Brian Dennis
Last modified 18 May 2002
This section demonstrates how to determine the location of the flare on the disc of the Sun. This information is needed in making high resolution images of the flare where the size of the image may be as small as 2 or 3 arcminutes square. Eventually, the source location will be available in the quicklook data, at least for the larger flares where the automatic data processing determines this information in a similar manner to that described here.
The technique described here assumes that the aspect solution is accurate. In particular, is assumes that the roll angle as determined from the PMTRAS is accurate for the duration of the interval of interest. If this is not the case, then a manual technique can be used to obtain a relative roll angle using the spin period and an assumed uniform rotation rate. This is described in separate instructions.
The technique for finding the flare is to make a coarse map of the whole Sun using only the coarsest collimators, say 6, 7, 8, and 9. This we do from the IMAGING window by clicking on the "Change" button in the box that shows the "Collimators and Detector Front/Rear Segments Selected." You should then get the following window:

Click on the appropriate boxes to make check marks appear for collimators 7, 8, and 9. Be sure that the "Front" box is checked. Note that currently you can only make images using the front detector segments. Leave the "Time Bins (bus)" numbers at their default values. ("bus" stands for "binary microseconds" where 1 bus = 1.024 microseconds.)
Click on the "Accept" button after you have made all of the changes and the collimator numbers to be used should update in the IMAGING window.
Now click on the "Change" button in the box with the Pixel Size indicated.
The "Image Size and Location Options" window will appear allowing you to change
the pixel size, image dimension, offset from Sun center.
Click on one of the buttons with the ^ symbol on it next to the words "Pixel
size (arcsec):" and select "32.0." In the same way, set the Image Dim (pixels)
to 64 x 64. Note that the default is for square pixels and a square map.
Finally, if it's not already set, change the "Offset of map center from Sun
center (arcsec)" to X = 0.00 and Y = 0.00. The window should look as follows
once all of the changes have been made.

Click on the "Accept" button to enter that changes into the IMAGING window.
At this stage the IMAGING window should look as follows:

Note that there should be one Energy Band from 12 - 25 keV and the "Image Algorithm" should be "Back Projection." If it is not, click on the down arrow to the right of the box and select "Back Projection." Check that all of the other options in this box are as indicated above. If not, click on the "Change..." button and select the options indicated in the following window:

Click on the "Accept" button once you have made the correct selections in this window.
Now we are finally ready to make our first coarse image of the whole Sun. Press the "Make and Display Image(s)" button in the IMAGING window and wait for the following image to appear in the GUI Main Window.

Our large field of view shows the outline of the Sun in red, and a strong X-ray source at the upper right, near the solar limb. Note also the location of the spin axis marked with a + and the ghost source an equal distance from the spin axis but in the opposite direction. The concentric circles about the real and ghost sources are the side-lobe artifacts of the back projection image reconstruction process.
(Note that the following lines should appear along with a lot of other "% Compiled module:" statements in the IDL display window:
HSI_RD_RASPMT: 34 packets of PMTRAS data were found.
PMTRAS_ANALYSIS: Compensating for P-angle = -19.163 degrees
PMTRAS_ANALYSIS: 1049 points ==> Period= 4.332801 s.
RMSscatter= 0.664 degrees
This indicates that the correct PMT RAS information was used to determine the aspect solution.)
We are interested in determining the X and Y coordinates of the real source. These can be read off this image accurately enough to allow us to keep the source in a smaller window with finer resolution. Position the curser cross-hairs over the center of the source in the image and press the right mouse button. The X and Y coordinates are printed next to the source together with the intensity at that pixel. (Note that these values are also printed in the IDL display window.) The coordinates in arcseconds from Sun center that we will use for this flare are as follows:
X: 910, Y: 270
Provided good SAS and PMT RAS data are available for the time interval of interest, the absolute aspect solution should be accurate to an arcsecond or better. However, this is not always the case and the source location should be verified against the known position from other observations.