MARX - HAK Example (Flight and MARX)

Update, Feb. 15, 2001

For a MARX simulation to produce accurate simulated wavelengths when analyzed by HAK or CIAO software, the MARX RowlandDiameter should be set to 8637.00 mm :

Change the following line in the marx.par file in your distribution
  from:
RowlandDiameter,r,h,8632.48,1000,,"Enter Rowland Torus Diameter (mm)"
  to:
RowlandDiameter,r,a,8637.00,1000,,"Enter Rowland Torus Diameter (mm)"
                  ^ ^^^^^^^

This change is a temporary expiedient while updated ACIS pixel sizes and parameters are prepared and tested.

Summary

A realistic simulation of an HETGS observation can be made with MARX 3.0 and then analyzed with the HAK software to produce a variety of .gif plots and ISIS-compatible spectra. The .gif plots can give insight into the observation in terms of count rate, zero-order and streak FWHM, pileup likelihood (newly added), spectral resolution, chip-gap locations, background effects, etc.

MARX simulation

A MARX simulation of the Capella observation, Obsid 1318, was carried out using MARX version 3.0 . The simulation was carried out by commands in the files marx_commands and marx_background by typing

source marx_commands

at the unix command line. These files largely consist of pset commands to modify the distributed marx.par for the desired simulation. These command files and the spectral model files used in the simulation are given here:

• marx commands
  • capella model spectrum
• marx background commands
  • ACIS background spectrum (approx)

Note 1. The ACIS background spectrum given here is approximate and serves primarilly to demonstrate one way in which a MARX simulation can include background events which have a non-astrophysical origin (i.e. do not come through the HRMA...); see the web page mentioned in the file and/or the CXC Calibration pages for more on the ACIS background.

Note 2. The simulation here has the same RA, Dec, and Roll as Obsid 1318. It is, however, run with the detector shifted by +3.0mm in the Z direction to move the aim point closer to the framestore regions to somewhat reduce CTI effect in the FI CCDs (the current nominal HETGS recommended offset.) The simulation also has the nominal Y offset of -0.33 arc minutes to put the zero-order fully in node 1 of S3.

The result of the simulation commands are the FITS files Capella_nobkg.fits (no background events) and Capella_wbkg.fits (with background).

The MARX parameter DitherBlur has a default of 0.35 rms radius blur which includes an estimate of the blur introduced by the combination of:

• ACIS pixel quantization
• Aspect reconstruction errors
• Ground processing pixel randomization

The result is that the Sky coordinates, X and Y, in these files have a PSF blur that very accurately approximates the blur seen in the Capella Obsid 1318 FITS event file.

HAK Analysis

Given the MARX-created FITS files, HAK routines were run to create the output below. The commands typed at the IDL command line are given for these examples by the "HAK Commands" link in the top right region of the summary tables. These MARX simulations were analyzed essentially identically to the flight data: using the already aspect corrected Sky (X,Y) coordinates flipped and rotated to detector coordinates. As comparison of the various products shows, the MARX simulation is very faithful to reality.

Capella Flight Observation, Obsid 1318

Capella MARX Simulation

This web page is: http://space.mit.edu/HETG/HAK/hak_marx.html.

Comments may be sent to Dan Dewey at dd@space.mit.edu.