Create false color temperature images

[mgalloy]

Create two false color images per day (R-G-B):

• 706-1074-789
• 1074-789-637

Requirements:

• The component images should take center wavelength from the average file for the day. For 1074, use the waves average.
• The images should be placed in the fullres directory.

Annotations:

• upper left: "MLSO UCoMP Daily Temperature Image\n[DATE]"

• upper right: for 706-1074-789:

  Corona hottest in regions colored white or red Corona coolest in regions colored blue

  for 1074-789-637:

  Corona coolest in regions colored blue

• lower left: for 706-1074-789:

  RED = 706 mm FeXV log Teff=6.20-6.63 GREEN = 1074 mm FexIll log Teff=6.08-6.41 BLUE = 789 mm FeXl log Teff=5.92-6.30 Where Teff=Effective Temperature

  for 1074-789-637:

  RED 1074 nm FeXIll log Teff=6.08-6.41 GREEN = 789 mm FeXl log Teff=5.92-6.30 BLUE = 637 nm FeX log Teff=5.80-6.24 Where Teff=Effective Temperature

[mgalloy]

From @jburkepile:

Occulter sizes:

I made scans of the UCoMP occulter radius in different wavelengths in IDL and here is what I got:

• 1074 nm occulter radius = 346.5 pixels
• 789 nm occulter radius = 339.5 pixels (harder to measure due to low signal-to-noise)
• 706 nm occulter radius = 340.5 pixels
• 637 nm occulter = 334.5 pixels

From the ballpark measurements I made in IDL for the occulter radii I get:

• 789 radius = 0.9798 * radius of 1074
• 706 radius = 0.9827 * radius of 1074
• 637 radius = 0.9657 * radius of 1074

Unsharp Mask:

img637 = unsharp_mask(img637, radius=5, amount=1)
img1074 = unsharp_mask(img1074, radius=5, amount=4)
img789 = unsharp_mask(img789, radius=5, amount=4)

I do not apply anything to 706. It enhances the noise as much as the corona (in my opinion).

Intensity scaling:

Cooler line image:

tv, bytscl(img1074^0.5, 0, 10)
tv, bytscl(img789(12:1291, 10:1033)^0.5, 0, 7)
tv, bytscl(img637(23:1302, 18:1041)^0.5, 0, 5)

Hotter line image:

tv, bytscl(img706(11:1290, 11:1035)^0.5, 0.2, 3.0)  

FOR 706: need to write mask occulter in 706 to block out unwanted scattered light:

tvcircle, 340, 639, 510, 0, /fill

tv, bytscl(img1074^0.5, 0, 10)
tv, bytscl(img789(12:1291, 10:1033)^0.5, 0, 7)

[mgalloy]

After fixing the indexing, I believe the temperature maps created from the mean/median files are correct now.

Using Joans method, I have also recently started performing the enhanced intensity (UNSHARP_MASK) on most of the channels (637 706 nm does not have enough signal to do it).

[mgalloy]

We also need a way to create temperature maps over the day, and hence movies, but that is probably another issue. See #107.

[jburkepile]

The 637 does have enough signal to perform an unsharp mask. ... you just want to set the amount = 1. The line that doesn't have enough signal-to-noise is the 706. Don't use unsharp mask on 706. For the 1074 and 789 lines you can set the amount = 4

I use radius = 5 for 637, 789 and 1074.

[mgalloy]

I wrote the wrong thing in the comment above, but the code is correct. It enhances the other lines, but not 706 nm.