How Post-Flashing Can Improve CTE in Some UVIS images Jay Anderson John MacKenty Sylvia Baggett Kai Noeske.

How Post-Flashing Can ImproveCTE in Some UVIS images

Jay AndersonJohn MacKentySylvia Baggett

Kai Noeske

Background on CTE

• Understanding evolving, improving– Photometric corrections– Pixel-based corrections– Corrections never perfect

• Better…– Minimize losses in first place

• Place source close to amplifier• Maximize signal (fewer exposures)• Avoid low backgrounds• UVIS charge-injection• Post-flash options

CTE TRAILS:

ACS-vs-UVIS(May 2012)

WP

0 1 2 3 4 5

WP LOSSESDOWNCHIP:ACSUVIS

The magic ofshort-long

dark studies…

1000s

100s

−5%

−75%

Finer-scale

CTE-losstrends

Finer-scale

CTE-losstrends Flat?

monotonic

Line

ar g

row

th N

trap

s α

Nq

Line

ar g

row

th N

trap

s α

Nq

Way to think about it…

12 in trail

18 total

6 in head

✕ Transfer out of board

Number of traps encountered isproportional to

the cross-sectionarea

Want to spread out

Prefer to squeeze rather than go up

Prefer to go up and squeeze more

Way to think about it…

18 totalAbove bkgd

6 in bkgd

12 in trail

18 total

6 in head

New ObservationsOmega Cen center, from SM4-calibration --- F336W

Two Kinds of Obsns

1) 10s + 700s at same pointing with var bkgd

2) three 9pt dither stacks: (a) 700s deep (b) 10s (no bkgd) (c) 10s (12 e- bkgd)

DeepStack (700s)

A: 22 electrons above skyB: 28 electrons above skyC: 65 electrons above skyD: 13 electrons above sky

Short UnflashedStack (10s)

Short Post-FlashedStack (10s)

DeepStack (700s)

A: 22 electrons above skyB: 28 electrons above skyC: 65 electrons above skyD: 13 electrons above sky

Trends from ~5×5-pixel Aperture

Photometry on ×2-Sampled Stacks

Losses as a Function of

Background:

Survival of a 100 e- source at top of chip

= sweet spot!

Next steps…• UVIS

– Help users re-plan Phase-II’s to hit sweet spot– Use observations to calibrate losses-vs-bkgd– Constrain model better

• Take more long/short darks-vs-background

• ACS– Following in UVIS’s footsteps– No sweet spot: compromise

• Add noise, preserve signal• More complicated analysis for us and users

– Post-flash issues:• 50% variation across field• 25% dependence on shutter (A or B)