RHESSI Spectroscopy - OSPEX User Guide

From RHESSI Wiki

Revision as of 19:27, 9 June 2010 by Agopie (Talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Contents

Introduction

This is a general introduction to X-ray spectroscopy from RHESSI front-segment data using the Object-oriented version of the SPectrum EXecutive (OSPEX). It covers the steps necessary to fit RHESSI count-rate spectra, starting with the creation of the spectrum and the Spectrometer Response Matrix (SRM) fits files and going through the processes of defining the background spectrum, the time intervals, and the fit functions to be used. The OSPEX - OBJECT SPECTRAL EXECUTIVE GUIDE contains additional information on using OSPEX. Note that similar techniques can also be used for rear-segment data but the detailed analysis of gamma-ray spectra is not covered here.

Much work has already been done by fitting spectra to data from multiple detectors simultaneously. Detectors are chosen for inclusion in the analysis based on their energy coverage and the energy range of interest. Detector #2, for example is generally not usable below ~20 keV, detector #7 below ~10 keV, and detector #8 is occasionally contaminated by pickup from the spacecraft transmitter during passes over a ground stations. This type of analysis is generally adequate above ~20 keV and provides for the greatest sensitivity, but it does not allow various differences between detectors to be taken into account. It should not be used where the highest spectral accuracy is desired or where it is important to estimate uncertainties on the fitted spectral parameters that depend on the systematic differences between detectors.

The general spectral analysis philosophy adopted here is to analyze the count-rate spectra for each of the nine RHESSI front segments independently. Only in this way can the different characteristics of the nine detectors be fully accounted for. Currently, the following detector characteristics can only be taken into account by analyzing the data from each detector separately:

  1. native 1/3 keV energy bins that have different energy edges for each detector,
  2. conversion factors from the detector pulse amplitude to energy loss in keV,
  3. FWHM energy resolution,
  4. relative central-to-total attenuation of each attenuator, and
  5. pulse pile-up.

A further advantage of analyzing spectra from each detector separately is that each spectrum so obtained is an independent estimate of the true solar spectrum. Thus, the scatter of the values obtained for any one of the various spectral fit parameters gives perhaps the best indication of the uncertainties in that parameter. This is superior to the uncertainties derived from the least-squares fitting routine used in OSPEX since it includes any systematic uncertainties in the response of each detector in addition to the statistical uncertainties considered in OSPEX.

Links to documentation for fitting RHESSI Data

Creating a Spectrum File Using the HESSI GUI

This document goes through the steps necessary to create a spectrum file using either the HESSI GUI or the IDL Command Line. The documentation is for creating a spectrum file and a spectrometer response matrix (SRM) file for a single detector, but is easily expandable to creating files for multiple detectors. There is also an ssw routine which can be used to create single detector spectrum files. Each file is created with native 1/3 keV energy bins for the detector and 4 second time binning. It covers the whole daytime part of the orbit containing the flare plus the available before and after nighttime data for use in background estimation. Examples of calling the routine from the command line are given in the file header.

User Guide for OSPEX

Using a spectrum file and SRM file this document goes through the steps necessary for using the OSPEX GUI to fit a RHESSI spectrum, including defining background, time intervals, a fit model and doing a fit. This section also includes information on individual Fit function components, Fitting multiple intervals, Viewing Fit Results, and using the final fit results from one detector to fit another.


Contact Andy Gopie with any questions, corrections, or suggestions

Personal tools
Namespaces
Variants
Actions
Navigation
Toolbox