JWST Mosaic Overview
JWST mosaic observations, a series of pointings covering an area larger than an individual instrument field of view, can be specified using parameters in the relevant APT observation templates.
See also: APT Mosaic Planning
JWST offers the capability to mosaic regions that are larger than a single field of view (FOV). This is useful not only for imaging instrument modes but also for integral field units (IFUs) in NIRSpec and MIRI MRS. A mosaic is made up of a series of pointings, called tiles, each of which may be dithered. These mosaics can be planned in the Astronomer's Proposal Tool (APT).
Table 1. APT science templates allowing mosaics
Wide field slitless spectroscopy (grism)
|MIRI||Medium resolution spectroscopy (IFU)|
|MIRI||Low resolution spectroscopy|
|NIRSpec||Integral field unit (IFU)|
|NIRISS||Wide field slitless spectroscopy (grism)|
MIRI MRS is a special case since the field of view sizes for the 4 MRS wavelength ranges are different from each other.
Time-series modes, coronagraphy, NIRSpec multi-object spectroscopy, and templates for observations of individual targets are not useful for mosaics; to avoid confusion, mosaics have been disallowed in those templates.
Mosaics are defined by a single coordinate, an assumed tile footprint (size and orientation), and a designated number of rows and columns of the tile footprint. Each instrument and mode that allows mosaics have a FOV size and reference axis defined in the Science Instrument Aperture File (SIAF), which is the official repository for all defined instrument-related positional and angular definitions.
If a dither pattern is defined for a given tile, it is the combined region covered by the FOV and the dithers that define a tile footprint for a given mosaic. This is the footprint that should be considered when addressing the overlap of tiles in a given mosaic. Because the various instrument FOVs have different shapes and dimensions, it's useful to always view their defined mosaics in the Aladin viewer in APT to assess the overall shape and coverage of their mosaic.
Because the overall mosaic footprint will rotate with time, it's helpful to define your mosaic footprint to be approximately symmetrical on the sky. If a non-symmetrical mosaic is needed to cover a particular target, the user should fix the position angle of the mosaic and define the tiles to cover the region of interest.
However, it may or may not be possible to find guide stars simultaneously for all tiles in a highly constrained mosaic. Also, fixing the orientation of mosaics essentially requires fixing the time of the observations, and many such observations can overly constrain observation scheduling. Users should only fix the position angle of mosaics when necessary for science. Consider obtaining insight into the available range of position angles using one of the JWST Target Visibility Tools.
For imaging mosaics, it is usually the case that the separation of the tiles is large enough that each tile will be a separate visit† (requiring a separate guide star and GS acquisition activity). For simplicity in this discussion, assume this is the case. Because each visit (tile) requires its own guide star (and thus a guide star acquisition), part of the schedulability assessment is that all tiles must simultaneously have guide stars available in order to be declared schedulable. With the addition of Gaia DR2 stars to the guide star system, experiments have shown that the problem of missing guide stars for mosaic tiles has nearly disappeared, but we cannot guarantee that this problem will not occur. As a general rule, the larger the mosaic, the more likely it is that one or more tiles will be missing guide stars at the same time as the rest of the tiles, and hence there will be no time when the entire mosaic can schedule simultaneously. In such cases, APT declares the entire mosaic as unschedulable because, by definition, the entire mosaic is a single observation.
There are ways to remove a tile or tiles into separate associated observations to fill the gap(s) in the original mosaic if needed. However, another option is to consider a pattern of smaller overlapping mosaics to cover larger regions, leaving each of the smaller units free to rotate and schedule at different times when guide stars are available for that portion. Some experimentation is required to understand the effectiveness of this option.
† This is not always the case, as it depends on such things as the amount of overlap you specify and the visit splitting distance calculated by APT, which is a function of the target's Galactic latitude. However, since APT performs the splitting of proposed observations into visits, it is a detail that only becomes an issue when a tile needs to be removed or split from a parent mosaic.
For many smaller spectroscopic fields of view, such as the NIRSpec IFU and MIRI MRS (also an IFU), mosaicking is allowed but the situation is handled quite differently by APT. The footprints of most mosaics with these small FOVs will remain within a region of a single guide star's availability, and so the entire observation stays within a single visit. (Recall that the visit splitting distance changes with ecliptic latitude of the target, but even the smallest values of visit splitting distance (30"–40") would require a huge IFU mosaic to need more than a single visit.) IFU mosaics in APT take advantage of a reorganization of the activity ordering to greatly reduce mechanism motions and improve efficiency. These changes occur by default and require no special handling in APT, but you should be aware of the difference as it is more efficient than large imaging mosaics which may require multiple visits.
Mosaics versus dithers
See also: JWST Dithering
The mosaicking capability for JWST is intended to provide users with a convenient way to cover larger regions of the celestial sphere. It should not be applied to make very small steps in position, to mimic normal dithering. While this could, in principle, result in somewhat lower overheads, in practice, it is a false gain.
Filter mechanisms have the potential to be life-limiting factors for JWST instrumentation. This is particularly the case for NIRCam, which serves as the wavefront sensing imager for the telescope. As a consequence, observers must not circumvent standard observation logic that minimizes filter wheel moves. Programs with either MIRI or NIRCam that involve dithered observations or “mini-mosaics” (that can be observed within a single visit) should step through the dither positions with a single filter before moving to a second filter. Creative alternative ordering for improving overhead efficiency will not be allowed.