NIRSpec Target Acquisition Recommended Strategies
Help is provided to select from the available JWST NIRSpec Target Acquisition (TA) options for a given observation. Practical advice for choosing the parameters is also given in a variety of science scenarios.
NIRSpec target acquisition (TA) options are described in the NIRSpec Target Acquisition article, and this should be consulted for technical details of the acquisition methods themselves.
This article is intended to provide advice on which of the available TA options to select for a given observation, as well as practical advice for choosing the parameters. Normally after the initial slew to the target, a guide star will be acquired by the Fine Guidance System (FGS) to correct the overall JWST pointing, (see Figure 1). Because only a single guide star will be acquired by one of the FGS detectors, the precision of the initial spacecraft roll will depend on the star trackers. The accuracy with which the desired science aperture will be placed on the sky will depend primarily on the accuracy of the guide star coordinates, but also on the separation between the science aperture and the guide star, and the accuracy and stability of the instrument-to-FGS alignment calibration.
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.
Pointing accuracy before target acquisition
For JWST instruments it is estimated that, without an instrument-level target acquisition, the pointing accuracy for cycle 1 will be between 0.3'' and 0.45'' (1σ radial error; post guide-star acquisition pointing accuracy), depending on the distance between the science instrument and the guide star. At the location of the S1600A1 aperture used for WATA, a more typical error is expected to be ~0.34''.
For Cycle 1 planning purposes it is currently recommended that for NIRSpec observations the observer should assume a 1σ radial pointing uncertainty of about 0.34''. This corresponds to about 0.22'' per axis.
It is expected that most NIRSpec observations will require significantly better placement of the target in the NIRSpec aperture plane that can be supported by this expected initial pointing and so will need to include a WATA or an MSATA target acquisition.
The coordinates for a majority of guide stars are anticipated to be updated using the Gaia DR2 release prior to JWST launch. When one of these guide stars is used, it is likely that the blind pointing accuracy will be considerably better than described above. However, a substantial fraction of observations will still need to rely on guide stars not in the Gaia catalog, and so all observations should be planned assuming the initial pointing performance described here.
Implications of this expected pointing performance for WATA observations and planning of IFU observations without a target acquisition are discussed below.
Typical time requirements for acquisition options
Detailed overheads for NIRSpec operations are discussed in the NIRSpec Overheads article, but applying those results in practice requires a detailed understanding of the operational sequences. When adding an acquisition, in addition to the time for the acquisition itself, there will often be additional overheads for instrument and detector changes between the acquisition and science observations. For a given observation, total overheads are best determined by using APT, but we give here approximate times from a few examples to illustrate typical costs for adding an acquisition.
Adding VERIFY_ONLY in place of TA = NONE will typically require 8 to 14 minutes or more, depending on the exact exposure parameters selected. Using MSATA will typically add between 24 and 30 minutes to the observation. Adding WATA to a Fixed Slit observation will typically require between 8 and 15 minutes depending on the exposure setup, while total overheads for adding WATA to an IFU observation tend to be about 3 minutes higher than for the Fixed Slit template. These overheads are likely to change near-term after further ground testing is completed. The most up-to-date overheads will always be found in the current version of APT.
WATA specific considerations
WATA is available with the Bright Object Time Series (BOTS), Fixed Slits (FS), and Integral Field Unit (IFU) templates, and with the Micro-Shutter Assembly (MSA) template when observations are created directly instead of with the MSA Planning Tool (MPT). For fixed targets, the acquisition target and science target may differ, but for moving targets, WATA must directly acquire the planned science target.
WATA compactness requirements
NIRSpec TA algorithms acquire the target by finding the flux-weighted center of the brightest 3 × 3 pixel, (0.3'' × 0.3'' on the sky) checkbox in the 32 × 32 pixel region that covers the footprint of the 1.6'' × 1.6'' aperture. If the acquisition target is extended, predicting the results of WATA will require understanding the spatial structure of the source in the TA bandpass at a resolution comparable to or smaller than the 0.3'' checkbox size. This may require pre-existing HST or JWST imaging of the field. Note that even given such information, there are currently no tools available from STScI to perform a detailed simulation of the acquisition.
The use of WATA to directly acquire extended sources that lack a clearly defined central bright spot or for which the spatial structure is poorly understood on 0.1'' to 0.3'' spatial scales is not recommended. In such cases, it may be better to use a nearby point source as the acquisition target.
Acquisition target isolation
To ensure that the correct object is acquired, in the bandpass used for the acquisition exposures the acquisition target should be the brightest compact object within a radius of ~2''. That is, no nearby 0.3'' × 0.3'' region should be comparable to or brighter than the acquisition target. If it is necessary to observe a faint target close to a brighter object, either an offset WATA acquisition or MSATA should be used; although in many cases observing both objects simultaneously with the IFU may be the best approach.
Required accuracy of WATA acquisition target coordinates
To provide a high (> 99%) probability that the target is sufficiently within the S1600A1 aperture slit at the initial blind pointing to allow WATA to succeed, it is recommended that the acquisition target’s coordinates, including any corrections for proper motion, be on the ICRS system and accurate to at least 0.15'' 1σ radial, (or equivalently about 0.1'' 1σ per axis). If the acquisition target and the science target differ, the precision of the final centering will also be limited by the relative accuracy of the two targets’ coordinates. If it is impossible to provide coordinates of the required precision for a fixed target, pre-imaging with a JWST or HST imager to obtain improved coordinates is strongly advised. For moving targets with an ephemeris too uncertain for WATA, users might consider using TA = NONE with the IFU and select a small enough dither pattern to account for the expected coordinate uncertainty.
The pointing improvements supplied by any instrument target acquisition only apply as long as the same guide star is used to control the pointing. This normally means that a new acquisition is needed for each visit and observation, although in some circumstances the use of target groups may permit multiple observations in a sequence using the same guide star and a single initial target acquisition. For WATA observations, this means that the desired science pointings must all be within visit splitting distance of the target acquisition exposure and that at least one guide star must actually be available that can support all the needed spacecraft pointings.
WATA using an offset acquisition target
For targets where a direct acquisition using WATA is not an option either because the source is not in the required brightness range or is too extended, acquisition of a nearby offset star may provide an alternative. Note that use of an offset target is only allowed for fixed targets; this option is not available for moving targets.
Impacts on schedulability
Adding an offset target may require an additional slew or a longer slew between the target acquisition and the science exposures. This may reduce the number of guide stars available to support the visit.
Additional error sources when using an offset TA target
Any relative error in the coordinates of the acquisition target and the science target will directly affect the centering of the science exposure. In principle, the coordinates for both targets should come from the same image (with the same relative astrometric accuracy, ideally from HST or JWST imagers). In addition, errors in the spacecraft roll will add a pointing error that scales with the angular separation between the two targets, e.g., for an expected 1 σ 60'' roll error, a 30'' target separation would lead to offset of the science target location in the NIRSpec aperture plane equivalent to 9 milliarcsec.
Practical advice for choosing an offset acquisition target
There are three points to keep in mind when choosing an offset target:
- The relative coordinates between the science and acquisition target must be known to sufficient accuracy to support the desired final pointing requirements. This includes consideration of the proper motion of both targets.
- The brightness of the TA target in the acquisition bandpass must be well enough understood to confidently estimate the exposure level. Remember that WATA can support acquisitions of point sources with an ABmag as faint as 25.6 (or J band magnitudes of about 24.9).
- The required slews between the TA and science exposures must be less than the visit splitting distance and small enough that a usable guide star can be found for one of the FGS detectors. If possible, it is best to pick an offset target that would be within a few arc-seconds of the science target. APT computes the worst case slew (between the TA and science exposures) to determine if the visit splitting distance is violated for the selected offset acquisition target. If it is, APT will also calculate the range of aperture position angles that would satisfy the constraint and will instruct the user whether an added special requirement would work. See more information about this in the section on Considerations for WATA acquisition for the IFU, below.
Ideally, an HST or JWST IR image would be available that allows the relative coordinates of the two targets to be measured in the same frame, direct determination of the IR flux of the offset star, and verification that the intended offset star is really a simple point source. In practice, it will often be necessary to use catalog information to supply some of this information and to check the targets for significant proper motion.
The Aladin interface in APT allows information for a number of catalogs to be over-plotted on images of the target field and the JWST aperture planes and approximate separations between objects to be easily estimated. Available catalogs in this interface that may be especially useful for choosing an offset target include the 2MASS Point Source Catalog, which includes JHK photometry and accurate coordinates, (but not proper motions), and the Gaia-DR1 catalog which should contain highly accurate coordinates for more than a billion stars as faint as 21st magnitude, although proper motions are currently only available for a subset of these. The Gaia-DR2 released in Spring 2018 is not yet included in the latest GSC version used by APT and Aladin. GSC 2.4.2, the planned release which will include Gaia-DR2, will become available in May or June 2019. This will provide highly accurate position and proper motion data for most of these stars.
Acquisition strategies by template
Bright Object Time Series template
For the BOTS template, the allowed TA options are WATA or NONE. Because good aperture centering is essential to achieve the desired photometric performance, and because spatial offsets much larger than 20 milli-arcsec can shift the zero-point of the wavelength scale by more than the nominal wavelength accuracy, WATA should be used if at all possible. If the science target is too bright to be directly acquired with WATA, the observer should attempt to find a suitable offset star to use for the acquisition. It is recommended against using the BOTS template with TA = NONE, as the blind pointing has a significant chance of putting the target too close to the edge of the 1.6'' × 1.6'' aperture. Even with a modest offset of 0.3-0.4" from the aperture center could cause, extremely low levels of telescope spatial jitter (7mas) could cause a signal modulation that is comparable to or greater than an exoplanet transit depth. Hence, if a BOTS target does not have a nearby offset star, then it may be preferable to use WATA on the saturated science source. In this case, executing the target acquisition on a slightly saturated PSF is estimated to give better performance than the blind pointing. Even if the science target will saturate in the shortest WATA exposures, we still recommend still executing WATA rather than using TA=NONE. However, if the JWST blind pointing turns out to be significantly better than the projected 0.3-0.45" value, then it may be possible to remove the TA on a saturated star.
Fixed Slits template
For the FS template, WATA, MSATA, and NONE are the available TA options. Normally pointing accuracy of 20 milli-arcsec or better will be desired with the FS template, and this will require a WATA or MSATA target acquisition. If using TA = NONE for an observation of a point source, even for the wide 1.6'' x 1.6'' S1600A1 aperture, there is a significant chance that the target will fall close enough to the edge of the aperture to significantly vignette the PSF wings, especially at long wavelengths. Likewise, for the smaller 200–400 milli-arcsec wide slits, there is a good chance the target will miss the aperture altogether. In addition, with TA = NONE there is no direct way to measure the displacement of the target in the dispersion direction, and this will cause significant uncertainty in the wavelength zero-point. For extended targets or fixed slit mosaics, more relaxed pointing requirements may sometimes be applicable depending on the details of the science scenario. It may make sense to consider TA = NONE when WATA or MSATA are not possible, but otherwise use of TA = NONE with the FS template is not advised.
Choosing between TA options for NIRSpec FS observations
If the science target itself or a nearby offset star is suitable for direct acquisition, WATA should normally be preferred. Note that WATA can be used to acquire sufficiently compact moving targets. When it is possible, advantages of WATA include:
- No need to provide a catalog of faint reference stars with coordinates accurate to 15 milli-arcsec relative to the science target.
- No need to wait for an assigned aperture position angle before finalizing observations.
- If directly acquiring the science target, WATA can correct the final pointing for modest uncertainties in the target’s position.
- Significantly less overhead time for WATA compared to MSATA. (Overheads will be larger for WATA with offset targets, due to the need to perform an additional slew after the target acquisition).
If WATA is not practical for a particular FS observation, then MSATA should be considered. This does require that a catalog of reference stars visible through the MSA be provided, and very accurate planning coordinate precision between these reference stars and the science object. Advantages of MSATA include:
- MSATA is independent of science target’s structure or brightness.
- MSATA is always done at the first science pointing, so no extra slew is needed.
- More tolerant of a larger than expected error in the initial spacecraft pointing (3.2'' square subarrays for each reference star vs 1.6'' square aperture for a single target).
- Reduces the error in the spacecraft roll from an expected 60'' rms to no more than 7'' rms. For the FS template, this may be important if significant slews are involved, or if multiple observations in a target group depend on a single acquisition.
Depending on the exact scenario, observers of extended targets may be willing to accept a pointing accuracy coarser than the normal 20 milli-arcsec accuracies expected from WATA or from MSATA with a highly accurate reference catalog. This might allow the use of MSATA with a reference catalog of coarser accuracy, or perhaps even the use of TA = NONE. For moving targets that are unsuitable for WATA, TA = NONE will be the only choice with the FS template. In such cases, the observer might consider whether an IFU observation with its coarser pointing requirements might be a better choice. If use of a narrow 200 or 400 milli-arcsec wide fixed slit is required for an extended fixed or moving target observed with TA = NONE, a mosaic in the dispersion direction could be used to map out the object, but this would need to be padded on each side by an amount sufficient to cover the possible pointing error. We recommend adding at least 0.5'' on each side in the dispersion direction for this purpose.
Integral Field Unit template
The IFU template supports the MSATA, WATA, VERIFY_ONLY, and NONE options for target acquisition. Many science observations will not require precise centering of the target at an exact position in the IFU ~ 3'' × 3'' FOV, and this can lead to relaxed pointing requirements relative to other NIRSpec templates. However, if considering blind pointing without a TA, it is still necessary to take into account the pointing accuracy, target size and dither pattern used. While choosing between WATA and MSATA involves many of the same considerations discussed above for the FS template, there are a few differences that should be considered.
Considerations for WATA acquisitions with IFU:
- The ~32'' separation between the S1600A1 aperture used for WATA and the IFU center may require a significant slew between the ACQ and science observations
- This may push the allowed limits of the visit splitting distance and guide star availability. To assist in using WATA, beginning with the APT 25.4.2 release in January 2018, the visit splitting distance for all moving targets was increased to 38'', while for fixed targets a minimum visit splitting distance of 38'' is also used whenever WATA TA is selected in any NIRSpec template. This allows WATA to potentially be used for any NIRSpec IFU observation.
- If the science is being done in any aperture other than the S1600A1 and the science and acquisition targets differ, then the required slew length will depend on the spacecraft orientation. If possible, keeping the scalar sum of the target separation plus aperture separation smaller than the visit splitting distance is advantageous, as it ensures that the visit will work at any aperture position angle.
It may sometimes be necessary to restrict the aperture position angle range to put a suitable acquisition target near the S1600A1 when the science target is in the IFU. This can potentially minimize the size of the needed slews at the cost of restricting scheduling. APT recognizes that WATA may work at a restricted position angle. When the sum of the target and apertures separations is greater than the visit splitting distance (the worst case), APT will generate an error message similar to the following: ''The worst case slew between the acquisition target and the farthest science target is 57.427 Arcsec (target separation=24.736 Arcsec, aperture separation=32.690 Arcsec). This exceeds the maximum of 38.000 Arcsec to maintain guiding at this latitude. Setting a position angle for the observation will reduce the worst case slew. If this observation is restricted to 165.934 to 328.780 degrees V3PA, then all activities will be within the limit." APT requires you to specify this explicitly with the special requirement on this observation, not via orient links with other observations.
- As noted above, the overheads for WATA are significantly lower than those for MSATA.
For MSATA acquisitions with the IFU:
- Unlike WATA, no additional slew is required between the TA and the first science observation, and visit splitting limits will not normally be a concern.
- Reduced coordinate accuracy will often be sufficient for IFU observations, allowing reference catalog coordinate accuracy as coarse as 50 milli-arcsec relative to the science target as opposed to the < 15 milli-arcsec catalog accuracy recommended for most FS/MOS template observations.
IFU Observations without a TA:
The source size, target coordinate accuracy, dither pattern, and the wavelength of the observation will affect the needed pointing margins when designing an IFU observation using either the VERIFY_ONLY or NONE options which do not perform a target acquisition.
The NIRSpec PSF FWHM increases with wavelength from 0.03'' to 0.16'' over the NIRSpec bandpass. For the best data quality at long-wavelengths it may be useful to keep the target at least 2 X FWHM away from the edge of the aperture. The radius of the target should be added to this margin.
Even with perfect pointing, the two and four-point nod patterns put the target as close as 0.71'' from the aperture edge, reducing the available margin for source size and coordinate or pointing errors.
Observers skipping the TA may wish to consider using the first two to four points of one of the cycling patterns or the four-point dither as a smaller alternative to a full sized nod pattern. Observers will find all the relevant information in the NIRSpec IFU Dither and Nod Patterns article.
If relying on blind pointing without a TA to observe a source that is small compared to the IFU field of view, the selected NIRSpec IFU dither pattern needs to allow room for the PSF width, the source size, and the accuracy of the target coordinates. For the preliminary estimate of the blind pointing performance discussed above, the minimum 1 σ radial target coordinate accuracy needed to give a 95% chance that all four dither positions will fall within the desired margin from the IFU edge is provided in Table 1.
Table 1. Required target coordinate accuracy for dithering with blind pointing
Recommended 1 σ radial coordinate accuracy (″)
Large cycle (1–4)
Medium cycle (1–4)
Small cycle (1–4) or 4-point-dither
† This is the minimum margin in arcseconds from the position of the source to the IFU edge needed for the PSF width and source size.
For the Cycling patterns, it is assumed that the first four points in each pattern are being used. Requirements for the success of 2-point nods or the first two points of the cycling patterns are similar to those for the 4-point patterns. This table is based on preliminary estimates of the JWST blind pointing accuracy. If in doubt, users should select one of the smaller dither patterns.
- Observers of solar system objects, in particular, are advised to be sure of the accuracy of the target ephemeris used and are responsible for ensuring that it is of sufficient precision to support the planned observations. This responsibility applies even when the moving target ephemeris is provided through APT.
- When using an IFU mosaic without a target acquisition to map a region, it is recommended to allow for as much as 0.5'' pointing error in any axis. Similar margins should be allowed for pointing offsets between visits, as they may use different guide stars.
Using VERIFY_ONLY with the IFU:
This mode does not perform any target acquisition. Instead, a verification image is taken at the final science pointing with the IFU aperture open and a user-selectable MSA configuration. In the undispersed light, the MSA quadrants and the IFU slices project onto different locations on the NIRSpec detectors and in principle this allows the IFU field of view to be precisely aligned relative to field objects imaged through the MSA, (although currently there are no tools to support this alignment in the standard pipeline processing).
When not performing a WATA or MSATA target acquisition, it may be important to include such a verification image to allow registration with other JWST/non-JWST observations if the morphology of the source doesn’t allow for accurate astrometry from the collapsed IFU spectral images alone.
For sources where precise alignment on the sky can be done from the IFU spectral images, an additional image of the IFU in undispersed light may be redundant and in such cases the user might wish to consider using TA = NONE rather than VERIFY_ONLY. Alternatively, if a direct IFU image is still desired, it is possible to use VERIFY_ONLY with an ALLCLOSED MSA configuration.
While imaging through open MSA shutters will not interfere with the direct IFU image, very bright sources imaged through the MSA may result in persistence that can affect subsequent NIRSpec observations. In such cases, it may be desirable to use a custom MSA configuration to block the bright stars in an image while still allowing fainter stars to be seen. However, an MSA configuration other than ALLOPEN or ALLCLOSED will require an aperture position angle assignment entailing a planning process similar to MOS observation planning. That process requires the use of MPT to design an MSA configuration at the last pointing of every visit. For this reason, and because of increased demand on the scheduling system, the use of protected MSA configurations should be a last resort.
Using TA = NONE with the IFU:
- This is normally recommended for cases where the absolute registration of the source on the sky is not critical or the registration can be done from the collapsed IFU spectral image alone.
- Most moving target observations not using WATA should use TA = NONE as trailed sky verification images will be difficult to interpret, and because most Solar System objects will be sharp-edged enough to allow the collapsed spectral cube to be used for image registration.
Multi-Object Spectroscopy template
Visits generated from a plan
- Most MOS visits will be generated using the MSA Planning Tool (MPT), and will normally use MSATA to precisely place sources in individual micro-shutters. The standard recommended TA method is the MSATA.
- For a few science cases generated with a plan, blind pointing using VERIFY_ONLY or TA = NONE may be adequate if the sources are sufficiently extended and exact placement of the micro-shutters on the sky is not needed.
- WATA is not available for observations generated from the MPT, and the MPT cannot generate observations for moving targets.
MOS visits not generated from a plan
It is also possible to specify MOS visits directly using the MOS template rather than generating them with the MPT. When specifying a MOS observation without the use of MPT, supported TA options include WATA, VERIFY_ONLY, NONE, and MSATA. The option of not using MPT is provided for a few special cases like moving targets, fixed extended targets, and targets that may not require the pointing accuracy delivered by MSATA. It is expected that most of these cases will use an MSA configuration forming a "long slit" from a column of open shutters. The inclusion of two built-in longslit configurations, and their associated MOS pointing centers (Q4 field point 1 and Q4 field point 2) within APT makes the specification of these observations straightforward. Q4 field point 1 is 28" away from S1600A1, allowing WATA to be supported for suitable targets. However, Q4 field point 2 is at too large a distance from the S1600A1 aperture to support WATA, and therefore is expected to be used primarily with VERIFY_ONLY or NONE. Finally, we note that caution should be used when selecting MSATA for MOS programs generated outside of MPT. A catalog will have to be included to select reference stars for MSATA.