NIRSpec MSA Target Acquisition

The NIRSpec target acquisition (MSATA) process uses reference stars to center targets within the MSA, IFU, or FS apertures for science observations. In cases where few reference stars are available, it is possible to use other compact sources with relaxed stellarity.

On this page

See Also: NIRSpec Target Acquisition Recommended Strategies

Words in bold are GUI menus/
panels or data software packages; 
bold italics are buttons in GUI
tools or package parameters.

The NIRSpec target acquisition (TA) method MSATA will be used for most NIRSpec science observations performed with the micro-shutter assembly (MSA) in the MOS observing mode, and is optionally available for the NIRSpec fixed slits (FS) and IFU observing modes.

During the MSATA procedure, a set of TA reference objects (typically stars) are observed through the open micro-shutters. 

Imaging data used to derive the astrometry for your catalog of reference stars and science targets needs to be carefully vetted against Gaia stars. In particular, the MSA target acquisition procedure has a very limited ability to correct for errors in the initial roll, so it is strongly recommended that the rotation of the reference stars on the sky be aligned with that of the Gaia frame to better than one arcminute (60", 0.0167°) in rotation. Roll offsets larger than 500" (0.14°) are likely to fail. Care should be taken to consider any proper motion of the Gaia stars and reference stars to avoid introducing a spurious roll offset larger than this. This is required to produce the best astrometry possible, and to correct for any lingering small roll offsets that may affect target placements.

The achievable TA performance also depends significantly on the relative astrometric knowledge of both the reference stars and the science source positions used to plan the TA and the science. The MSATA procedure is designed to work best when the relative astrometric accuracy of the target field is in the range of 5 to 50 mas and the absolute rotation of the catalog is aligned to the Gaia frame within 60 arcseconds of roll or less. The optimal TA accuracy of about 20 mas requires a relative astrometric accuracy of 5–10 mas in the planning Catalog. 

MSATA is not available for the bright object time-series (BOTS) observing mode because (1) the wide aperture TA available for this template in the JWST Astronomers Proposal Tool (APT) is equally accurate for a single target and takes less time to execute and (2) the required full frame detector readout would likely saturate on any BOTS science sources.

Operational sequence of MSATA

The MSATA procedure uses a set of reference stars distributed over the MSA quadrants, determines their observed centroids on the detector, and compares them to the desired positions to calculate a small-angle maneuver that adjusts the initial pointing and position angle. The process acquires 2 exposures to centroid reference stars in order to accurately align the science targets of interest within their apertures. The centroids of 5 to 8 reference stars will be measured in a full frame imaging exposure obtained through the "all open" MSA quadrants (or a protected configuration to block all operable shutters except those with reference stars) using both detectors NRS1 and NRS2

The 2 exposures are offset by one-half of an MSA shutter "pitch" (the distance between adjacent shutter centers) in both coordinates. This “MSA half facet slew” is necessary to properly derive the location of the reference stars, minimizing the effect of the physical bars between micro-shutters that could otherwise bias the measured positions of the stars. The description of the micro-shutter assembly presents additional information on the MSA structure. The NIRSpec observation planning tool software will preferentially choose reference stars that are not affected by failed closed micro-shutters.

Both NIRSpec MSATA images are processed by the onboard software to determine precise locations of the reference stars. These locations are initially measured in pixel coordinates and must be transformed by the software into the distortion-free tangential coordinate system on the sky before they can be compared to the desired sky positions. This coordinate transformation requires precise knowledge of the combined distortion effects of the telescope (Optical Telescope Element; OTE) and NIRSpec, as well as the measured tilt of the NIRSpec target acquisition imaging mirror. The measured centroids of the reference stars will be transformed to sky coordinates and compared with Catalog positions expressed in the same sky frame to determine offsets for each reference star and a combined mean pointing offset including a possible roll adjustment.  

These quantities are calculated using a least squares methodology. After a sigma-clipping and threshold outlier rejection procedure to determine the mean offset and residuals, the telescope will be offset by the final derived mean slew value. The procedure may be repeated once to improve centering accuracy if either the overall offset exceeds a specified threshold (1.1") or if the correction to the roll is greater than 120". Subsequently, a reference image of the centered target field is automatically obtained to conclude the MSATA operational process. The overall timing to execute MSATA is estimated to be 2158 s per science visit in observation planning. When the program is finalized, this timing will be adjusted to take into account the detector readout pattern selected, and the number of reference stars actually used.

The available filters for these acquisition images are F110WF140X, or CLEAR. NIRSpec TA images are always acquired in full frame readout with Ngroups = 3 and detector readout pattern options of NRSRAPID, NRSRAPIDD6NRSRAPIDD1 and NRSRAPIDD2.  The last two are available for MSATA only.

Figure 1. Graphical description of NIRSpec MSATA

Schematic view of the MSA layout, with target acquisition reference stars shown in red, and science targets (blue) shown in their open MSA shutters (green). Some science targets are not observed because they lie in the same row as other sources (spectral overlap) or because their flux is attenuated by the MSA support grid. The target acquisition process acquires exposures and carefully measures the pointing of the reference stars to correct the pointing and position angle to the best science position.

Expected Accuracy of MSATA

Figure 2 shows the NIRSpec target acquisition planning uncertainty as a function of input Catalog astrometric accuracy, derived and updated from the estimated error budget definition. 

If target acquisition uses 5 or more reference stars, the estimated NIRSpec acquisition accuracy will be better than ~20 mas (<1/10 shutter width) if the input planning Catalog astrometric accuracy is 5–10 mas. The estimated delivered TA accuracies of better than 25 mas (called optimal TA) are achieved only if planning Catalog astrometry is better than about 15 mas for all reference stars and science sources. Delivered estimated TA accuracy of 25–50 mas is possible to achieve using input relative astrometry with 15–40 mas accuracy. 

An estimated TA accuracy level of 50 mas is 1/4 of an MSA shutter width (or 0.2" FS width), and this would result in very significant flux calibration (i.e., slit loss) uncertainty and probable degradation in wavelength calibration accuracy. Hence, this TA regime (called NIRSpec “relaxed accuracy TA”) should primarily be used for MSA science on very extended sources and use cases that can tolerate a reduced flux calibration accuracy. A TA accuracy beyond 50 mas is recommended primarily for IFU mode observations. (Nevertheless, WATA is usually preferred over MSATA for IFU observations.)

Table 1 describes these MSATA accuracy classifications based on input astrometric accuracies for reference stars and science targets. The Astronomer's Proposal Tool (APT) planning interface with the MSA Planning Tool requests an astrometric Catalog accuracy, and these MSATA classifications are presented to users when target acquisition is defined in visits.

Figure 2. Expected accuracy of the NIRSpec MSATA process as a function of planning Catalog accuracy

Expected NIRSpec target acquisition accuracy as a function of input planning Catalog (relative) astrometric accuracy. The multiple curves show the expected accuracy for the number of reference stars used in MSATA, from 5 to 20.

Table 1. Description of the "optimal," "relaxed accuracy," and "IFU only" categories used to describe the accuracies achieved in MSATA

Accuracy classification

Input catalog astrometric accuracy (mas)

Estimated TA accuracy (mas)




Relaxed accuracy



IFU only



Approximate brightness limits for MSATA

The brightness limits for the reference stars in MSATA observations are defined by the instrument filter used to acquire the images, and the detector readout pattern used for the exposures (IRS2 readout patterns are not used for MSATA or WATA.) All NIRSpec TA exposures are acquired with Ngroups = 3 (specified using the Acq Groups/Int parameter in APT), which also defines the achieved brightness range.

Table 2 provides the approximate limiting AB magnitudes of sources that might be used for MSATA, and the corresponding instrument settings. Magnitude limits for a source with a S/N equal to  20  and for a source near saturation were estimated based on pre-launch PSF and throughput predictions, for all available TA filters, and these values are shown in Table 2. In practice, the in-flight JWST PSF is significantly sharper than these pre-launch estimates and the effective throughput is higher. These updated estimates are now used for JWST ETC calculations and demonstrate that the estimated magnitudes limits shown here should in principle be fainter by typically 0.7 to 0.8 magnitudes. However, in practice, it does not appear that allowing reference stars slightly above the ETC estimated saturation limit causes actual issues. In part this is because the reference stars are rarely perfectly centered in a pixel, causing average peak count rates to be lower than ETC estimates, and also because the centroiding algorithm appears to be tolerant of modest amounts of saturation. At the faint end, the performance with reference stars fainter than the pre-launch estimates has yet to be established. For these reasons, as well as to maintain consistency for existing programs, the pre-launch limits shown here continue to be (or remain) the ones used by APT to determine reference star bins.  This is likely to be updated as we improve our understanding of MSATA performance.

In addition to the NRSRAPID and NRSRAPIDD6 readout patterns described in the NIRSpec detector readout patterns pages, MSATA has 2 additional detector patterns available called NRSRAPIDD1 and NRSRAPIDD2. These 2 patterns were added to more efficiently observe the magnitude range achievable for NIRSpec in the MSATA images, particularly for TA in extragalactic deep fields. The NRSRAPIDD1 and NRSRAPIDD2 patterns are available only with the CLEAR filter. The NRSRAPIDD6 readout pattern was created to replace NRS for improved cosmic ray rejection during target acquisition. (NRS can still be used in science exposures.)

Table 2. Brightness ranges for NIRSpec MSATA filter and readout pattern options

Readout modeS/N = 20Saturation






Table note: Limiting bright and faint magnitudes for target acquisition reference stars. Readout patterns NRSRAPIDD1 and NRSRAPIDD2 are only used with the CLEAR filter for TA exposures. All TA exposures are acquired with Ngroups = 3.  NRSRAPIDD6 replaced NRS for MSATA for improved cosmic ray mitigation. Numbers based on ETC default coordinates (RA = 00:00:00 and Dec = 00:00:00). Numbers may vary with coordinates.

Observation Planning for MSATA 

See Also:  NIRSpec MPT - MOS and MSATA Program UpdatesNIRSpec MPT - Catalogs, NIRSpec MOS and MSATA Observing Process

In the MOS, IFU, and FS science APT templates, MSATA is an option for target acquisition. MSATA is the recommended target acquisition method for most NIRSpec MOS science mode observations, especially when multiple targets need to be placed precisely in individual shutters. In all cases, the NIRSpec science observations will be defined in a standard way via the Astronomer's Proposal Tool (APT). The MSATA parameters are defined using a Catalog with reference stars, ingested in the MSA Planning Tool (MPT).  For MOS observers, this is the same Catalog that is used to define the MOS science observation.  More details are described in the NIRSpec MPT - MOS and MSATA Program Updates article.

NIRSpec MSATA parameters, including reference star selection and exposure parameters, are typically not defined at proposal submission. Science observations with the NIRSpec MSA, FS, or IFU that include target acquisition using the MSATA method must be oriented observations. However, the final execution orientation will be assigned after proposal acceptance, and observers will be asked to resubmit programs with fully defined TA parameters. Hence, MSATA reference stars do not need to be included in the Catalog for proposal submission, but must be included in final flight ready programs. An exception are programs with a fixed angle special requirement for which pre-imaging is already available at proposal submission.  In these rare cases, MSATA parameters could potentially be defined at proposal submission. Science observations that seek to restrict the aperture position angle are allowed at proposal submission, but discouraged because of their scheduling impacts. Ideally, if an orient constraint is desired, a range of angles no smaller than 30 observable degrees is recommended.

Latest updates

  • Under "Approximate brightness limits for MSATA", added text to better explain why we are still using pre-launch reference star magnitude limits in APT

  • Revised text to explicitly mention the 60" roll accuracy requirement for catalogs

    Content about MSA Program Updates moved to new article on the topic.

  • Added link to NIRSpec MSATA Reference Star Selection Recommended Strategies article.

    Added the last update about reference stars (above), and fixed a reference to the readout patterns available for MSATA.

    Changed the number of reference stars needed for MSATA, and the average timing value for MSATA with 8 stars.
Originally published