The combination of the upgrade in sensitivity and the increased angular resolution provided by the A+ configuration opens an extremely exciting area of investigation that cannot be exploited by any other existing or planned instrument - milli-arc-second resolution imaging of thermal radio emission. This capability is placed in context in the angular size versus brightness temperature plot in Fig. 3.3. The two diagonal dashed lines in the figure show the brightness temperatures required to produce flux densities of 100 mJy and 1 mJy within a given angular radius. Curves for the most powerful interferometric arrays characterize their maximum resolution (horizontal portion) and maximum sensitivity (diagonal portion). For a source with a given angular radius and brightness temperature to be resolved by an array (and therefore imaged), it must lie above the curve. It is an interesting astrophysical result that at a brightness temperature of a few K, radio sources can be broadly divided into non-thermal and thermal. Photo-ionized, nebular sources of thermal radio emission generally have equilibrium, kinetic temperatures of a few K. Hence, compact nebular radio sources have brightness temperatures at this value or lower. Photospheric emission from stars, which is optically-thick at kinetic temperatures of K, straddle this K boundary. Figure 3 shows the location of the stellar main sequence and giant branches for stars at a distance of 1 kpc.
Figure 3.3: Brightness temperature and angular size regimes accessible to imaging by interferometer arrays. The previously inaccessible region that will be opened up by an expanded VLA with the upgrade in sensitivity is shown by the shaded area.
It is clear from the diagram that, with current arrays, imaging at milli-arc-second resolution is restricted to high brightness temperature, non-thermal, radio sources. For brightness temperatures just over K, the MERLIN array provides images a resolution of tens of milli-arc-seconds, however, as brightness temperature drop below this value, the minimum angular dimension over which detectable flux can be obtained increases rapidly. Consequently, imaging at milli-arc-second resolution of the entire class of thermal-emitting radio sources is an area of astrophysical investigation that has yet to be exploited. To access this area requires order-of-magnitude improvements in both sensitivity and resolving power. The A+ configuration, in combination with the enormous improvements in sensitivity, will provide that facility. The new region of observational parameter space that would be accessible to imaging by the A+ configuration is illustrated graphically by the shaded region in figure 3. This region encompasses a wealth of phenomena in stellar astrophysics that have hitherto been revealed only as ``point sources'' of radio emission.
Angular resolution of a few milli-arc-seconds corresponds to a linear dimension of a few AU at a distance of one kpc. This distance is typical of the binary separation of semi-detached interacting binaries containing giant stars, such as symbiotic stars and recurrent novae. Other examples of thermal radio radio sources generated or affected by phenomena on AU scales are the winds and circumstellar environments of early-type stars and the very early stages of novae outbursts. At current resolving power, imaging investigations of these types of objects have been largely constrained to examination of ejected material that has propagated to large distance from the point of origin before becoming resolved. The A+ configuration will finally allow imaging of thermal radio emission on the scale of the interaction that gives rise to ejecta. Fig. 3.4 shows the maximum distance to which these and other types of object would be resolved by both the current and the A+ configuration. For the current VLA, typical distances range from a few tens to a few hundred parsec. Unfortunately, the space density of these objects is sufficiently low that there are essentially no examples within this resolvable distance. For the A+ configuration, the resolvable distances increase to a large fraction of a kpc to several kpc. Red supergiant photospheres can be resolvable to beyond a kpc, the binary separation of `D' type symbiotic stars and the very early stages of nova ejections can be resolved out to several kpc, and the circumstellar disks of Be stars out to several hundred pc. At these distances, significant populations of all of these types of sources can be probed on AU and smaller scales for the first time.
Figure 3.4: The resolving power of the current and A+ configuration compared to the linear dimensions of several known, thermal radio emitting astrophysical phenomena.
Dimensions of order 1 AU have crossing time of order 1 yr for velocities of 10's of km/s. Hence, milli-arc-second resolving power will allow direct measurement of the angular expansion rate of slowly expanding objects such as compact planetary nebulae and wind interaction shells in the ISM. Combined with spectroscopic data, such measurements can provide accurate estimates of distance for a significant population of objects.