NIRCam Imaging APT Template

Instructions for designing JWST NIRCam imaging observations using APT, the Astronomer's Proposal Tool.

On this page

Imaging is one of the five NIRCam Observing Modes. Each mode has a corresponding template in the Astronomers Proposal Tool, APT, for users to design their observing programs. Instructions for NIRCam Imaging are given below. Complete listings of allowed values are documented in the NIRCam Imaging Template parameters page.



Step-by-step APT instructions

Generic

The following parameters are generic to all templates, and are not discussed in this article: Observation Number, Observation LabelObservations CommentsTarget Name, ETC Workbook Calculation ID, Mosaic Properties, and Special Requirements.

Module

Select ALL to observe with both NIRCam modules, returning data from the full field of view using all 10 detectors. To obtain data from only a single module or smaller detector subarrays, select Module B.

Subarray

When using Module B, users may opt to either read out all five detectors completely (FULL) or the smaller detector subarrays. Subarrays are read out more quickly than the full detector, providing brighter saturation limits for a given number of reads (see below). There are six subarrays available for imaging: SUB640, SUB320, SUB160, SUB400P, SUB160P, and SUB64P, where the number indicates the number of pixels on each side of the subarray.

The central subarrays were designed for bright extended sources (e.g., Jupiter or large star-formation regions). They return data from four short-wavelength detectors and one long-wavelength detector, with similar combined areas on the sky in each wavelength channel. Note that the is a gap in the coverage in the short-wavelength channel; these gaps can be covered with dithering or mosaics (see below). The corner subarrays are designed for small objects (named "P" for point source) and return data from just two detectors, one from each wavelength channel.

Figure 1. NIRCam detectors and imaging subarrays

Layout in the NIRCam field of view of the 10 detectors and Module B subarrays. The short and long-wavelength channels are shown in blue and red, respectively. Note each long-wavelength subarray covers 4 times the area as the corresponding short-wavelength subarray with the same number of pixels, because the short-wavelength pixels deliver twice the spatial resolution along each axis.


Dither parameters

Dithering (multiple exposures with shifted overlapping pointings) is required for NIRCam imaging. Dithering mitigates bad detector pixels and improves flat-fielding and PSF sampling by imaging each portion of sky with multiple regions of the detector. Larger primary dithers and smaller subpixel dithers also serve additional purposes described below.

Primary Dithers

Primary dithers serve to fill gaps in sky coverage in the field of view between the detectors and to mitigate flat field uncertainties. The primary dither patterns were designed for different purposes (and the choice is restricted depending on the module selected). 

  • FULL (Module = ALL): Large fields, covering all gaps between detectors and the ~45" gap between modules. Designed for use with mosaics (tiled pointings) of larger areas, but inefficient, requiring visit splitting and large overheads (see below).
  • FULLBOX (Module = ALL): More efficient than FULL. Covers a rectangular field of view without gaps when performing 4 or more dithers.
  • INTRAMODULE (Module = ALL or B): Targets smaller than the field of view of a single module (< 110" across). Covers the 5" gaps between the short-wavelength detectors, but not the 44" gap between modules.
  • INTRAMODULEX (Module = ALL or B): Similar to INTRAMODULE, but more efficient when performing 4 or more dithers.
  • INTRAMODULEBOX (Module = ALL or B): Covers two square regions without gaps when performing 4 dithers. More compact than INTRAMODULE(X), yielding more area with maximal depth.
  • INTRASCA (Module = B): Compact targets that can be imaged at the four corners of the detector (and in between) to mitigate flat field uncertainties. The object size should be smaller than the the field covered by an individual detector (<50" or <100" across for short or long-wavelength observations, respectively).
  • SUBARRAY_DITHER (Module = B; designed primarily for SUB64P): NIRCam Subarray Primary Dithers, spanning about ±0.6" in each axis (x and y), were designed to ensure that sources are observed in both wavelength channels when using SUB64P.
  • NONE (Module = ALL or B): Only recommended if used only in conjunction with an appropriate mosaic or subpixel dithering (see below).

When subarrays (smaller than the full detectors) are used, then fewer primary dither options are available:

  • INTRASCA (SUB640, SUB320, SUB160): as described above, available for extended-source subarrays only
  • INTRAMODULEBOX: as described above, available for all subarrays.
  • SUBARRAY_DITHER (designed primarily for SUB64P): NIRCam Subarray Primary Dithers, spanning about ±0.6" in each axis (x and y), were designed to ensure that sources are observed in both wavelength channels when using SUB64P.
  • NONE: only recommended in conjunction with subpixel dithering (see below)

When dithering, consider the JWST Slew Times and Overheads, including these detailed examples for NIRCam Imaging. Note that the FULL dithers always require visit splitting (new guide star acquisitions), increasing overheads significantly.

Dithering results in uneven depth (exposure time) across the final combined image. Only the INTRASCA dither patterns are designed to provide uniform depth (full coverage in every exposure) across the small area containing the science target.

Figure 2. Example NIRCam primary dither patterns

Exposure map for dithered observations in the short-wavelength channel. Corresponding long-wavelength observations are not shown. [Figures appear in three columns in a wide browser window.]

1) Top left: FULL with 3TIGHT Primary Dithers
2) Top right: FULL with 9 Primary Dithers and a Mosaic of 3 rows × 3 columns (or greater)
3) Middle left: FULLBOX with 6TIGHT Primary Dithers
4) Middle right: INTRAMODULEX with 4 Primary Dithers
5) Bottom left: INTRAMODULEBOX with 4 Primary Dithers
6) Bottom right: INTRASCA with 5 Primary Dithers. The science target should fit within one of the larger black regions.

Subpixel positions

Smaller subpixel dither patterns include subpixel shifts designed to optimally improve image sampling and resolution. This is especially useful for images with undersampled PSFs below the Nyquist wavelengths: 0.6–2 µm in the short-wavelength channel and 2.4–4 µm in the long-wavelength channel. Subpixel dithering is not implemented by default but may be added to an observing program by selecting more than one subpixel position.

In addition to the standard subpixel dither patterns, more compact small-grid-dither (SGD) patterns are also offered. SGD patterns are executed more quickly and precisely using the Fine Steering Mirror without slewing the telescope. They are limited in size to 0.06" or less, and are expected to be two times more precise than regular subpixel dithers. New SGD patterns were introduced in APT 25.4. They preserve the optimal subpixel sampling of the original pattern, while reducing overheads but offering less mitigation of bad pixels. Prior to APT 25.4.1, the SGD patterns designed for coronagraphy were available for imaging as well.



Filters and exposures

NIRCam uses a dichroic to observe simultaneously in the short-wavelength channel (0.6–2.3 µm) and long-wavelength channel (2.4–5.0 µm). For each exposure, users select two NIRCam Filters (one in each channel) as well as parameters which control the exposure time (identically for both channels) via readout patterns, which serve to reduce data volume. The exposure hierarchy is:

  • Exposure – with chosen filter pair; all instruments stationary and telescope locked on target
    • Integrations – each ends in a detector reset
      • Groups – saved data: average of one or more reads
        • Reads – non-destructive; charge continues to accumulate

Figure 3. NIRCam filters

Preliminary total system throughput for each NIRCam filter. Filters marked "P" are located in the pupil wheel, requiring transmission through a second filter in the filter wheel, either F150W2, F322W2, or F444W. In these cases, the combined transmissions are plotted. Use of F150W2 on its own (with a CLEAR pupil wheel filter) is discouraged. High photometric precision will not be supported in F150W2 due to the transmission dips and the variable PSF across the wide wavelength range.


Readout patterns

Use the Exposure Time Calculator (ETC) to determine which readout pattern, number of groups, and number of integrations will avoid saturation and/or achieve the signal-to-noise required for your science.

Multiple integrations may be most useful for brighter sources to avoid saturation. However for fainter sources, multiple dithers are generally preferable, with one integration per exposure per dither position.

Multiple groups are desirable to enable "up-the-ramp" fitting to observed count rates. This facilitates cosmic ray rejection, reduces the effective read noise for the integration, and increases the dynamic range of the final image (sampling count rates of bright sources before they saturate).

The 9 readout patterns are detailed below, including integration times for full detectors, which are read out in 10.73676 s. Subarray integration times are shorter. Five group sizes are designed for short to long integrations: RAPID, BRIGHT, SHALLOW, MEDIUM, and DEEP. Based on current assumptions, RAPID, BRIGHT2, SHALLOW4, MEDIUM8, and DEEP8 are recommended as yielding higher signal to noise for faint sources  (Robberto 2009, 2010; and more recent tests with the ETC).

Integration times are reported as in the ETC: they account for the integration time of the entire detector/subarray rather than an individual pixel.  For example, 3 groups of SHALLOW4 consist of 14 groups (two groups of 5 reads plus a final group of 4 reads). The first pixel completes its integration in 14 frame times.  The last pixel finishes one frame time later. So the total integration time is given as 15 frame times, or 161.0514 s for the full detector.


Table 1. NIRCam Readout patterns and rounded integration times for full detectors

Readout pattern
Reads per group
Frames co-added in each group
Time of first group (s)Time of each subsequent group (s)
Groups per integration
Integration time (s)
RAPID1110.710.71–2 or 1021.5–118
BRIGHT12110.721.51–1021.5–215
BRIGHT22221.521.51–1032.2–225
SHALLOW25221.553.71–1032.2–515
SHALLOW45442.953.71–1053.7–537
MEDIUM210221.5107.41–1032.2–999
MEDIUM810885.9107.41–1096.6– 1063
DEEP220221.5214.71–2032.2–4112
DEEP820885.9214.71–2096.6–4177

The RAPID pattern is limited to two groups per integration (to limit the data rate) when reading out the full detectors in both modules. This limit increases to 10 groups when using a single module. DEEP2 and DEEP8 allow up to 20 groups. Scroll right to view full table, if needed.

Mosaics

Mosaics are used primarily to cover areas larger than the 5.1' × 2.2' NIRCam Field of View (including the ~45" gap between modules). For NIRCam Mosaics, the spatial extent of each tile is defined as 5.115033' × 2.221150' in APT. Tile overlaps as well as shifts (pattern rotation angle) may be adjusted for each axis.  Primary dithers should be used in conjunction with mosaics to fill the gaps between detectors.

FULLBOX primary dithers are recommended for this purpose. If these patterns would split visits, then INTRAMODULEBOX should be considered as well.

A less efficient option is the original FULL pattern, which was designed specifically to yield roughly even observing depth over large areas with mosaics. Use FULL with mosaic tile spacings of 5.8' × 2.25', input as –13.4% column overlap and—1.3% row overlap in APT. The negative overlap leaves a gap between each tile; these gaps are filled by the FULL dither pattern. This strategy was designed by Anderson (2009) before the observing overheads were known. More efficient mosaics are possible using other dither patterns, though these provide less even depth across the field. Additional background information on principles of dithered observations with JWST are also described in Koekemoer & Lindsay (2005), Anderson (2011)Anderson (2014), and Coe (2017).

Figure 4. NIRCam Mosaic Examples

Examples of 3 × 3 NIRCam mosaics combined with primary dither patterns. Top: FULLBOX 6TIGHT with 0% overlap in rows and columns. Bottom: FULL 3TIGHT with –1.3% and –13.4% overlap in rows and columns, respectively. This yields relatively uniform depth across the field, at the expense of significantly higher overheads.


References

Anderson, J. 2009, JWST-STScI-001738
Dither Patterns for NIRCam Imaging

Anderson, J. 2011, JWST-STScI-002199
NIRCam Dithering Strategies I: A Least Squares Approach to Image Combination

Anderson, J., 2014, JWST-STScI-002473
NIRCam Dithering Strategies II: Primaries, Secondaries, and Sampling

Coe, D. 2017, JWST-STScI-005798
More Efficient NIRCam Dither Patterns

Koekemoer, A. M. & Lindsay, K. 2005, JWST-STScI-000647
An Investigation of Optimal Dither Strategies for JWST

Robberto, M., 2009, JWST-STScI-001721
NIRCAM Optimal Readout Modes

Robberto, M., 2010, JWST-STScI-002100
NIRCAM Optimal Readout II: General Case (Including Photon Noise)




Published

 

Latest updates
  •  
    Added Subarray Primary Dithers

  •  
    Added new dither patterns available in APT 25.4.

  •  
    Clarified that Primary Dithers NONE may be recommended in some cases.