9. Using RHESSI Data of a Different Orbit for Background Subtraction

In the following, we will describe how you can customize RHESSI data of a different orbit for background subtraction for analysis within SPEX. One possibility is to use data of one orbit before/after the flare (the duration of a RHESSI orbit is about 96 min 40 sec). If you are using the data of one day before/after flare, you have to adjust your time interval to ensure that you duplicate the same range of geomagnetic parameters as for the flare observation itself (check the observing summary). Be sure to select an orbit that does not contain a flare or an electron precipitation event.

In the GUI, save the spectrum and response matrix data of the interval covering the flare as well as the time interval used as background estimate. Be sure that both time intervals are exactly of the same length and that you use identical settings (detectors, energy bins, etc.)

The best thing is you save the observed count rates (+ the uncertainties) of the interval used as background to a file. One way to accomplish this is to call SPEX, read the spectrum and srm data of the background interval (lets assume we use one orbit before the flare) and then exit SPEX:

SPEX> data, hessi, front
SPEX> _1file, hsi_spectrum_1orbbef.fits
SPEX> dfile, hsi_srm_1orbbef.fits
SPEX> graph
SPEX> exit
 

Now, within IDL load the variables "flux" (observed count rate) and "eflux" (Gaussian uncertainties on "flux") from SPEX into your IDL session, rename the variables, and save them to an idl save-file:

IDL> flux_1orbbef = spex_current('flux')
IDL> eflux_1orbbef = spex_current('eflux')
IDL> save, flux_1orbbef, eflux_1orbbef, file='flux_1orbbef.sav'
 

With this procedure you have saved your data which you intend to apply as background count rates for later use.

In the following will be describe how to force SPEX to use these user-selected background data instead of one of the polynomial models which are inherent to SPEX. Call SPEX and read the spectrum and srm data of the flare interval. Do any background subtraction (SPEX needs that step, but later we will overwrite this background with the data from one orbit before), then exit SPEX:

SPEX> data, hessi, front
SPEX> _1file, hsi_spectrum_flare.fits
SPEX> dfile, hsi_srm_flare.fits
SPEX> graph
SPEX> background, null ;skip background subtraction
SPEX> exit
 

Within IDL, we will now read the background data from the IDL save-file (which yields us the variables "flux_1orbbef" and "eflux_1orbbef ") and load the flare count rate data together with the uncertainties (variables "flux" and "eflux") from SPEX.

The new data for the background subtracted count rate (variable "rate") has to be calculated and together with the new background count rates (variable "back") to be loaded into SPEX. Note that we have to provide SPEX also with the actual data for the uncertainties ("erate" and "eback").

IDL> restore,'flux_1orbbef.sav'
IDL> flux = spex_current('flux' ) ; flare count rates
IDL> eflux = spex_current('eflux' ) ; uncertainties of flare count rates
IDL> back_new = flux_1orbbef ; new background count rates (=count rates of one orbit before)
IDL> eback_new = eflux_1orbbef ; uncertainties of background
IDL> rate_new = flux - back_new ; new background subtracted count rates
IDL> erate_new = sqrt( eflux^2. + eback_new^2. ) ; uncertainties of background subtracted rates
IDL> test1 = spex_current('rate', input=rate_new )
IDL> test2 = spex_current('back', input=back_new )
IDL> test3 = spex_current('eback', input=eback_new)
IDL> test4 = spex_current('erate', input=erate_new)
 

Here is an IDL procedure that does all of the above steps in IDL session. You can use it directly by cutting and pasting it into an IDL/SSW session.

If the loading of the new background and rate variables into SPEX was successful, then the variables "test1" to "test4" are set to the value 1. Now you can call SPEX again, and all the variables should have the intended values (i.e. the user defined background data are applied).

Note: Saving the actual SPEX session, e.g. with the command "save_event", does not save the new background and background subtracted count rates! However, if you save your results obtained within SPEX to an IDL save file, then you have all the data as you defined them.

This page last updated: June 27, 2011