Create quick-look synoptic maps

[detoma]

The goal is to provide quick-look synoptic maps in v1.1 to meet Joan's PSP grant requirement.

We will create maps at 2 heights (1.1 and 1.25 solar radii) for multiple observables for 1074 and at one height (1.1 solar radii) for intensity only for the other lines.

The changes from the current code are the following:

• the heights selected are 1.1 and 1.25 solar radii instead of 1.08 and 1.30solar radii (the lower height was raised to 1.1 to avoid issues with saturation/nonlinearity near the occulter, and the higher height was lowered to increase the SNR)
• the range in height is +/- 0.01 instead of +/-0.02 (0.02 was too broad)
• the angle is 1.0deg instead of 0.5deg (0.5deg was too narrow and was leading to poor statistics)
• use peak intensity instead of center wavelength intensity (this is a more meaningful quantity but needs testing to ensure we do not introduce too much noise by doing this)
• use the median instead of the mean in one bin to avoid bad pixels (this also needs testing given the relatively small number of data averaged)
• some smoothing will be performed in time only.

[detoma]

Suggested strategy:

• test the effect of using peak intensity and median on a single image per each wavelength (we may revert to center wavelength intensity for weaker lines)

• once the choice of intensity and averaging scheme is finalized, process 2-3 rotations to be ready in early January 2026

• create synoptic maps for this rotations and finalize synoptic maps choices (including smoothing and min max for display)

• include new choices in pipeline and run entire UCOMP mission in mid-late January 2026

[detoma]

I did a quick test on the use of mean and median in scans for synoptic maps for the 1074 intensity.

We MUST use the median, not the mean. There are some bad values in peak intensity, especially at higher heights, in noisy pixels where the gaussian fit gives bad results. These bad pixels results in a bad average value if the mean is used instead of the median to average point. They appear as obvious spikes.

This is the a scan at 1.3 solar radii using the mean (there are spikes at ~100 and huge spikes at ~260)

This is the same scan using the median

The same test needs to be done on all the other lines. Fainter lines have more noisy pixels and bad gaussian fits can happen more often. I do not know if we have enough good pixels in each bin for the fainter lines without testing it.

A similar test needs to be done on the other quantities we like to use for synoptic maps.

Joan said she will try to look into it.

[detoma]

Minor comments on ucomp_annulus_gridmeans.pro

The code default for bins is 720 in the current version. We want to change it to 360 and this should be reflected in the keyword description.

The default for the width was correctly changed to 0.01 but the keyword description is not updated

The comments:
; radius : in, required, type=float
; number of solar radii to create annulus at
are not accurate.

This should read for clarity as follow:
; radius : in, required, type=float
; height of the annulus in Rsun

Instead of
; sun_pixels : in, required, type=float
; number of pixels corresponding to a solar radius

I would say
; sun_pixels : in, required, type=float
; solar radius in pixels for the image analyzed

[detoma]

I see negative intensity in some level2 images. The code ucomp_annulus_gridmeans.pro MUST require that the intensity used are positive and below a reasonable maximum.
We need to add the condition that the pixels used of for the average are greater than 0. and less than 500. millions of Bsun.

the lines:
imax=500
bad = where(image le 0 or image gt imax)
image[bad]=!VALUES.F_NAN
will do the trick since we neglect NANs in the average

The value imax=500 should work for coronal lines. A larger imax (e.g. imax=3000) would be needed for chromospheric lines, but it does not make sense to make synoptic maps for 10830 and 656.

For version v1.1 and v1.2 we will NOT create scans and synoptic maps for 530, 656, 691, 706, and 10830.

637 images are already problematic for synoptic maps. I grabbed a 637 image and more than half of the pixels in the fov have negative intensity at line center. Even at 1.1 Rsun there are many negative intensity pixels. Things become worse if we use the peak intensity. I do not think we can use peak intensity for weaker lines but this needs a proper in-depth analysis.