MIRI-Specific Time-Series Observations

MIRI-specific recommended strategies for preparing time-series observations (TSOs)

Main article: MIRI ImagingMIRI Low-Resolution Spectroscopy
See also: MIRI Imaging TSOsMIRI LRS TSOs

With the low-resolution slitless spectroscopy mode ("slitless LRS"), MIRI provides a dedicated observing mode for mid-IR (5–12 µm) high precision spectro-photometry. In the sections below we provide some strategy tips for TSOs with this mode. Limited support is also available for imaging TSOs (~5–25 µm) ; we describe here the current capabilities and limitations.



MIRI TSO imaging

Main article: MIRI Imaging TSOsMIRI Imaging APT Template
See also: MIRI Detector SubarraysMIRI Imaging Target Acquisition

Time-series observations can be performed in all imaging subarrays and with all available imaging filters, which cover approximately 5 to 25 µm. Such observations are defined in the MIRI Imaging APT template, in which the user can indicate that an observation is a TSO by selecting the time-series observation special requirement. 

There are 3 important items to note for such observations:

  • The user should select the Time Series Observation special requirement in the MIRI imaging APT template to indicate a TSO. Note that the No Parallel special requirement should always be selected alongside Time-Series Observation. Selecting the Time-Series Observation special requirement waives the 10,000 s exposure time limit, allowing single exposures to cover multi-hour observations.

  • Target acquisition is not available for any type of MIRI imaging observations. The placement of the target in the center of the field is limited by the blind pointing accuracy of the telescope. This may hamper calibration and analysis of observations which aim to combine exposures over multiple periods or transits. We expect this capability to be available in the future.

  • The option to disable dithering is available for all subarrays on selecting the Time Series Observation special requirement. Note that APT may report an error if the Dither table contains no entries; to avoid this error create a dither pattern in this table, but keep the selection in the Filters table set to "None". 

We expect that full implementation of all TSO-specific settings for MIRI Imaging, including target acquisition, will be available for future cycles.



Exposure setup

Main articles: MIRI Detector Readout OverviewMIRI Detector Subarrays

To define the NGROUPS, NINTS, NEXP setting, the following rules of thumb should be considered for both TSO imaging and spectroscopy:

  • For well-calibrated data, we recommend NGROUPS ≥ 5. For very bright targets, NGROUPS of 2–4 are permitted, but we expect the absolute flux calibration, linearity correction, and cosmic ray detection to be non-optimal.
  • Where possible, the observations should be carried out in a single exposure (NEXP = 1). This provides more accurate time recording for the observation, and having many integrations in an exposure is better for detector stability
  • Users should plan observations of very bright targets with care and consider the effects of saturation. Whilst saturation should be generally avoided, it can be tolerated in some circumstances. For very small numbers of groups (e.g. NGROUPS=2), the benefit of adding an additional group to the integration may outweigh the effects of mild saturation in a small number of pixels. Users should also consider avoiding saturation by using a smaller subarray with a shorter read time (see below).

  • Users should also keep in mind there is a potential for additional systematics in data from bright targets. Detector artifacts, such as pull-up/pull-down effects along the rows and columns of the pixels receiving the target flux, and latent images can occur with all bright targets, regardless of whether a pixel saturates.

Read mode

For FULL array, both FAST and SLOW mode are available for time-series observations; the 3rd MIRI read mode, FASTGRPAVG, is available as a limited access option (enabled by the "Allows restricted" check box on the Proposal Information page in APT). Use of any limited access options requires explicit justification in the proposal text. We do not expect use of FASTGRPAVG to be warranted for TSOs. 

In general, FAST mode is recommended for Imaging TSOs. For faint targets, SLOW mode can be used, however having more up-the-ramp samples (i.e. groups) for a given integration length may improve calibration. As per the general comments above, the user should never select SLOW mode if the resulting NGROUPS would be < 5. 

For subarrays only the FAST read mode is available. Selecting SLOW mode alongside a subarray will results in an error in APT.  

Subarray

A range of subarray sizes is available alongside the default FULL array configuration, providing a large dynamic range. Each subarray has its own characteristic group read time; the SUB64 subarray has shortest read time, and is therefore most suitable for the very brightest targets. The choice of subarray should mostly be driven by the brightness of target, to allow for NGROUPS > 5 where possible. For example, if a target saturates in 0.5 s, the user in principle has the choice of choosing NGROUPS = 4 with the SUB128 subarray (frame time = 0.119 s) or NGROUPS = 5 with SUB64 (frame time = 0.085 s), the SUB64 subarray would be the more appropriate choice (all else being equal).

A smaller subarray also prevents saturation from other targets in the field. 

However some observations may benefit from the additional astrometric information provided by a crowded field— this may make a larger subarray more attractive.



MIRI TSO spectroscopy with LRS slitless mode

Main articles: MIRI Low-Resolution SpectroscopyMIRI Low Resolution Spectroscopy Template APT Guide
See also: MIRI LRS Slitless Target Acquisition

LRS slitless basics

The slitless LRS mode is a variant of the low-resolution spectroscopy mode in MIRI, particularly suited to time series observations. Like LRS performed with slit, the mode shares a focal plane array with MIRI imaging and coronagraphic imaging. The 2 variants of LRS share the same disperser element: a double prism mounted in the imager filter wheel. The spectral resolving power is ~100, varying from approximately 40 to 160 over the 5–12 µm range.

For slitless LRS, the target is placed in a dedicated 72 × 416 pixel-sized subarray that has a frame time (in FAST mode) of 0.159 s, approximately 17.4 times faster than the FULL array read time in FAST mode (2.77 s). The saturation limit of slitless LRS is therefore a factor of ~17.4 brighter than LRS with slit. Conversely, the sensitivity of the slitless LRS is an order of magnitude worse than LRS with slit. The absence of a slit mask allows the background radiation irradiating the entire subarray to be dispersed over the science spectrum, reducing the signal to noise ratio (SNR). 

A noted difference between slitless and slit LRS is the absence of the slit mask filter for slitless LRS. The dispersion profile of the LRS is characterized by a sharp turnover shortward of 4.5 µm; wavelengths below this are folded back onto the spectrum, effectively "contaminating" the longer wavelengths. A filter is mounted on the slit mask blocking wavelengths below 4.5 µm; however the slitless LRS does not mitigate for this effect. Whilst the throughput of the double prism is low at these short wavelengths and the effect is thought to be minor, this could cause some calibration uncertainties in the 5-6 µm region. 

The main advantages of the slitless LRS mode for TSOs compared with traditional slit spectroscopy are:

  • the absence of the slit avoids pointing-related flux variations
  • the subarray is read out faster, providing a larger dynamic range

LRS slitless in APT

Slitless LRS observations use the low-resolution spectroscopy template in APT. The choice between slit and slitless is made in the selection of the subarray. If the user chooses the SLITLESSPRISM subarray, the observation will be carried out in slitless mode; choosing FULL selects the LRS slit mode. The SLITLESSPRISM subarray option is also linked to the Time Series Observation special requirement.

Selecting the slitless subarray automatically disables the along-slit dither pattern, and waives the 10,000 s exposure time limit. 

LRS slitless operations

Target acquisition

Given the importance of target placement for TSOs, target acquisition (TA) is mandatory for slitless LRS observations. For this mode, TA is performed in the same subarray as the science observations (SLITLESSPRISM), i.e. with the same detector readout cadence of 0.159 s. In common with all MIRI modes, 4 filters are available for TA: 

  • F560W
  • F1000W
  • F1550W
  • FND (neutral density filter)

The first 3 of these filters are available for imaging science; the FND filter is a dedicated filter for TA on very bright sources. See the MIRI Filters and Dispersers article for further details of these filters.

As the TA region is shared with the science spectrum, it's important to avoid saturation during TA. Users are therefore strongly recommended to perform TA calculations in the JWST ETC to ensure successful and non-saturated TA exposures. The minimum number of groups required for the TA centroiding algorithm is 3, and the recommended minimum SNR for TA is 20. Whilst the FND is provided for TA on very bright targets, given the saturation limits for slitless LRS we expect the F1550W filter to be more practical for this mode. Stellar-type targets fainter than the saturation limits at the shortest wavelengths can typically be observed with the F1550W filter without saturating. Furthermore, this filter is adjacent to the double prism in the filter wheel, thus minimizing the risk of causing persistent images as the filter wheel moves from the TA filter to the double prism position.

Following TA and placement of the target to the nominal pointing position in the subarray, the procedure is configured such that an exposure will be taken through the TA filter, prior to the filter wheel move, as verification of the target placement before beginning the science exposure. This image can be used to fine-tune the calibration during the data analysis.

TA is carried out in FAST readout mode, or a dedicated FASTGRPAVG mode that averages 4 groups before the downlink. FASTGRPAVG is generally required for long TA exposures on faint targets, where data volume may be an issue; we do not anticipate the mode to be required for typical slitless LRS observations.



Exposure setup

Main articles: MIRI Detector Readout OverviewMIRI Detector Subarrays

To define the NGROUPS, NINTS, NEXP setting, the following rules of thumb should be considered for both TSO imaging and spectroscopy:

  • For well-calibrated data, we recommend NGROUPS ≥ 5. For very bright targets, NGROUPS of 2–4 are permitted, but we expect the absolute flux calibration, linearity correction, and cosmic ray detection to be non-optimal.
  • Where possible, the observations should be carried out in a single exposure (NEXP = 1). This provides more accurate time recording for the observation, and having many integrations in an exposure is better for detector stability
  • Users should plan observations of very bright targets with care and consider the effects of saturation.  Whilst saturation should be generally avoided, it can be tolerated in some circumstances. For the case of MIRI TSOs with slitless LRS specifically, saturation at the shortest wavelengths in the spectrum may still provide scientifically useful data at longer wavelengths. 

  • Users should also keep in mind there is a potential for additional systematics in data from bright targets. Detector artifacts, such as pull-up/pull-down effects along the rows and columns of the pixels receiving the target flux, and latent images can occur with all bright targets, regardless of whether a pixel saturates.

For slitless LRS, further choices are fairly constrained:

  • detector read mode is always FAST
  • subarray is always SLITLESSPRISM



Published

 

Latest updates

  • Updated advice regarding saturation in "Exposure setup" sections for both imaging and LRS TSOs