Last modified: December 2023

AHELP for CIAO 4.16 Sherpa


Context: models


A CCD pileup model for the ACIS detectors on Chandra.




This model is based on the work by John Davis ( [1] ). It is intended only for modeling the pileup in one-dimensional spectra obtained in imaging mode (i.e. no grating data), but can be used with the zeroth-order spectrum of a grating data set.


>>> create_model_component("jdpileup", "mdl")
>>> print(mdl)

Create a component of the jdpileup model and display its default parameters. The output is:

   Param        Type          Value          Min          Max      Units
   -----        ----          -----          ---          ---      -----
   mdl.alpha    thawed          0.5            0            1           
   mdl.g0       frozen            1  1.17549e-38            1           
   mdl.f        thawed         0.95          0.9            1           
   mdl.n        frozen            1  1.17549e-38          100           
   mdl.ftime    frozen        3.241  1.17549e-38            5        sec
   mdl.fracexp  frozen        0.987            0            1           
   mdl.nterms   frozen           30            1          100           


The attributes for this object are:

Attribute Definition
alpha The alpha parameter parameterizes "grade migration" in the detector and represents the fraction of piled-up events that result in a good grade.
g0 The probabilty of assigning a grade of zero. This should remain frozen.
f The fraction of flux falling into the pileup region. This should remain frozen.
n The number of detection cells. This parameter can not be fit.
ftime The frame time in seconds (as given by the exptime keyword of the event file). This parameter can not be fit.
fracexp The fractional exposure that the source experienced while dithering on the chip (as given by the fracexpo keyword in the ARF file). This parameter can not be fit.
nterms The maximum number of photons considered for pileup in a single readout frame. This should not be changed from its default value of 30. This parameter can not be fit.


As the pileup model is inherently non-linear, it is strongly advised that multiple optimization methods are used to thoroughly investigate the search space for the model.

The alpha parameter should vary with photon energy and detector position, but for simplicity it is treated as independent of energy and location.

An example of using this model to fit a Chandra spectrum is provided in [2] .



See the bugs pages on the Sherpa website for an up-to-date listing of known bugs.