JWST Slit Spectroscopy
JWST provides slit spectroscopy in 2 instruments: Near Infrared Spectrograph (NIRSpec) and Mid-Infrared Instrument (MIRI).
Spectroscopic slits are narrow apertures in the optical path designed to allow light only from the intended scientific target onto the detector array. These slits are usually narrow in the dispersion direction, which limits the size of the target in the wavelength direction and improves the spectral resolution. JWST offers spectroscopy in fixed slits in 2 instruments, NIRSpec and MIRI, which generate spectra covering wavelength ranges from 0.6 to 5.3 μm and 5 to ~12 μm, respectively.
Available spectroscopic slits
NIRSpec offers slit spectroscopy in 5 apertures, which are described in detail in NIRSpec Fixed Slits Spectroscopy. MIRI offers slit spectroscopy in a single aperture with the low-resolution spectroscopy mode.
Table 1. NIRSpec and MIRI spectroscopic slits
|Instrument||Slit||Size (arcsec)||Size (pixels)|
|NIRSpec||S200A1, S200A2, (S200B1)||0.2 × 3.2||2 × 32|
|NIRSpec||S400A1||0.4 × 3.65||4 × 36|
|NIRSpec||S1600A1||1.6 × 1.6||16 × 16|
|MIRI||LRS||0.52 × 4.7||5 × 43|
The NIRSpec S1600A1 slit is optimized for time-series observations.
Summary of available options
The NIRSpec slits can produce spectra with resolving powers of roughly 100, 1,000, and 2,700. At the lowest resolution, a full spectrum from 0.7 to 2.3 μm can be generated in one grism setting. At the higher resolutions, 4 settings are required for full wavelength coverage.
The LRS on MIRI produces a spectrum from 5 to 12 μm, with spectral resolving power increasing from ~40 at 5 μm to ~200 at 12 μm. The long wavelength cut-off for the LRS is a soft number because the prism transmission decreases rapidly from 10 μm to longer wavelengths while the resolution grows higher. The quality of the flat fields and photometric calibration begins dropping past 10 μm and will not be reliable past 12 μm.
Comparison to ground-based slit spectroscopy
The primary advantage of spectroscopy from JWST is that telluric absorption does not limit the wavelength coverage, making it possible to observe in spectral regions inaccessible from the ground and to obtain continuous spectra from 0.6 μm all the way to the LRS sensitivity cut-off past 12 μm. In addition, since JWST is unaffected by atmospheric turbulence, the spectroscopic slit truncates less of the light from a target and the angular resolution in the cross-dispersion direction is closer to the diffraction limit. Finally, the cold operating temperatures of JWST and the resulting low backgrounds from the telescope leads to much better mid-infrared sensitivity than can be achieved from the ground.
Dithers and nods
Dithering in slit spectroscopy may help mitigate the effects of bad pixels, detector effects, and improve spatial and spectral sampling. In addition, the different pointings can provide background observations around point (or compact) sources. When the step size is relatively large and exposures can be pairwise subtracted, this background removal strategy is typically referred to as nodding.
NIRSpec fixed slit spectroscopy offers both nodding with large-scale offsets and subpixel dithering. A variety of options are available for both types of dithers. The user can select either one strategy, or combine both.
The MIRI LRS template offers, for the LRS slit, an ALONG SLIT NOD* option, where the target is placed alternately at 30% and 70% of the slit length (approx. ±0.9" or 8.25 pix from the slit center). A MAPPING mode is also offered, where the user can define a number of offsets in the spatial and/or spectral direction. One of these two dither types must be chosen. These dither types can be combined with mosaic settings to perform a dither pattern at each mosaic pointing position.
For more information, please see these articles:
* Bold italics style indicates words that are also parameters or buttons in software tools (like the APT and ETC). Similarly, a bold style represents menu items and panels.
It is inevitable that the spectroscopic slit will truncate some part of the point spread function (PSF), leading to transmission losses in the resulting spectrum. These losses are a function of wavelength because the size of the PSF is a function of wavelength. Transmission losses also depend on the location of the source within the slit, with greater losses closer to the edges of the slit. Given the importance of source placement in the slit, target acquisition is recommended for slit spectroscopy observations.
For JWST, these pointing errors can be expected to be on the order of 7 mas (the precise value of the mean error depends on the distance the telescope moves). Pointing errors on this scale are almost negligible compared to the size of even the smallest fixed slit, 200 mas on NIRSpec. Consequently, it's expected that the transmission function (transmission vs. wavelength) will be almost the same from one spectrum to the next and thus straightforward to correct in the pipeline.
A couple of caveats are important. First, the transmission function for an extended target, even if it is compact, will differ from that for a point source, which could lead to more significant issues, especially if the target is mispointed. Second, mispointings are more likely when an offset source is used for target acquisition. In that case, any errors in relative positions between the source chosen for target acquisition and the scientific source could seriously impact the quality of the spectrum. As an example, even relative positional errors of only 50 mas are halfway from the centerline of the smallest NIRSpec slit to its edge, if the user is unlucky enough for the error to be in that direction. For this reason, target acquisition on the science target is usually the safest option, provided that it is a point source.
Scattered light and saturation
The NIRSpec slits are relatively safe from scattered light because they are in the gap between the 2 halves of the MSA and most of the shutters will be closed.
For the MIRI LRS, on the other hand, contamination from the imager field of view should be considered. During LRS observations, any sources in the imager field will also be dispersed. Because of the broad passband of the LRS prism, these sources will easily saturate in the field. Very bright or saturated sources can cause detector artifacts along rows and columns around the bright source. These can, in principle, cause systematics in the LRS spectrum during slit spectroscopy observations. Very bright sources located in the imager field may also cause some light to scatter into the LRS slit spectral region. If possible, observers should avoid having very bright sources in the imager portion of the detector field of view whilst exposing with LRS, in particular in the region immediately adjacent to the slit (this can be checked using the Aladin visualization option in APT). If they find a potentially troublesome source, it may be necessary to constrain the roll angle of the telescope to keep this source off of the imaging array during an integration. The full detector read out will be available to observers, allowing any such sources, if present, to be identified, which can help with data analysis.