12.3 Time-Resolved Observations

There are two ways to obtain time-resolved spectroscopic and imaging observations with STIS:

  • Use the MAMA TIME-TAG operating mode (described in Section 11.1.3) in the UV.
  • Take a series of multiple, short, identical repeated observations (described in Section 11.2.4) of your target in ACCUM operating mode with either the CCD or the MAMAs.

Both the data products received (an event stream in the case of TIME-TAG and a series of individual images from each ACCUM mode exposure in a repeated sequence) and the basic parameters of the time-resolved observations (e.g., sample time, interval between samples, total number of samples or equivalently duration) differ dramatically in these modes. In Table 12.1 we summarize and contrast the ranges of parameter space covered by the different methods of obtaining time-resolved observations.

The information presented in Table 12.1 can be summarized into the following set of guidelines for performing time-resolved observations with STIS:

  • If you wish to observe variability on second-or-less time-scales, observe in the UV using TIME-TAG mode. Figure 12.3 shows an example of a time profile of the Crab Pulsar obtained with STIS G140L in TIME-TAG mode overlaid on prior HSP observations.
  • In the optical, variability can be observed on tens-of-seconds time-scales using subarrays and multiple exposures with the CCD.

  • In the UV, variability on the several-minute or more time-scale can be observed by multiple ACCUM mode exposures with the MAMAs or using TIME-TAG (subject to count rate limitations).
  • A series of identical exposures that are each less than three minutes in length incurs additional data management overheads when compared to longer exposures (see Table 9.2).
  • Single MAMA visits are limited to five orbits (see Chapter 2.7) so continuous variability (though interrupted by occultation for non-CVZ visits) can be tracked for only that duration with the MAMAs.
  • Spatial scanning with the STIS CCD is another option in the optical/NUV – though apparent slight variations in the scan rate may limit both the useful time scales and the accuracy of the timing within individual scanned spectra (see results from test observations in the March 2017 STAN, spatial scanning performance in the September 2020 STAN, and detailed information on this mode in Section 12.12).

Table 12.1: Summary of Time-Resolved Imaging and Spectroscopy.



Range (Å)

Time (τ)

Samples (Δτ)

Total Duration of
Uninterrupted Time Series1




0.1 second

full frame: 45 second
1060 × 32 subarray: 20 second

For τ > 3 minutes, no limit
For τ < 3 minutes:
dur = (τ+Δτ) × 7, full frame (1062 × 1044)
dur = (τ+Δτ) × 256, 1060 × 32 subarray




0.1 second

30 second for τ > 3 minutes
2 minutes for τ < 3 minutes

No limit2
No limit




125 μsecond


6.0 ×107/R seconds if R < 21,000 counts/s
4.0 ×106/R sec if R > 21,000 counts/s,
where R is rate in counts/s

1 τ = duration of an individual exposure.
2 Note, with BINAXIS1=BINAXIS2=2, 7 images separated by 1 minute and each of duration τ < 3 minutes can be obtained.
3 Above 30,000 counts/s timing accuracy suffers.

Figure 12.3: Crab Pulsar Observed Using the STIS FUV-MAMA with G140L in TIME-TAG Mode.

The resulting integrated time profile is shown superimposed on the early results from the HSP. The full period of the Crab pulsar is 33 milliseconds. (Figure courtesy of Don Lindler and Ted Gull, see also Gull et al. 1998, ApJ, 495, L51).