VII. 3-Band Cryo Data Release

3. 3-Band Cryo Data Processing

f. Artifact Identification (ARTID)

i. Diffraction Spikes
ii. Scattered-Light Halos
iii. Optical Ghosts
iv. Latents (Persistence)
v. Glints
vi. Cautionary Notes

Bright stars produce a number of artifacts in the both the single exposure and atlas images. These artifacts include diffraction spikes, scattered-light halos, optical ghosts, persistence (latents), and glints. The algorithms used to predict and flag artifacts in the processing of the 3-band cryo data is essentially unchanged from the All-Sky Release, thus we refer the user to the All-Sky Release ARTID documentation for a description of the flagging procedures. However, there are important changes to the 3-band version of ARTID. These include:

There is no artifact flagging for Band 4, since there is no data for Band 4 (obviously). The 'cc_flags' field in the catalog retains its four character format, but the fourth character will always be '0'.

For Band 3 only, this version of ARTID uses the instrumental magnitudes (i.e., non-zero-point-corrected magnitudes) for the purpose of predicting and flagging artifacts. The increasing temperature of the Band 3 array leads to an increasing background, which suppresses artifacts (e.g., the faint parts of diffraction spikes, ghosts, and halos are no longer visible). The increased temperature also decreases short-term latent decay times. Thus, the behavior of artifacts more closely follows the instrumental magnitudes of the parent objects, rather than their actual magnitudes.

Some of the parameters used for 3-band artifact flagging do change. Below we include the tables which list the values of the ARTID parameters used in 3-band processing; the numbering of the tables is identical to the All-Sky Release ARTID documentation, for easier comparison.

The cautionary notes from the All-Sky processing are also applicable to 3-band processing. They are listed again here for your convenience. Read them!

i. Diffraction Spikes

log₁₀L_S = a_L * m_p + b_L

W_S = a_W; m_p ≤ m₁
W_S = b_W; m₁ < m_p ≤ m₂
W_S = c_W; m_p > m₂

**Table 1** - Length and Threshold Parameters for Diffraction Spikes, Single-frame
Band	a_L	b_L	m_{thr_d}
1	-0.195	3.38	9.0
2	-0.178	3.14	9.0
3	-0.177	2.71	5.5

**Table 2** - Width Parameters for Diffraction Spikes, Single-frame
Band	a_W	b_W	c_W	m₁	m₂
1	25.0	20.0	10.0	2.0	5.0
2	25.0	20.0	10.0	2.0	5.0
3	30.0	15.0	7.0	1.0	4.0

i.1.b Spurious sources vs. Real/Contaminated sources

Δm_{spur_d} = (a_spur * r_parent^b_spur) + c_spur

**Table 3** - Parameters for Spurious vs. Real Determination in Diffraction Spikes, Single-frame
Band	a_spur	b_spur	c_spur
1	0.4	0.5	3.5
2	0.4	0.5	3.5
3	0.4	0.5	3.5

i.1.c Background Level Dependence

L_{S_bg} = B * L_S

B = c_bg * log(medint) + d_bg

**Table 4** - Parameters for Background Level Scaling of Diffraction Spike Lengths, Single-frame
Band	c_bg	d_bg	B_min	B_max
1	-0.95	1.95	1.1	0.22
2	-1.5	3.08	1.1	0.3
3	0.0	1.0	N/A	N/A

i.1.d Secondary Spikes (Band 3)

L_sec = a_sec * m_p + b_sec W_sec = c_sec * m_p + d_sec

**Table 5** - Length and Width parameters for Single-frame Secondary Spikes, Band 3
Parameter	Value
a	-47.0
b	261
c	-3.3
d	22.0

i.2 Multiframe Flagging

**Table 6** - Length and Threshold Parameters for Diffraction Spikes, Multiframe
Band	a_L	b_L	m_{thr_d}
1	-0.195	3.38	9.0
2	-0.178	3.14	9.0
3	-0.177	2.83	5.5

**Table 7** - Width Parameters for Diffraction Spikes, Multiframe
Band	a_W	b_W	c_W	m₁	m₂
1	45.0	20.0	10.0	2.0	5.0
2	40.0	20.0	10.0	2.0	5.0
3	50.0	15.0	7.0	1.0	4.0

**Table 8** - Parameters for Spurious vs. Real Determination in Diffraction Spikes, Multiframe
Band	a_spur	b_spur	c_spur
1	0.4	0.5	4.6
2	0.4	0.5	4.6
3	0.4	0.5	4.6

i.2.c Background Dependence

**Table 9** - Parameters for Background Level Scaling of Diffraction Spike Lengths, Multiframe
Band	c_bg	d_bg	B_min	B_max
1	-0.95	1.38	0.9	0.19
2	-1.5	2.18	1.1	0.27
3	-0.30	1.44	1.0	0.55

i.2.d Spike "Fanning" and High Ecliptic Latitude Length Scaling

L_{S_elat} = E₈₅ * L_{S_bg}; 85 ≤ elat ≤ 89
L_{S_elat} = E₈₉ * L_{S_bg}; elat > 89

**Table 10** - High Ecliptic Latitude Scaling for Diffraction Spike Lengths
Band	E₈₅	E₈₉
1	0.4	0.1
2	0.4	0.1
3	0.6	0.1

ii. Scattered-Light Halos

ii.1 Single-frame Flagging

ii.1.a Halo Radius

log10(r_h) = a * m_p + b

**Table 11** - Parameters for Single Frame Halo Radii
Band	a	b
1	-0.144	2.76
2	-0.113	2.49
3	-0.157	2.38

ii.1.b Halo Real vs. Spurious Sources

Δm_{spur_h} = a * log₁₀(d) + c,

**Table 13** - Real vs. Spurious Source Parameters for Single-Frame Halos
Band	a	c
1	6.3	-2.6
2	7.3	-4.3
3	5.6	-2.7

ii.1.c Background Level Dependence

r_{h_bg} = B * r_h

B = c_bg * log(medint) + d_bg

**Table 13** - Parameters for Background Level Scaling of Halo Radii, Single-frame
Band	c_bg	d_bg	B_min	B_max
1	-0.35	1.35	0.01	1.0
2	-0.57	1.78	0.01	1.0
3	0.0	1.0	N/A	N/A

ii.2 Multiframe Flagging

**Table 14** - Parameters for Multiframe Halo Radii
Band	a	b
1	-0.144	2.76
2	-0.113	2.55
3	-0.157	2.48

**Table 15** - Real vs. Spurious Source Parameters for Multiframe Halos
Band	a	c
1	7.0	-2.2
2	7.0	-2.8
3	5.6	-2.7

**Table 16** - Parameters for Background Level Scaling of Halo Radii, Atlas Coadds
Band	c_bg	d_bg	B_min	B_max
1	-0.57	1.10	0.3	1.1
2	-1.35	2.30	0.3	1.1
3	-0.22	1.51	0.3	1.1

iii. Optical Ghosts

iii.1 Single-frame Images

**Table 17** - Optical Ghost Parameters, Single-frame Flagging
Band	Δx	Δy	R_ghost	m_{thr_o}	Δm_{spur_o}

1	114.0	0.0	13.0	1.0	12.0
2	114.0	0.0	13.0	4.5	9.0
3 (1st ghost)	0.0	-206.0	29.0	3.9	6.5
3 (2nd ghost)	0.0	-411.0	45.0	-1.6	7.5

iii.2 Multiframe Flagging

**Table 18** - Ghost Parameters, Multiframe Flagging
Band	Δx	Δy	R_ghost	m_{thr_o}	Δm_{spur_o}
1	313.5	0.0	35.8	0.6	12.0
2	313.5	0.0	35.8	3.1	9.0
3 (1st ghost)	0.0	-566.5	75.6	3.2	6.5
3 (2nd ghost)	0.0	-1130.3	123.8	-2.0	7.5

iv. Latents (Persistence)

iv.1 Single-frame Flagging

**Table 19** - Single-Frame Latent Flagging Parameters
Band	Δm_spur	Parameter A	Parameter B	Scale Background	Decay Time (sec)
1	7.0	0	2.0	15.0	1.5
2	7.0	0	2.0	30.0	1.5
3	6.0	-15.0	35.0	2400.0	1.75

iv.2 Multiframe Flagging

**Table 20** - Multiframe Latent Parameters
Band	Δm_spur	Parameter A	Parameter B	Scale Background	Decay Time (sec)
1	7.0	0	2.0	15.0	1.5
2	7.0	0	2.0	30.0	1.5
3	7.5	-9.6	38.4	1500.0	2.5

v. Glints

v.1 Single-frame Flagging

**Table 21** - "Feather" Glint Parameters, Single-frame, Bands 1 and 2
Band	x (above; below)	y (above; below)	FR(above; below)	FR_Δx (above; below)	m_p range (above; below)
1	400-1024; 0-1024	200-412; 801-830	1306x435; 413x110	-69.0; -13.8	<4.75; 3.4-6.4
2	400-1024; N/A	200-412; N/A	1169x358; N/A	-69.0; N/A	<3.65; N/A

**Table 22** - "Fan" Glint Parameters, Single-frame, Band 1 ONLY
Parent Location	x	y	FR	FR_Δx	m_p range
above	0-1024	120-205	1500x420	-99.0	<3.4
below	0-1024	795-890	1500x420	-99.0	<3.4

**Table 23** - "Needle" Glint Parameters, Single-frame, Bands 1 and 2
Band	x (above; below)	y (above; below)	FR(above; below)	FR_Δx (above; below)	m_p range (above; below)
1	0-400; N/A	200-400; N/A	825x80; N/A	0.0; N/A	<4.25; N/A
2	0-400; N/A	200-400; N/A	825x138; N/A	0.0; N/A	<3.0; N/A
3	0-1024; 0-1024	117-130; 885-904	275x83; 275x83	-22.0; 11.0	<3.5; <3.85

v.2 Multiframe Glints

Glints are not flagged in multiframe processing.

vi. CAUTIONARY NOTES

If possible, examine all your sources by eye on the actual images. There are unflagged artifacts, as it is not possible to perfectly flag every single affected source.

For the cleanest possible search results (with respect to artifacts), we recommend requiring cc_flags = '0000' in your catalog searches. However, as noted above, there are unflagged artifacts, so examine your images.

Above we describe a method of accounting for variable depth-of-coverage in artifact flagging; we refer to this as parent magnitude scaling. This method has deficiencies in cases with low (< ∼10 frames) coverage or very high coverage (> ∼100 frames).

high-elat scaling

In the galactic plane, artifacts are often overflagged, due to the high background level in all bands. Our background scaling procedure somewhat mitigates this effect, but does not do a perfect job for the highest background levels. You may wish to allow contaminated (lower-case) flags for all artifacts if your search is within a degree or two of the galactic plane.

Treatment of near-ecliptic-pole diffraction spike lengths is NOT ideal, and does not completely account for shortening of diffraction spikes in this region. Spikes can be significantly overflagged within a few degrees of the ecliptic pole.

Large (> 1 arcmin) fuzzy bright objects surrounded by dark rings in the Si:As image data should always be presumed to be short-term latents from a bright source on a previous image. Although the size of the flagged region varies with parent and background brightness in the all-sky catalog, the algorithm is tuned for point sources only. Resolved, saturated nebulae will certainly cause a larger region to be affected on subsequent images than the automated pipeline will flag for latents. In addition, W1 and W2 latents, while small, are underflagged for very bright sources which saturate regions larger than 6 pixels in diameter. These appear as a clump of equally bright pixels on a frame following one with a saturated source on the same pixels. The best way to confirm that a source is affected by a latent is to look at the image and several preceding images.

Glints appear, but are not flagged in the multiframe pipeline (which produces the atlas images and catalog). Examine your images, especially when your sources are known to be near bright objects!

Glints appearing on the cross-scan sides of single-frames are not flagged.

As noted above, glint flags are always lower-case. If your source has a 'g' flag, inspect the source on the image to make sure it is not a spurious extraction.
In rare cases, a source may be bright enough that it produces a ghost which has a latent. These latents are not flagged. This occurrence is most common in Band 3

In the case of very bright objects, the Band 3 images can sometimes show a vertical "bleed" from the source. Although these artifacts look like diffraction spikes, they are not flagged.

Last update: 2012 July 6