Example of 20 January 2005 Flare

Making the spectrum fits files

• Use binning code 22 for energy binning.  The energy bins are then 0.33 keV wide up to 15 keV for detailed study of the iron-line complex at 6.7 keV. The bins go up to 300 keV.
• Use 4-second time intervals for the full duration of the flare (the whole orbit) and include some nighttime date to be used for background determination (see under OSPEX below).
• Enable pile-up when making spectrum file of the flare in HESSI to enable the default correction for pulse pile-up. This is particularly important just before the thick attenuators are moved above the detectors and at the peak of the flare. Best fits can be obtained between 10 keV and 300 keV.  We are still studying the difference in the model and the data below 10 keV.
• Use detector front segments for energies up to 300 keV. Use rear segments for energies above 100 keV.  DO NOT use both fronts and rears together in making a spectrum FITS file - OSPEX is not set up to handle them both together.

Using OSPEX

• Do not fit below 6 keV because most counts there are from photons with energies above the 11-keV germanium K-edge that suffer K-escape and deposit only a fraction of their energy in the detector.
• Use background at nighttime when possible. This is particularly important when the attenuator state changes during the flare, as in this case, since the count rate during the flare in the A3 attenuator state (called F3 for 'filter' 3 in the figures below) can be below the preflare level in the A1 or A0 state if there was some low level solar counts at that time.
• Intervals should only be between 4 seconds and ~1 minute long.
• DO NOT try to do spectral fits for time intervals that include an attenuator change.
• Example below is for the front segment of detector #4, i.e., 4F.
• Do not use detector 2 for analysis below 20 keV since it has poorer resolution than the other detectors.
• Do not use detector 7 for analysis below 10 keV since it has poorer resolution than the other detectors.
• Use functions vth + 3pow for fitting.
• Keep the third parameter of the vth function (the abundance relative to coronal) fixed at the default value of 1.
• We usually set the second parameter of the 3pow function (the slope of the low energy power law) fixed at 1.5.
• Except for times when significant flare flux is seen above ~50 keV, set the fifth parameter of 3pow to fixed at 400 and fix the sixth parameter. This makes the third slope inoperative.

**************************************************************************

The following is an example of analysis for 20 January 2005 using OSPEX.

Background levels shown below are chosen at night between 07:28:20 to 07:30:40 UT.

This plot shows the 6 one minute intervals chosen for analysis.

Count-rate spectrum showing the best fit vth + 3pow function between 5.67 and 30 keV for Interval 0 in attenuator state A1.

Best fit for Interval 1 between 5.67 and 50 keV in attenuator state A1.

Best fit for Interval 2 from 15 to 300 keV in attenuator state A3. We are not sure yet why the model fits so poorly below ~12 keV.

Best fit for Interval 3 from 15 to 300 keV in attenuator state A3. Again, we are not sure yet why the model fits so poorly below ~12 keV.

Best fit for Interval 4 from 15 to 300 keV in attenuator state A3. Again, we are not sure yet why the model fits so poorly below ~12 keV.

Best fit for Interval 5 from 15 to 300 keV in attenuator state A3. Again, we are not sure yet why the model fits so poorly below ~12 keV.