Jonathan Gimeno Boal, UCM

MODEL PSF EFFECTS OF PILE-UP

Jonathan Gimeno Boal, UCM

Tutor: Martin StuhlingerTutor: Martin Stuhlinger

Jonathan Gimeno Boal, UCM

Index:

• Selection of the observations.

• Determination of the counts per second.

• Selection of point sources.

• Determination of pile-up.

• Aim of the project.

• Next steps.

Jonathan Gimeno Boal, UCM

Aim of the project:

- Pile-up: Arrival of two or more photons within the readout time of the detectors.

- For very bright sources there is a considerable amount of pile-up.

- The centre of the PSF (Point Spread Function) gets distorted:

Loss of counts at the centre of the PSF.Clean PSF.

Jonathan Gimeno Boal, UCM

Aim of the project:

Finally a 3D model of the PSF will be created:

Jonathan Gimeno Boal, UCM

What is pile-up?

Pattern pile-up: Two or more photons arriving next to each other in the same readout time.

Photon pile-up: One single photon causes multiple pixel activation

SAS task Epatplot output.

Consequences:

• Less counts at the PSF centre.

• Shifts the spectrum towards higher energies.

Jonathan Gimeno Boal, UCM

Selection of observations:

● Distinguish point sources:

• Observation Modes

• Filters

Jonathan Gimeno Boal, UCM

Selection of point sources:

In order to automatically determine if an observation is of an extended source or a point source the counts per second of two annuli around the source were calculated and then the ratio of these two count-rates.

Point source: ratio =counts outer

counts innerExtended source: >X%=[5% , 15%]

Jonathan Gimeno Boal, UCM

Determination of counts per second:

Once the number of events registered by the detectors and the exposure time is known, the count rate is determined.

Jonathan Gimeno Boal, UCM

Source coordinates determination:

Annuli extraction needs PSF centre, position of the source in the detector.

First Attempt:

Header sky coordinates

Detector coordinates

SAS task esky2det

Jonathan Gimeno Boal, UCM

Chosen Solution:

Find the centre of mass using IDL software

Work on the Generated Image

( , )

( , )ij

ij

I x y x dxdy I ix

II x y dxdy

( , )

( , )ij

ij

I x y y dxdy I jy

II x y dxdy

Jonathan Gimeno Boal, UCM

Similar ratio as in the extended source case:

Avoid the identification of extended sources as strong pile-up observations

Strong pile-up case

Chosen Solution:

Extended source

singles

doublesratio:

inner annulusouter annulus

Jonathan Gimeno Boal, UCM

Ratio between the single events and double events in energy band of 1keV to 1.8keV. (TBD)

singles

doubles~=2

Determination of pile-up:

Pile-up singles

doubles

Clean PSF:

Jonathan Gimeno Boal, UCM

Investigating in using vertical and horizontal bipixels: (S. Molendi & S. Sembay, 2003 “Assessing the EPIC spectral calibration in the hard band with a 3C 273 observation”)

Determination of pile-up:

Clean PSF: Probability true horizontal bipix. < True vertical bipix.

Pile-up: Probability horizontal bipix.from 2 piled up photons = Vertical bipix.

from 2 piled up photons

To assess pile-up calculate:vertical bipix.

pattern 0On the source and inside an annulus, then calculate the ratio of both results.

Repeat with horizontal bipix. and compare with vertical

Equal ratios: pile-up free.

Different ratios: Pile-up case.

Jonathan Gimeno Boal, UCM

Sum up observations with clean PSF:• Superposition.• Normalize.

Next steps:

Summing up observations in different pile-ups and comparison with clean

PSF:

MOS camera examples.

Jonathan Gimeno Boal, UCM

Questions?