Montana State University Solar REU Program 2011
~Casey Donoven~
Week 10-
Week 9-
Week 7 and 8-
I noticed that my velocities seemed unreasonably high after
some simple calculations so Angela suggested that I smooth
the data for more accuracy.
This would reduce noise due to resolution problems with
Rhessi and changing footpoint shape.
I wrote a program that allowed me to average over several
positions and times so I could use these averages to do my
calculations.
These are graphs of the average speed vs number of positions/times
averaged in my program.
Clearly, the more I average, the more the speed is reduced,
but there is a point where the speed decreases less.
I am now trying to determine the point at which the speed
is most accurate.
This is a quick plot of the footpoints from the below movie.
The colorbar didn't want to cooperate so I left it off.
January 19, 2005 Rhessi Images
This is a movie of the first truly successful images I've made of
a SHS flare. Typically, SHSs flare are smaller than SHH flares
making it more difficult to get enough counts for images.
Week 6-
November 2, 2003 Rhessi Images
I apologize for the quality of the movie. I'm seeing if I can fix it.
For now, I suggest stepping through it slowly to see the individual frames.
Statistic | Bottom Left Footpoint |
x coord avg | 485.506 |
x coord stddev | 2.10160 |
y coord avg | -111.316 |
y coord stddev | 3.74826 |
x velocity avg | -2.89582 |
x velocity stddev | 51.8108 |
x speed avg | 35.9432 |
x speed stddev | 36.5604 |
y velocity avg | 13.4130 |
y velocity stddev | 74.0920 |
y speed avg | 49.6211 |
y speed stddev | 55.6162 |
speed avg | 70.3597 |
speed stddev | 56.3355 |
This is data from the May 29, 2003 flare at 01:05. The upper
right footpoint was so dim that it was barely, if at all, visible
for most of the flare.
Also, the looptop source was bright, so I had to use a higher
flux percentage, .97, to identify the bottom footpoint.
I managed to download Rhessi data for all of the x-class
flares on my list. All of the x-class flares, save one, show
SHH behavior.
This is the corrected data with the correct magnetogram.
Statistic | Right Footpoint- old | Right Footpoint- new |
x coord avg | 694.499 | 694.697 |
x coord stddev | 5.14565 | 5.13136 |
y coord avg | -159.733 | -159.690 |
y coord stddev | 3.91136 | 3.97782 |
x velocity avg | 16.1563 | -17.5636 |
x velocity stddev | 163.257 | 125.442 |
x speed avg | 115.862 | 93.2275 |
x speed stddev | 115.093 | 84.8038 |
y velocity avg | 2.54145 | 4.62923 |
y velocity stddev | 114.052 | 114.594 |
y speed avg | 82.2195 | 82.4741 |
y speed stddev | 78.2886 | 78.8721 |
speed avg | 154.945 | 134.349 |
speed stddev | 124.422 | 103.957 |
After noting the two times in which the right footpoint jumps
back to the left, I examined the original images to find that
the footpoints were ill defined in those times.
The centroids calculated during those times encompassed both
footpoints and the looptop source giving erroneous data.
I reran the calculations excluding those points.
This is a magnetogram overlayed with the positions the of the
centroids of the footpoint colored according to the time.
Please note, this is the wrong month and year for the
magnetogram. My apologies.
Statistic | Right Footpoint | Left Footpoint |
x coord avg | 694.499 | 670.277 |
x coord stddev | 5.14565 | 1.55149 |
y coord avg | -159.733 | -160.102 |
y coord stddev | 3.91136 | 1.80427 |
x velocity avg | 16.1563 | -3.43896 |
x velocity stddev | 163.257 | 39.0756 |
x speed avg | 115.862 | 31.4448 |
x speed stddev | 115.093 | 22.0187 |
y velocity avg | 2.54145 | 8.68796 |
y velocity stddev | 114.052 | 60.8459 |
y speed avg | 82.2195 | 42.5619 |
y speed stddev | 78.2886 | 43.0169 |
speed avg | 154.945 | 57.5358 |
speed stddev | 124.422 | 42.3225 |
I have revamped my program to calculate speed more accurately,
using x and y speeds rather than x and y velocities.
Here is the new output from my program.
Week 5-
Statistic | Right Footpoint | Left Footpoint |
x coord avg | 694.499 | 670.277 |
x coord stddev | 5.14565 | 1.55149 |
y coord avg | -159.733 | -160.102 |
y coord stddev | 3.91136 | 1.80427 |
x velocity avg | 16.1563 | -3.43896 |
x velocity stddev | 163.257 | 39.0756 |
y velocity avg | 2.54145 | 8.68796 |
y velocity stddev | 114.052 | 60.8459 |
speed avg | 16.3550 | 9.34382 |
speed stddev | 199.150 | 72.3127 |
I wrote a program that extracts data from flux output files,
calculates velocities of the centroids of the footpoints,
and makes a structure containing all of the data including statistics.
This is a table of the data from the March 18, 2002 flare.
The large standard deviations makes me worry about the
accuracy of my method.
Week 4-
 
 
The left footpoint's coordinates
The right footpoint's coordinates.
 
 
The left x coordinate v time
The right x coordinate v time
 
 
The left y coordinate v time
The right y coordinate v time
 
 
A typical image of both footpoints
Using the Rhessi GUI, I calculated the centroids of the footpoints
during the April 18, 2003 flare. Here are plots of the data.
(Note: the time scales are completely off and are in unix time,
while the position is in arcsecs.)
Week 3-
This is a magnetogram overlayed with contours of the footpoints
(blue, green, and then yellow) and the looptop (red).
The flare moves upward during this time frame and the looptop
shows what direction the flare is pointing.
I also looked into the "hump" after the flare but the data
seemed to be just noise and many images couldn't even be made.
I explored different energy ranges to separate the footpoints
and the top of the flare, but to no avail.
The top can easily identified using the 20 to 40 keV range
but using higher energies resulted in fairly messy images
without clear footpoints.
I made a higher resolution image using all nine detector
on Rhessi, which made the image much, much cleaner.
The bottom footpoint is dim though and making contour plots
didn't seem to help identify it very well.
During the flare, what I believe to be the top of the flare
shows up in the data and makes the footpoints harder to see.
Even making contour plots of the image doesn't help as the
top is so bright it hides the footpoints.
The bright spots are the footpoints of the flare. I will use a magnetogram to
determine which footpoint is positive and negative or if one is actually
the top of the flare.
I spent a bit of time examining the August 3, 2002 flare that occured at 19:07.
This flare had some well defined footpoints as well as a slight 'hump' in xray
emissions after the majority of the flare was over.
Week 2-
Angela and I were able to find data on when the
attenuators change during a flare, which I will use
to make sure future plots are less messy.
Here is a lightcurve that shows how many photons are detected
by the colliminators on the rhessi satelite on July 3, 2002
from 02:09 to 02:16
The attenuators switched on and off during the time interval
making my first attempts at making images go haywire.
I was able to make images (to be made into contours) of the April 17, 2002 flare
in 20 second intervals but 4 of the 131 images were blank due to a lack of data.
The July 3, 2002 gave me more trouble as roughly a third of the images were
blank because the attenuators kept changing to eliminate excess soft xray data.
From the contours of the fotpoints, I can calculate the centroid of each consecutive
footpoint.
With this data, coupled with their positions on the sun,
I will eventually be able to calculate velocities.
This is a magnetogrma with the footpoint of a flare overlayed.
You can see the footpoint of the flare move, from blue to purple,
over 6 different 12 second intervals.
The footpoints come from hard xray data collected by the Rhessi satelite.
Week 1-
This is a magnetogram of the sun with an overlayed contour plot of xray data
from the Rhessi satelite.
These xrays are mostly soft xrays, highlighting the top of the flare.
I used my journal entries from IDL to help make a script that will replicate
this image for my project.
I started some of Angela's tutorials this week.