth (1-2 GHz) with coarse sampling (128 channels). Both types of experiment appear to be accomodated by exisiting proposals for the spectrometer.
Fast Frequency Switching: it is desirable for a number of experiments to be able to tune to a different frequency within a given band on a timescale of 1 s. It is desirable to switch frequency bands on a similar timescale.
Multi-band Observations: we support the idea of a dual-band observing capability for observations of solar flares, during which variations occur on a timescale faster than the electronics could be switched to another frequency or band. A dual-band capability is achievable through the use of a dichroic mirror and/or the use of subarrays. Our goal is to obtain 3-point spectra through the course of a flare event.
Gain Control and Field-of-View: the antenna temperature will be about
5000 K when pointing at the quiet Sun. Since the system temperature
will be about 50 K, the signal must be attenuated by roughly 20 dB for
solar observing. The VLA uses a switched 20 dB attenuator -- a
similar scheme could be employed by the MMA. essentially of of the
quiet Sun could be observed in this mode using mosaicing or OTF
mapping techniques. More challenging is the active Sun. During the
impulsive phase of a flare, the system temperature can increase to a
few times 
to 
K over a timescale of several seconds to several
10s of seconds. Accurate and calibratable automatic gain control is
probably needed. With the aforementioned numbers, a total of 30--40 dB
of gain reduction will be needed at times. Another problem is the
antenna FOV, which is 32
. A enlarged FOV is desirable --
at a wavelength of 1 mm -- in order to improve the detection
efficiency of flare events. An enlarged FOV reduces the forward gain
of the antenna which feeds back into the problem of gain control.
Submm Performance: submm performance is desirable for i) access to radio recombination lines in the solar atmosphere; ii) probing the temperature minimum region.
Instrument Siting: The Chilean site at Chajnantor is clearly superior to the Mauna Kea site by any number of measures. It appears to be the only site for which design and planning for performance at submm-wavelengths is warranted. Furthermore, it is the only site under consideration on which outrigger antennas can be easily accomodated.
REFERENCES