Now let's create the background spectrum by telling SPEX to accumulate it during intervals before and after the flare when there is nothing going on. This won't always be possible, but when such intervals exist within 10 minutes or so of the flare it's usually the best way to get an accurate background measurement. SPEX can fit a polynomial background model of order N for each energy bin given at least N background time intervals.

Initially, we will choose one background interval before the flare and one after it is over. We will then choose to linearly interpolate (order one) between the spectra in the two intervals in order to arrive at a background spectrum to subtract at each time interval during the flare. If you have more background intervals, then you can fit them with a higher order polynomial to try and get a better estimate of the background during the flare. Also, you can use different background intervals for four different energy ranges to better estimate the background, particularly at higher energies where the flare duration is shorter. This technique is described below but for now, to keep it simple, we will just choose one interval before the flare and one afterwards and use them globally for all energies.

First, set the order of the interpolation polynomial to one:

SPEX> back_order, 1
 

Then, to choose background intervals before and after the flare, first type

SPEX> background
 

The following window will appear:

Click on the "Continue by Selecting from Plot Window" button.

Then move the mouse over the window with the active display ("Time History Window"). Push the left mouse button to define each side of the time intervals to be used for background determinations. Choose the first interval inside the first 1-minute interval of the time history and the second interval inside the last 1-minute interval. Note that the left edge will automatically revert to the data interval to the left of where you select; the right edge will revert to the next edge of a data interval to the right of where you select. Also, be careful that you don't click too fast. Wait after each click to see the vertical line appear on the plot and see the message "Select points for discrete intervals using any mouse button." appear in the IDL output log window. After clicking four times, you should see the following figure:

Once you have selected the two intervals for background determination, move the cursor out of the Time History Window and click "DONE" in the same CURSOR CONTROLS window shown above.

The time histories should immediately change as the background is subtracted for each time interval and energy range. The resulting figure is shown below.

Note that the resulting count rates are negative in some time intervals and energy ranges indicating that the background rate has been overestimated. We will not worry about that at this stage since we still have a positive signal at all energies in the time interval of prime interest at the peak of the flare.

This page last updated: June 27, 2011