Page History

...

your favorite astronomical target. See example screenshot below:

Done.

Homework 2

Understanding JWST detectors questionnaire

(some questions have JDox links you might want to check).

1. What is the difference between the JWST NIR and MIR detectors?

The near-infrared detectors are made of mercury-cadmium-telluride (HgCdTe), and are used for wavelengths of up to 5 microns, beginning at around 0.70 microns. The mid-infrared detectors are made of arsenic doped-silicon, and are made for the wavelength range 5-28 microns.

1. What is the fundamental difference between a CCD and a JWST detector readout?

From the Jdox: “The infrared-sensitive detectors in JWST science instruments operate very differently from the CCDs that many astronomers are familiar with from ground-based work or HST's ACS and WFC3/UVIS. These IR detectors, similar to those in the Spitzer Space Telescope instruments and HST WFC3/IR, are read out using a non-destructive up-the-ramp readout technique that provides a number of advantages.”

These advantages enable a greater dynamic range as bright stars can be read before saturation, a reduction of read-out noise, and a reduction of sensitivity to cosmic rays.

1. What is a frame?

A single read of all pixels in the detector array or sub-array.

1. What is a group?

On-board average of the multiple frames, which is used to reduce the readout pattern.

1. What’s the meaning of “group gap” or “dropframes” in the JWST NIR detectors?

Drop-frames are between groups, they are sampled, but are not included in the average that gets downloaded to the ground.  

1. What is a reset?

Closing the detector readout, and reduces all charge counts to the bias level.

1. What is an integration?

An integration is a set of groups between a reset.

1. What is an exposure?

Set of identical integrations, separated by a constant set of resets.

1. Suppose you have data from a CCD and from a JWST detector. They both reach the

saturation level in one-half of the total integration time. Can you describe what the

main difference is? Can you recover information in the saturated pixels?

For the CCD, you’ll have to do saturated star photometry.

For the JWST detector, you may be able to get the photometry from the individual groups, if you have more than two groups per exposure.

1. What will be the impact of a cosmic ray in a JWST integration? Can information be

recovered?

A cosmic ray will cause a large jump, like a theta function, between two frames of a group. It can be accounted for as long as there are multiple groups in the exposure.

1. a) What is the practical difference between a MIRI SLOW mode exposure and a NIR

exposure with NFRAMES=8? b) Which has a higher data rate: a single MIRI Si:As

detector running in SLOW mode or a single NIRCAM H2RG detector using the

MEDIUM8 readout pattern?

Nsample=9 in SLOW mode (first frame is lost) and cannot be altered, observers can alter Ngroups and Nintegrations.

 Nframes=8 refers to MEDIUM8 and DEEP8, which actually have more skipped frames, they have 10 and 20 total frames, so they technically go “deeper”.

NIRCam requires ~80 megabytes of download per group. For Medium8 and Deep8, the maximum comes out to 34 and 68 gigabytes per day.

MIRI takes up ~2 megabytes of data per group per detector, presumably you’re using one at a time. I don’t know what the minimum exposure time is …

1. Given a certain readout pattern, why is the group time different for full and subarray

mode?

https://jwst-docs.stsci.edu/near-infrared-camera/nircam-instrumentation/nircamdetectors/

nircam-detector-subarrays

Subarrays are read out more quickly than the full detector, allowing for shorter integration times. Shorter integration times can allow brighter objects to be observed without saturating the detector. Each pixel takes up 10 microseconds of readout time.

1. If a user defines a single NIRCam exposure (i.e. no dithers) with all modules in FULL

array and BRIGHT1 readout pattern, that uses 10 groups and 1 integration, the

exposure time is 203.99 second. 10 groups and 2 integrations result in 418.73 seconds.

Why the total time of 2 integrations is not twice as long as one? Can you guess why

that would not be the case for MIRI?

https://jwst-docs.stsci.edu/jppom/visit-overheads-timing-model/instrument-specificoverheads/

nircam-overheads

Full Array, 2048x2048 array, t_frame = ((N_columns / N_outputs + 12) × (N_rows + 1) + 1) × 10.00 µsec = 10.737 seconds per readout.

Bright1 means 2 samples per group.

I think that there’s one extra readout time …

Answer …  Readout mode: BRIGHT1, 1 integration of 10 groups takes 203.99 seconds, two integrations take 418.73 seconds. The clock time of two integrations is more thantwice one integration, because there is a reset in between. Thiswouldnot be the case for MIRI because aread-reset (rather than a reset) is executed between integrations, adding no extra time.

Done.