Continuum emission

IMAGER naturally handles many spectral channels. However, the astronomer may be interested in broad-band continuum emission. Extracting the properties of such a continuum emission may be a difficult issue in some cases. IMAGER offers a number of tools to do so, including the notion of a CONTINUUM image.

To first order (over a limited frequency coverage), continuum emission can be represented by a a flux (at some reference frequency) and a spectral index. Both flux and spectral index may vary spatially. Such a representation is often valid over a factor 2 or 3 in frequency. In general, the spatial distribution of these properties of the continuum emission may be widely different from those of the spectral lines. Furthermore, spectral lines can be easily optically thick at some velocities, and may hide the continuum emitted from behind, and thin at others. The continuum itself may sometimes have sufficient opacity to obscure line emission.

Thus a proper extraction of the continuum properties can formally only be done in the image plane, after imaging the combined continuum and spectral line emission, and cannot rely on a separation in the uv plane at the level of the visibilities.

In general, this cannot be done by simply imaging channel per channel. The reason for that is that continuum emission is in general fainter than the spectral line one, and often faint enough to have only limited signal to noise in a single channel. Deconvolution of such emission is noise limited. Yet, when averaging over many channels (PolyFix on NOEMA offers at least 2048 channels per wide band window, 16384 in survey mode, ALMA has up to 4096 per spectral window), the S/N changes by a large amount, and the undeconvolved sidelobes from the continuum emission appear.