Description

Class:

jwst.emicorr.EmiCorrStep

Alias:

emicorr

Overview

The emicorr step corrects for known noise patterns in raw MIRI ramp data. The majority of the MIRI subarrays have a 390 Hz or other electromagnetic interference (EMI) noise pattern in the raw data. The effect of these EMI frequencies is to imprint periodic noise into each ramp image, depending on the readout timing of each pixel. For short integrations in the LRS slitless mode, the correlated noise is quite apparent in the rate images. For longer integrations the net effect is to increase the read noise by about 20%.

In the long term, it may be possible to eliminate the most significant sources of EMI noise by changing the sizes and locations of the subarrays, to get readout times in phase with the known noise frequencies. For existing and near-term observations, this step is required to correct for the noise. In the future, it is likely this step will still be useful to correct for lower level EMI noise sources.

The known frequencies to correct for are stored in the EMICORR reference file. This reference file is expected to be in ASDF format, containing the exact frequency numbers and the corresponding 1D array for the phase amplitudes in the reference waveform. The reference file also contains a subarray_cases dictionary that maps a specific subarray, read pattern, and detector to the necessary frequencies to correct for.

There are two algorithms available for fitting EMI noise in the input ramps: a “sequential” fit and a “joint” fit.

For the sequential fitting algorithm, the process to correct for EMI noise is the following (repeated iteratively for each discrete EMI frequency desired):

1. Read the image data.

2. Make a very crude slope image and fixed pattern bias image for each integration, ignoring the effects of nonlinearity, saturation, bad pixels, etc.

3. Subtract the scaled slope image and bias from each frame of each integration.

4. Calculate the phase of every pixel in the image at the desired EMI frequency (e.g. 390 Hz), relative to the first pixel in the image.

5. Make a binned phased amplitude (PA) waveform from the cleaned data.

6. Measure the phase shift between the binned PA waveform and the input reference waveform. [1]

7. Use a look-up table to get the aligned reference waveform value for each pixel (the noise array corresponding to the input image). [1]

8. Subtract the noise array from the input image and return the cleaned result.

[1] (1,2)

Alternately, if a reference waveform is not available, use the binned PA waveform to “self-correct” the noise.

The joint algorithm proceeds similarly, except that the linear ramps and the EMI noise are fit simultaneously. It works by choosing pixels with modest scatter among the reads, and then finding the amplitude and phase of a supplied reference waveform that, when subtracted, makes these pixels’ ramps as straight as possible. The straightness of the ramps is measured by a chi-squared metric, after fitting lines to each one. As for the sequential algorithm, the EMI signal at each frequency is fit and removed iteratively, for each desired frequency.

The optimal choice for the fitting algorithm depends on the input. The sequential algorithm can be used without a reference waveform, generating a new reference file on-the-fly for user-specified frequencies, but it requires 10 or more groups for a reliable fit. The joint algorithm requires a reference waveform but can successfully fit EMI in ramps with 3 or more groups.

Input

The input file is the _uncal file after the dq_init_step step has been applied, in the calwebb_detector1 pipeline.

Output

The output will be a partially-processed RampModel with the EMI-corrected data in the SCI extension. All other extensions will remain the same.