This software is tailored for time-series spectroscopy and is hosted on my machine tomo and should be accessible from any HEAD machine. Currently only Linux versions are up and running. MOLLY is to be used to prepare a time series of spectra for Doppler tomography. Most of this software is courtesy of Tom Marsh, who always appreciates an acknowledgement when the software is used for scientific analysis.
Danny Steeghs @ CfA
Pamela is an implementation of the optimal extraction algorithm for long-slit CCD spectroscopy and is well suited for time-series spectroscopy. It properly implements the optimal extraction algorithm for curved spectra, including on the fly cosmic ray rejection as well as proper calculation and propagation of the errors.
The Pamela commands are initialised by typing:
source /data/tomo/Software/TRMSOFT/pamela/Pamela
Pamela makes use of the starlink SDF file format, and a familiarity with starlink and figaro is recommended in order to prepare raw data for extraction with Pamela.
Some envrionment variables need to be defined:
setenv MOLLY_START /data/tomo/Software/TRMSOFT/F77/molly/startup
setenv MOLLY_DIR /data/tomo/Software/TRMSOFT/F77/molly
the molly executable is then: /data/tomo/Software/TRMSOFT/molly/molly
Help is available within molly by typing help, or in HTML format online. Data can be imported in SDF format as produced by Pamela or in multi-column ASCII files if the spectra were extracted with other software.
ADDING MOLLY HEADERS
Headers can best be imported through a multi-column ASCII file listing one header per columns and in sequence with the list of spectra to be imported. The following headers (and their molly names) are recommended to be defined as a mimimum: Object Name (Object), UTC time at mid of exposure in decimal hours (UTC), RA of telescope in decimal hours (RA), Declination in decimal degrees (Dec), Equinox of coordinates (Equinox), Exposure time in seconds (Dwell) and UTC date of observations split into Day, Month and Year. The first line of the file contains the name of the header items, the second line whether they are strings (C), integers (I), reals (R) or double precision entries (D). The actual values then follow in the remaining columns, one line per slot.
Example:
Object UTC RA Dec Equinox Dwell Day Month Year
C D D D D R I I I
GX339-4 1.1775 17.047 -48.78969 2000 1300.0 11 08 2002
...
Some recommended papers and reviews discussing Doppler tomography as a technique include Marsh (2001), Steeghs (2003), Steeghs (2004) and Marsh (2005).
Doppler software needs to be initialised by running:
source /data/tomo/Software/TRMSOFT/doppler/Doppler
The doppler commands are now available from the shell. Help can be found here.
In order to display the maps and data files, these are in the Starlink SDF format and thus software that recognises this format should load Doppler files OK. To setup the supporting starlink software, type:
source /star/etc/cshrc
source /star/etc/login
Gaia is a nice image display program that recognizes Doppler maps, but other starlink packages such as figaro can be used to manipulate the SDF files. Alternatively, the starlink convert package can convert SDF files into FITS files.
A program called trailer is available to help with making some plots of .sdf format Doppler maps. It allows one to load in the data and map files from Doppler and arrange them on a page. One can also overplot the gas stream etc. Trailer can be launced as:
/data/tomo/Software/TRMSOFT/trailer
This is an extension to Doppler tomography that incorporates flux variations across the binary orbit. It decomposes the emission into a constant and variable part which are mapped independently. See Steeghs (2003) for details. The linux executable is at:
/data/tomo/Software/ModMap/bin/modmap
Modmap is menu-driven and does not use the Doppler datafile format, but can read in molly files directly for convenience. The spectra in the molly files do need to be binned to either a constant wavelength scale or constant velocity scale and the continuum level should be zero. ModMap automatically only loads up the relevant data pixels, and there is no need to chop the spectra to only contain a particular line. It also does not require regular phase binning, and will map to the exact phase arrays as provided by the data.