Jenna's Daily Research Log

Thursday June 5

Today I learned how to find and transfer XRT (Hard X-ray Telescope) images aboard Yohkoh onto my workspace.  Then I was taught how to view the fits files and remove bad images from the set.  I also wrote an IDL program to find Supra Arcade Downflows (SADs). "FakeData" is a sample array of images that I used, but the program will work with any 3D array of images if you change the value that the for loops run to. FakeData was a 256x256x50 array, for example.  The images in FakeData form a data cube (a 3D array) which is also a movie.The program finds the mean and variation of the pixels in each image and then averages those values for all of the images.    By finding the mean and variation in the pixels, the program is able to find all the pixels below the standard deviation of the mean.  This is how the downflows are detected.  Next the program changes the color of the pixels that make up the downflows.  It changes these pixels to a lighter color and all the background pixels become black.  When the program runs, you see a clear video, without noise, of all the  downflows that make up the data cube.

Friday June 6

I spent the day doing research online.  I was searching for a method of finding and transferring LASCO (Large Angle Spectrometric COronagraph) images.  I also wanted to find out if those images were already prepped or not and what format the images were in. The information I located is on the SOHO/LASCO homepage.  I did find out that I could transfer the images in FITS format, which is what we prefer.

Monday June 9

I started my day by attending the weekly Hinode meeting for the first time.  Every week a group of students/professors meets to discuss the Hinode spacecraft and their progress on certain tasks related to that particular satellite.  After that I spent the rest of my morning at a talk by Dr. Duncan Mackay from St. Andrews University.  I spent the afternoon reading through the Finding Supra-Arcade Downflows Manual that Sabrina Savage wrote.  As I went through the manual, I searched and viewed many of the files in IDL that Sabrina used to teach me last week.  Later in the afternoon I met with my mentor, David McKenzie.  he gave me the task of finding some specific FITS files taken by LASCO.  I found images for the last two dates, but not for the first one:

20-Jan-1999 19:00-23:59 UT
20-Apr-1998 09:00-23:59 UT
04-Nov-2003 19:00-23:59 UT

Tuesday June 10

Today I downloaded the images for April 20 (980420) onto Dr. McKenzie's computer.  The images I downloaded are LASCO images from the C2 coronagraph.  On the LASCO query form I chose the file type as level_1 and the source as NRL LZ.  I then transferred the files to my own workspace.  Next I searched LASCO's data archives for C1 images for the same date, but found nothing.  Then I downloaded the November 4 (031104) data onto my computer for the C2 coronagraph, but I did not find any C1 images for this date either.  After I obtained all these files for both dates, I started running the 980420 images through Sabrina's FindSADs program.  The first step involved viewing the FITS files.  After doing this, I realized there was only one image that was sized as 512x512; therefore, it was the only image in the data cube and could not be played as a movie.  There were 26 images that were 1024x1024, however.  So those images could be viewed as a movie.  Unsure of what else to do, I tried viewing the 031104 images.  There was only one image that was 512x512, but there were 10 images that were 1024x1024.  So I tried playing the movie consisting of the 1024x1024 images, but the changes in the corona as the movie played were very slight.

Wednesday June 11

I spent the morning trying to figure out what I was doing wrong with the LASCO images.  I did some research on LASCO's website because Sabrina's manual does not include how to work with these type of images.  In the afternoon, Dr. McKenzie realized that these LASCO images were already prepped so the next step was creating a map.  Finally, by the end of the day, I created a map for the 980420 data.

Thursday June 12

This morning I sat through a lecture on chapters 1-3 of The Cambridge Encyclopedia of the Sun.  Dr. Longcope gave the lecture this morning.  After that, I went to work on my LASCO data.  I realized that the map I made of 980420 was not saved so I created a new map.  After that I sharpened the data.  The program spits out numerous maps that have several tasks done to them.  I accepted the srmfmap.  This means that the map was flattened, then run-mean-differenced, and then smoothed.  I renamed this map because Sabrina's parallel processing script needs the data to be named sadsdata.sav.  The next step involved running the data through Sabrina's findsads program.  Unfortunately, an error kept coming up saying that a variable (npixmax) was undefined in line 1346 of her code.  That being said, I had to spend the majority of my afternoon searching her code for the problem.  After going to line 1346 and working my way up through the code, I realized that the variable was indeed undefined.  Dr. McKenzie told me to put in the value that npixmax should have been according to the code and continue running findsads.  After doing this, Sabrina's program worked nicely and found several SADs in that particular flare.  After this I decided to create a map for the 031104 flare.  Then I sharpened those images and chose the srmfmap again.   Finally, I renamed the sharpened data to sadsdata.sav.  The map is below.



 

Friday June 13

That's right, Friday the 13th.  So far my day is off to a good start though.  I started by running the 031104 data through the FindSADs program.  Just like yesterday, the program printed out the same error so I typed in npixmax=14 and continued on.  FindSADs detected SADs for this flare also.  After this I went back to the 980420 flare.  I ran this data through the combine_troughs program and then through the traject_search program.  The programs both seemed to work fine, but it printed out that no trajectories were found.  Since I detected several downflows in the first part of the program, I highly doubt there are no paths for these downflows.  This would mean that I am doing something wrong even though the programs both ran fine.  I think I might possibly be entering the wrong dataset when I run these programs.  To test this theory, I ran the 031104 data through both programs to see if any trajectories were found for this flare.  None were found.  Now it's a matter of figuring out the problem.  Several hours later, after getting frustrated with my problem, I decided to download the rest of the LASCO data that David told me to do after this data.  I got images for the following dates from the LASCO archives:

020421 00:00-02:00 UT     (21-April-2002)
020723 00:00-01:00 UT     (23-July-2002)
010402 21:00-23:59 UT     (02-April-2001)

All of these images are in FITS format and are from the C2 coronagraph. 

Monday June 16

I started my day by looking at the LASCO data from Friday.  I'm not sure whether there is actually a problem or there just really isn't any trajectories.  After looking at that data, I started on the other flares because without trajectories I cannot move on with the old data.  First, I made a map for the 020421 data.  Then I sharpened those images and renamed the data sadsdata.sav.  When I sharpened the data, I picked the srmfmap.  Next I ran the 020421 data through the findsads program and the traject_search program.  I am happy to say that the program detected trajectories for this flare; however, the program spit out an error saying that the problem was a floating divide by zero. Because the program did find flares, however, this means that I did not do anything wrong with the flare data from last week they just did not have any trajectories.  Dr. McKenzie showed up shortly after this and we fixed Sabrina's program by adding the line npixmax=fix(maxpix).  We saved the new program in my own directory.  After this I moved on to the 020723 data.  As usual, I made a map for this data.  Then I sharpened the data, choosing the srmfmap, and renamed it sadsdata.sav.  When I tried running this data through findsads it printed back some errors I have never seen.  This data set only contained 3 images though so I went back to the LASCO website and downloaded more data for this flare to see if this would help.  I expanded the time frame to 00:00-05:00 UT.  I made a new map of this data and then sharpened the data choosing the srmfmap.  I also renamed the data as usual.  I then went ahead and ran the findsads program.  After that I ran the data through combine_troughs and traject_search.  Traject_search spit out the same error for this data as it did for the 020421 data, but it did detect trajectories.  There is an error somewhere on line 1996 of the code so I checked the code, but according to what is written, the variable should not be divided by zero.  This has me stumped.  After this I tried running the last flare I have data for at this point in time, 010402.  I created a map, sharpened the data (choosing the srmfmap), renamed the data, and combined troughs.  After getting a bunch of error messages on traject_search I decided to spend the rest of the day reading chapter 4 of The Cambridge Encyclopedia of the Sun to get ready for tomorrow's lecture.

Tuesday June 17

I started off my day with one of our weekly lectures on The Cambridge Encyclopedia of the Sun.  The lecture this morning covered chapter four.  Then I headed off to my weekly meeting for the Hinode spacecraft.  During the meeting we always give an update on our work for the past week so I told Dr. McKenzie about the problems I had with the traject_search program the day before.  He said that the program always prints out errors like that (division by zero) because they are unavoidable.  He said that as long as the program completed first and found trajectories then I shouldn't worry about it.  That means the 020421 data and the 020723 data both have trajectories and the error is unimportant.  After getting back from my meetings I ran the 010402 data through traject_search and found no trajectories, but I might try extracting a different section of the image later and trying to find trajectories there.  After that I downloaded data for 12-July-2000 (000712).  There were two different flares on this day so I downloaded two different datasets:

01:50-05:00 UT
18:40-23:59 UT

Both of these datasets are are level_1 and from the C2 coronagraph.  Next I went to work on the first flare from 000712.  I created a map, sharpened the data (srmfmap), renamed the data, ran the data through the findsads, combine_troughs, and traject_search programs.  The last program found trajectories for the first 000712 flare.  After this I moved to the second flare that day. I repeated the exact same steps for this flare starting with creating a map.  After running the data for the second flare on 000712 through all the programs, it returned no trajectories found.  Here is a list of the results I have so far for the LASCO data:

Flare
Trajectories
980420
None
031104
None
020421
Yes
020723
Yes
010402
None
000712 (first flare)
Yes
000712 (second flare)
None

After I finished looking at the second flare on 000712 I came up with a list of things I wanted to do.  The first thing I did was check the 980420 data and the 031104 data for trajectories again.  I noticed the first time that when I renamed the sharpened data for these two flares, they had a different variable being restored than all the other data.  When I went back through and renamed the data this time, however, they both had the same variable.  This did not have any effect on the results though because I still found that both flares have no trajectories.  The next thing I wanted to do was run the 010402 data through the findsads program again so that I could extract a different area.  I don't think the first area I picked was the right section of the sun so I tried a different section.  After doing this and then running the data through traject_search I still found no trajectories in this flare. 

Wednesday June 18

I talked to Dr. McKenzie on my way out yesterday and told him I planned on going back through all of the flares above to make sure that the program picked up real trajectories and didn't miss any that should've been picked up.  So that is how I started out my day.  After going through all of the flares I decided that I don't see any trajectories in any of the flares.  During my meeting with Dr. McKenzie we decided that the SADs in the LASCO images would be very hard to see so we came up with a plan.  First I was going to make a new map for each flare that includes the whole image instead of making smaller images that cut off some of the data.  Then when I sharpened this data I need to go through all the possible choices instead of choosing the srmfmap.  Dr. McKenzie thinks that smoothing the data might make get rid of some of the downflows that we might otherwise see.  So I started with the 980420 flare and worked my way through most of the list.  Late in the afternoon I met with Dr. McKenzie again and showed him the different sharpened datasets that the sharpen_flows program puts out because I told him there were all horrible.  Since he agreed, we came up with a different plan.  I need to search through Sabrina's code for the sharpen_flows program and see if there is an easy way to output a map that has just been run-differenced and not run-mean-differenced.  Dr. McKenzie says that map should look the best.  If that is not already in Sabrina's code, we're going to add it in and then put all of the data back into findsads.

Thursday June 19

This morning I sat through a lecture on chapter five of The Cambridge Encyclopedia of the Sun.  After that I went back to work on figuring out Sabrina's program.  There was an easy fix to our problem.  When I started the program, I just had to add /rdiff and the program would find the run-difference instead of the run-mean-difference.  When I viewed the new map of the 980420 data, however, Dr. McKenzie and I decided that the image needed to be rescaled.  The program Dr. McKenzie found to do this kept printing out errors so now I am trying to write a program of my own.

Friday June 20

At the end of the day yesterday Dr. McKenzie looked at my program.  It made the images better but still not good enough.  He could not make them any better so he said he'd work on it and let me know today.  While I was waiting to see what he came up with this morning I went ahead and made run-difference maps for all of the data in the table above.  Shortly after this was finished, David came in with a way of fixing our images.  Our new method involved taking the run-differenced images and then removing the bad images from the set.  After that we used the bytscl command and resaved our data.  I did this for 980420, 031104, and 020421.  Then I ran these three flares through all of the programs I have been using, but no trajectories were found.  I was convinced that I was doing something wrong so David told me to download a LASCO flare that we knew downflows were in.  I started running this new flare, 000101, through the programs, but I still found no trajectories.

Monday June 23

This morning I started my day off by working with the 000101 data again.  This is data that we know has downflows; therefore, traject_search should be detecting something, but its not.  I tried messing around with a bunch of the parameters in the GUI window including the pixel resolution.  None of these helped.  Later when Dr. McKenzie came we discovered that the default cadence time was smaller than the time between the LASCO images.  To solve this problem, I changed the default cadence time in Sabrina's code for the traject_search program.   We thought this would fix our problems, but traject_search still refused to detect the trajectories that we can see.  The rest of the day was spent doing detective work on the traject_search program and reading a paper about axions for the Hinode meeting tomorrow.

Tuesday June 24

It has been a crazy two days.  It's now Wednesday but the servers have been down off and on so I haven't been able to edit my page.  Tuesday I went to another lecture on The Cambridge Encyclopedia of the Sun.  After that I went to my weekly Hinode meeting.  David came up with an idea for what might be wrong with Sabrina's program: he thought maybe since the LASCO images have a resolution of 12 arcsec/pixel the trough that FindSADs detected was really large and the traject_search program might be rejecting it since it is so large.  I was going to start looking through Sabrina's code to see if this was our problem but since the servers were being moved I read the paper In/Out Pairs and the Detachment of Coronal Streamers (Wang and Sheeley 2006).  Towards the end of the day I started searching Sabrina's program, but found nothing.

Wednesday June 25

Well the servers were back up this morning but they still had a few issues and then they quit working completely during the afternoon.  That being said, I searched through Sabrina's program for something that might reject large troughs, but I still found nothing.  After talking to Dr. McKenzie, he emailed Sabrina about our problems and how we have not been able to fix anything.  While we wait for a response from her, I am going through the code line by line to see if anything suggests why the traject_search program would not return any trajectories when FindSADs detected a very clear trough for 5 consecutive images. 

Thursday June 26

I started today with another lecture on the sun.  Today's topic covered the solar wind.  Sabrina had still not gotten back to us this morning about the traject_search program so I continued looking for problems in her code.  She finally sent us a response after lunch.   She said she would be looking into the issue, but her first response was to try adjusting the pixel off values.  I spent the majority of the afternoon working on this, but with no luck.  Below is the third image of the 000101 data:



The trough that FindSADs detected is the "blob" above the upper left part of the sun.  This trough is detected in five consecutive frames, but the next program we run the data through says that no trajectories exist.  That is the problem we're trying to solve. 

Friday June 27

I am super excited to report that our problem has been solved!  Sabrina sent an e-mail this morning saying that she has been meaning to add a new keyword to her program that might help us out.  Traject_search bases the pixoff values by the major axes of the flows.  The new keyword allows us to use a bigger percentage of the axis size.  At first I thought this new keyword was still not finding anything but when I tried changing the pixel resolution on top of the the new keyword the program detected a trajectory for our troughs.  What a fantastic way to start the day!  After finding the trajectory for the 000101 flare, which was a small side project to make sure our programs were working correctly, I went back to my real data.  I had already created new maps for the 980420, 031104, and the 020421 flares so I started with those first.  I found trajectories for both the 980420 and the 031104 flares, but not for the 020421 data.  After lunch I started on the 020723 data.  I had to create a new map and resharpen the data following all of my usual steps before I could run the data through FindSADs, combine_troughs, and traject_search.  I had some problems sharpening this data, however, so I moved on to the 010402 data.  I found trajectories for this data but then got a new error message for traject_search.  Since right now we only care about whether or not trajectories are being found, I moved on to the 000712 flares.  The first flare on July 12 had no trajectories and I did not finishing running the second flare through all of the programs. 

Monday June 30

I started my morning out by finishing the second flare from 000712.  I did find trajectories for this particular flare.  After that I went back to the 020723 flare to find out what was wrong with the sharpened data.  After deleting the bad images I had a nice looking movie so I ran the images through Sabrina's programs.  This flare had no trajectories.  I sent David an e-mail earlier this morning telling him I would need new data soon so he gave me a list of flares that I tried to download.  Some of the flares I could not get data on:

980509 (03:00-07:00 UT)
980816 (17:30-21:00 UT) - No data
980818 (22:00-23:59 UT) - No data
980930 (13:00-16:00 UT) - No data
990120 (19:00-23:59 UT) - No data
990508 (10:30-13:00 UT)
000805 (23:00-August 7th 23:59 UT)
010403 (03:00-10:00 UT)
010404 (09:30-12:00 UT)

The flare on August 5th covered three days and I had trouble downloading the data because there is a restriction on how much data I can get at once.  That being said, I decided to download the data later when I finish with the other flares.  I spent the rest of the afternoon running as much of the above data as possible.  I found no flows for 980509 and 010403, but I did find them for 990508.  I started running the 010404 data, but I did not finish it. 

Tuesday July 1

Just like every other Tuesday, I started my day with a lecture on The Cambridge Encyclopedia of the Sun and my Hinode meeting.  The lecture this morning was on observing the sun or chapter nine.  At the Hinode meeting, David said that Sabrina and him were confused as to why changing the pixel resolution fixed the traject_search problem.  He was happy that the program is finding the trajectories but said that Sabrina and I would probably sort out why this fixes our problem when she returns from vacation tomorrow.  After both of these events this morning, I went back to running the rest of my data through Sabrina's programs.  The first flare, 010404, had no trajectories.  Before I could start the 000805 data, I had to download all of the images from LASCO's website.  This took way longer than expected because I can only download so much data at once.  There were a ton of images since this flare is on three different days so I had to download the images in groups.   This took up much of my afternoon.  Once it was all downloaded I started sharpening the data. 

Wednesday July 2

I spent all morning and part of the afternoon at a coalignment workshop.  Professors and grad students gave talks on how they have coaligned images because of differences in resolution, instruments, etc.  I also talked to Sabrina and David this morning.  Sabrina is planning to look at the traject_search program before Monday to see why changing the pixel resolution is fixing our problem.  Until then I am supposed to start analyzing  the  flares I have already worked on, but instead of working with LASCO data, I will be analyzing the flares using SXT (Soft X-ray Telescope) or TRACE (Transition Region and Coronal Explorer) data.   I spent some of the afternoon looking at Letisha McLaughlin's website to see what data she has already analyzed using either SXT or TRACE.  Letisha was the REU student working with David and Sabrina in 2006 and 2007.  Just to show all of my results in one place, an updated table is below:

Flare
Trajectories
980420*
Yes
031104
Yes
020421**
None
020723
None
010402*
Yes
000712 (first flare)
None
000712 (second flare)*
Yes
980509*
None
990508*
Yes
000805

010403
None
010404
None

A * above means Letisha has done some SXT analysis on these flares.  ** means that she analyzed some TRACE data. 

Thursday July 3

A group of REU students took the day off to spend a four day weekend at Glacier National Park.  It was amazing!

Monday July 7

When I came to work this morning I realized that I never finished the 000805 LASCO data because of other tasks that came up.  I spent the morning creating a map.  It took quite awhile because there are a lot of images for that flare.  I was going to run the data through FindSADs, but then I did not know what to do about the pixel resolution problem.  I was not going to run the data through combine_troughs and traject_search until Sabrina and I figure out what is wrong with traject_search but I never got to that point anyway.  We have a meeting this afternoon so hopefully we can finally resolve that problem.  I spent the rest of the morning looking over Letisha's website for information on the flares I am working with.  I also tried to figure out how to get SXT and TRACE data.  Sabrina's manual gives a quick overview, but some of it is unclear.  Finally I retried the 000101 data to see if it would work without changing the pixel resolution, but no trajectory was found for the very clear SAD.  Sabrina and I will have a lot to talk about this afternoon.  During our meeting Sabrina said she wanted me to run all of the above flares through FindSADs again.  I also need to recreate a new map for each one.  Instead of using an entire image, I will be using sub_map to make the images smaller and then I will zoom in on the image.  After this, the data will be sharpened and ran through FindSADs again.  After this is done for all of the flares, we will go from there.  I started this process with the first flare, 980420. 

Tuesday July 8

I didn't have any lectures or meetings this morning which is very unusual for a Tuesday.  I went back to work on the 980420 flare this morning but after creating a map, resizing the data, and sharpening the data, I found that there were still bad images in my set.  I sent Sabrina a quick message because she said this would no longer be a problem because of the method I am now using.  I might have an idea for fixing it it, but I wanted to double check with her.  While I waited for a response, I went through and created a data cube for each flare.  I got a response from Sabrina and she said exactly what I was thinking.  I spent the afternoon removing the bad images from 980420's data but when it came time to run the data through FindSADs I kept getting errors.  IDL kept saying that the variable sind was undefined, but it is defined, it is zero.  I've had this problem before but I've always been able to fix it easily.  I spent all afternoon fighting this battle but no matter what I did, FindSADs would not run.  It has been a very frustrating afternoon.  After I could not take anymore of those stupid errors, I tried moving on to 031104 thinking maybe I kept getting errors because something was wrong with the 980420 data.  Nope.  Still no progress. 

Wednesday July 9

Sabrina sent me an e-mail last night so now I have a few ideas for what to do with the 980420 data.  After extracting a sub region, I need to make the images bigger.  This should give the data a lower pixel resolution.  I finished doing this for the 980420 flare, but then FindSADs kept printing out the same error still.  While I wait to see what is wrong with that, I started creating maps for the rest of the data and sharpening the images.  Sabrina sent me another e-mail after lunch saying that she fixed the FindSADs program but has no idea how it got like that because the program should never have been able to run if that error always existed.  That being said, I finished running all of the flares from the table above through the sharpen_flows program.  Then I ran 980420, 031104, and 020421 through FindSADs.  Sabrina told me earlier that the extractions I have been using for all of the flares may still be too large.  We have a meeting tomorrow afternoon and she said she will show me what she means by this when we meet.  I will probably have more to say about that tomorrow when I figure out how to change them and why they are too big in the first place.

Thursday July 10

I started my morning by attending the last lecture on The Cambridge Encyclopedia of the Sun.  Today's lecture was on the connection between the sun and the earth.  After this, I started running the rest of the flares through FindSADs.  I started with 020723 and worked my way down the list above.  Later this afternoon, when I meet with Sabrina, I will probably resize all of the maps again, resharpen the data, and then run all of the flares through FindSADs yet again.  When I ran all of the flares through FindSADs today, I skipped 010402 because that flare has two bad frames and it only has four frames to begin with.  Also, when I was running all of these flares through FindSADs, I kept getting an error.  FindSADs had already detected the troughs so I ignored the errors, but I think Sabrina is probably right about the extractions being too large.  The error that printed was arithmetic error: floating underflow.  The reason I think this might be more of an issue is because when I try to run the next flare through FindSADs, I have to retall or FindSADs won't run.  This means that FindSADs is still running from the last flare and never finished what it was doing.  That being said, I only finished running the following flares: 020723 and 000712 (first and second flares).  After that, I decided to try extracting smaller regions instead of wasting time running the data through FindSADs when it is going to have to be done again anyway.  I started with the 980420 flare.  I extracted a much smaller section, blew the image up, and then resharpened the data.

This afternoon I had a meeting with Sabrina.  She showed me how big of a region to extract from my data.  Then she gave me a little bit more direction for my project.  Right now I am going to go through each of the flares above and extract a region, blow it up, and then run it through FindSADs.  If I detect any troughs, I am going to work on the same flare but in SXT and TRACE.  That way I can work more towards seeing if any of the downflows match up in the different instruments.  That being said, I am starting back at the beginning with the 980420 LASCO data.  Also, I am now saving the srfmaps like I used to instead of the rmaps.  While I waited for Sabrina to respond to a question I had about getting SXT data, I also ran the LASCO data for 031104 through FindSADs after extracting a smaller region.  After hearing back from Sabrina, I downloaded the SXT data for 980420 and created a data cube. 

Friday July 11

This morning I started with the 980420 SXT data.  When I downloaded the data yesterday, I was only able to get full images as opposed to partial images; therefore, when I started today, I had to extract a smaller region and then blow it up.  After that I ran the data through FindSADs.  The program detected troughs just like it did for the LASCO data, but I don't think any of those are actually SADs that line up with the LASCO data so I moved on to the 031104 SXT data.  Unfortunately, when I searched the SXT archives they only went through 2001.  Since I could not get SXT data for this flare, I moved to the 020421 LASCO data.   There is no SXT data for 020421 since it was in 2002 so I moved on to the 020723 LASCO data.  Since there is also no SXT data for 020723, I moved to the 000712 flares after I finished the 020723 LASCO data.  After I finished with the 000712 LASCO data, I started the SXT data for these flares.  I downloaded all of the fits files, but then I realized I already had info in my directories.  Sabrina had used this flare to train me my first week so all of the SXT data already exists for 000712.  Instead of continuing on with more flares, I decided to to work on the 980420 flare and the 000712 data because I have both LASCO and SXT data for these flares.  There are a couple of different things I want to try with these images so we'll see if this works...No luck.  I tried blowing up the LASCO data even more for the 980420 flare, but that was definitely a bad idea.  Time to move on.  I started working with the LASCO data for the 980509 flare and ended the day by running the data through FindSADs.

Monday July 14

Since I finished the 980509 LASCO data last week, I started my day by downloading the SXT data for this flare.   This took way longer than expected because I finally found a way to download the partial images instead of the full images but there is a ton of data.  I have been wanting to get the partial images because that's originally what Sabrina said to get but I was never able to find a way to obtain the data in that form until today.  Now I am using ftp to download all the data to my directories.  It took me quite awhile to figure this out.  After getting all of the 980509 SXT data, I started creating a map, sharpening the data, etc.  When I tried to compile the FindSADs program, however, it wouldn't compile and a weird error that I had never seen before kept showing up.  It said that FSNONE is already defined with a conflicting definition.  While waiting for Sabrina to sort that out, I went back to the 980420 and 000712 flares to see if I could get some new SXT data in the form of partial images.  Sabrina sent me an e-mail saying that she was working on the program when I tried to run it which is why I got an error.  That being said, I ran the SXT data for 980509 and the troughs that were detected seem like they might match up well with the LASCO data!  After this I went back to downloading the SXT data for the other two flares because I never finished that.  After I downloaded all of that data, I ran the 980420 SXT data through FindSADs.  I also chose to make a map for the 128x128 images.  I've been working with the 64x64 images because there are more of them, but I might try running the bigger images tomorrow. 

Tuesday July 15

I started today by trying to run the 128x128 images for the 980420 flare through the sharpen_flows program.  I noticed that there were tons of bad images and the video was shaking a lot.  I decided to not try that size  and I moved on to the 000712  SXT data.  Once I created maps and sharpened the data, I tried to run the data through, but FindSADs decided not to work this morning.  Sabrina made a bunch of changes to all of her programs so she warned me late yesterday afternoon that I would probably encounter some errors.  She fixed the problem pretty quickly so I went back to running that data.  When I finished with that I moved to the 990508 LASCO data.  When I finished with that I started downloading the 990508 SXT data.  After I ran the SXT data through, I moved on to the 000805 LASCO data.  Next I downloaded the 000805 SXT data.  When I created a data cube, I found out that there were over 600 64x64 images and over 800 128x128 images.  I decided, therefore, to run both sets of images.  The 128x128 set looked like it had a bunch of bad images so I moved on to the 64x64 images.  I think that I might actually need to be extracting a region so that I am using images with the same filter and resolution, but the commands for that in Sabrina's manual haven't been working so I assumed because the data was already prepped I did not need to do that step.  I think I will be asking her about that tomorrow because I only have two flares left and then I will need to meet with her anyway.  After this realization, I moved on to the 010403 LASCO data.  After finishing running that through FindSADs, I downloaded all the SXT data for 010403.  Then I created a map that was 64x64.  I also talked to Sabrina before I left and she said the extraction was meant to be for unprepped data, but she would think about it tonight; therefore, everything is fine and I can start with the TRACE data tomorrow for all of my flares. 

Wednesday July 16

Sabrina sent me a message last night saying that the program does not work for already prepped fits files.  She said for now I could just use the rm_bad_files program to remove the filters/resolutions I don't want.   I tried using this program for 010403 but it requires me to type in the files I want to remove by hand.  There are over a thousand images.  After I finished doing this, I still had files I did not want in the data.  Since it took me forever to do this and it still did not work, I decided to just download the SXT data for 010404 and then move on to the TRACE data for my flares.   I will probably have to go through the flares again for the SXT data since there are filters and resolutions I do not want included in my data but hopefully I can come up with a better method of doing this later.  I then downloaded all of the TRACE data for the following flares: 980420, 031104, 020421, 020723, 010402.  I stopped with the TRACE data there and went back to redo the SXT data for all of my flares because Sabrina created a new program to extract the correct data.  I started back at the 980420 flare.   I made a map for the 64x64, Full_Al12_1.00s extraction (resolution, filter used, exposure time), sharpened that data, and ran it through FindSADs.

Thursday July 17

I started my day by going to a talk about what to expect in grad school.  After that, I realized I never ran the 010404 LASCO data so I decided to do that before moving back to the SXT data.  Next I went back to the 980420 SXT data to see if there were any other extractions I wanted to run through FindSADs.  Instead of trying to keep track of everything I have done by typing it all up in paragraph form, I decided to make a table for the SXT data.  These aren't all of the extractions that I took, but after looking at each extraction, some of them did not looking promising; therefore, these are only the extractions I ran through FindSADs.  The promising column means whether or not the troughs found in SXT might match up with the LASCO troughs.

Flare
Size
Extraction
Promising
980420

64x64
Full_Al12_1.00s
No
64x64
Half_Al12_1.00s
No
000712 (first flare)

64x64
Full_Al12_1.00s
No
128x128
Full_Al12_1.00s
No
128x128
Full_Be119_1.00s
No
000712 (second flare)
64x64
Full_Al12_1.00s
No
64x64
Full_AlMg_1.00s
No
64x64
Full_Be119_1.00s
No
128x128
Full_Al12_1.00s
No
980509
64x64
Full_Al12_1.00s
No
64x64
Half_Al12_1.00s
No
128x128
Full_Al12_1.00s
No
990508
64x64
Full_Al.1_1.00s
No
64x64 Full_Al12_1.00s No
64x64
Full_AlMg_1.00s
No
64x64
Full_Be119_1.00s
No
192x192
Full_AlMg_1.00s
No
000805
64x64
Full_Al.1_1.00s
No
64x64
Full_Al12_1.00s
No
64x64
Full_AlMg_1.00s
No
010403
64x64
Full_Be119_1.00s
No
010404
64x64
Full_Al12_1.00s
No
64x64
Full_Be119_1.00s
No
128x128
Full_Al12_1.00s
No
128x128
Full_AlMg_1.00s
No

Friday July 18

I started today finishing the SXT data for 980509.  After that I started working my way through the rest of the table above.  I just added all of my results from today to the table above so that all of the SXT data would be in one place.  I want to look more at the 010403 data on Monday because there might be a connection between the LASCO and SXT data for that flare.  I also determined what extractions I wanted to use for the 010404 flare, but I didn't run them through FindSADs yet. 

Monday July 21

I started my day by making some tables of information for Sabrina.  My tables included all of the flares I've been working on and the directories they are located in.  One of the tables I created was basically the table above.  I also created a table with a list of the other flares that I will be looking at TRACE data for.  After that I looked back at the 010403 SXT and LASCO data.  I don't know what I was looking at on Friday, but I don't think there is any connection between the two.  When I was done looking at that, I moved to the 010404 SXT data.  The results are in the table above.  After that, I realized I needed to redownload some of the TRACE data.  I didn't select a specific filter the first time and we don't have a program to extract seperate filters.  Instead of redownloading all of the stuff I already did, I just started with the flares I could not get SXT data for: 031104, 020421, and 020723.  Unfortunately, after I spent a good deal of time downloading this data, Sabrina told me during our meeting this afternoon that I don't need to download TRACE data for 020421 or 020723 because they have already found SADs in that data.  During our meeting, Sabrina looked at the LASCO data for all of the flares.  Some of the flares that could be promising we don't have enough data for so I will search for more of that eventually.  The thing I need to check first deals with the 000805 flare.  She thinks this flare could be good in LASCO also, but she wants me to run difference the images instead of run-mean differencing.  I spent the rest of my day resharpening the 000805 data and then running it through FindSADs. 

Tuesday July 22

This morning I looked back at the newly sharpened 000805 data.  It looks a lot better but I'm still not sure about whether or not FindSADs detected any good troughs.  After that I looked for some more data on the LASCO website for both 020421 and 020723.  I can get more images but only if I change the parameters a little bit.  I have to go to the weekly Hinode meeting this morning soon so I will have to ask Sabrina what we can change.  More images might make it possible to detect some troughs because the images will look better after being sharpened.   Sabrina told me at the meeting to try downloading the unprepped data because that will give us more images and we can see if it looks good.  After the meeting I went to a lunch for just the women in the department.  They wanted to answer any questions we had about grad school.  When lunch was over, I downloaded the unprepped LASCO data for both of the flares.  I need to talk to Sabrina about whether or not these look good and then finding a method to prep LASCO data because right now we don't know how to do it.  While I waited for this, I started running the TRACE data for 031104, but the do_trace_prep program kept spitting out all kinds of errors.  I decided to go back to downloading TRACE data.  I started with 980420.

Wednesday July 23

I never got an e-mail yesterday about the TRACE data.  To download data for TRACE you have to send in a request and then they e-mail you a link so that you can ftp the data.  Usually I get an e-mail within five to ten minutes, but I still haven't gotten one for the 980420 data so I am going to retry and see what happens. Okay, so I think I typed in the wrong e-mail yesterday.  Anyway, so far today I have downloaded the following TRACE data: 980420, 000712 (both flares), and 980509.  After lunch I continued to download data.  I finished up the rest of the flares: 990508, 000805, 010403, and 010404. 

Thursday July 24

I started my morning by attending a lecture on ethics in science.  After that I started to play around with some stuff for my final presentation.  I looked at some examples they gave us and started working on my title page.  I took a break from that after finishing my first page.  I started searching the LASCO website again for anything that might give us more images for the 020421 and 020723 flares.  Then I checked the data I have already analyzed for both flares.  I realized that the extraction I took for the 020723 flare is a little off from where the known downflows are.  Sabrina told me earlier this week that the downflows are in the NE and I was just looking in the east so I tried rerunning the data through FindSADs.  I still did not find any SADs though because there are only 6 images or so.  After lunch I decided to try working with the final presentation some more.  I am meeting with David late this afternoon to talk about the presentation.  After successfully figuring out some html to create parts of my presentation, I decided to reread one of David's papers so that I can get some background information for my presentation.  I spent most of the summer running programs and watching videos of flares so I don't feel like I know that much about the SADs yet.  The paper I read was Signatures of Reconnection in Eruptive Flares. I met with David after that and he told me to review that paper before we meet again tomorrow morning so I'm glad I got that started!

Friday July 25

This morning I finished reading David's paper.  After that we met to discuss my presentation.  David explained a lot of the background information on SADs to me.  He gave me some direction for this presentation.  He also gave me some movies and pictures that I could use to help explain what goes on during magnetic reconnection.  I also got another paper to read or skim through that might help me out.  My goal for today and next week is pretty much putting together my presentation.  I'm given 30-40 minutes to speak plus another 20 minutes for questions and discussion.  It will take me quite awhile to prepare the html presentation and figure out what I'm going to say.  David also sent an e-mail to someone who used to be at MSU and worked on LASCO data a lot.  He might be able to help us with the unprepped data.  If we get a response from him sometime soon then I will work on the 020421 and 020723 LASCO data to see if we can find anything.  I spent most of the day working on my presentation and reading.

Monday July 28

I started working on my presentation this morning.  I can't figure out how to get symbols such as pi and sigma to show up when creating a webpage.  I really want to type in this one equation but I can't get it to work.  I've tried typing up the equation in Word and copying and pasting but I've had no luck.  I continued moving forward and so far everything is looking pretty good.  Now I'm stuck until I figure out where I want to go next.  I started working on a table that David wanted me to create.  It has information about all three of the instruments I used this summer: LASCO, SXT, and TRACE.

Instrument:
LASCO
SXT
TRACE
Already prepped?
Yes
Yes
Yes
Fits format?
Yes
Yes
Yes
Resolution:
11.9
 1.22750 (64x64, full res)
-
2.45500(64x64, half res or 128x128, full res)
Smaller extractions necessary?
Yes
No
-
Blow up the images?
Yes
Yes
-
Extract diff filters/resolutions? (after or before dl data)?
No
Yes(after)
Yes(filters before, resolutions after)
Best sharpened data?
If there are more than 100 images, use srmfmap, otherwise srfmap
If there are more than 100 images, use srmfmap, otherwise srfmap
-
Website for dl data:
http://stereo.nrl.navy.mil/cgi-bin
/swdbi/lasco/img_short/form
http://solar.physics.montana.edu/ylegacy/ http://trace.lmsal.com/trace_cat.html

I need to add some more data to this table but I haven't figured out what else I need to add yet.  It will also be formatted differently for my final presentation next week. 

Tuesday July 29

I started my morning by just looking at my presentation and trying to figure out where I want to go with it.  I think I've got some ideas for what to work on this afternoon but right now it's time for my weekly Hinode meeting...After the meeting I went back to my presentation.  I'm having some problems with slide 7 and the movies on that page.  The movies are working but the LASCO video skips certain images unless I use Firefox to open the website.  I spent a lot of time on this slide but decided to come back to it at the end of the week when I'm fixing last minute problems.  Now I'm also having problems getting images for my website.  I create the images, but then they have a huge black box around them that takes up too much space.  I tried using sub_map to clip off the extra space, but so far I'm not having much luck.

Wednesday July 30

David got an e-mail back from Tim Howard about the unprepped LASCO data.  Tim said he uses all of the level .5 data and has never prepped it.  That being said, we thought we'd run the 020421 and the 020723 flares again with the unprepped data.  All is not well this morning, however.  I keep getting errors when I run the 020421 data through FindSADs.  OK, I figured that out but when I ran the data through FindSADs I still did not find anything.  I didn't even bother running the 020723 unprepped LASCO data because there is less unprepped data than there is prepped data.   After I finished messing around with that data, I moved back to my presentation.  I feel like I accomplished a lot before lunch.  I continued working on the presentation after lunch.  I've got the last several slides planned out but I didn't get them all done yet.  I had some troubles picking good videos for a few of my slides.  I have a meeting with David tomorrow before lunch so hopefully he can answer a few questions and let me know how things are looking. 

Thursday July 31

Today I went right to work on my presentation.  I met with David late in the morning and we went through all of the slides I have created so far.  We came up with a few changes here and there.  After the meeting I started modifying some of my slides.  I was trying to create an mpg movie for some original LASCO data but it isn't working.  The movie keeps showing up all black.  I need to figure out how to fix that tomorrow.

Friday August 1

Back to work on the presentation.  I went through the whole thing and changed/fixed all of the things David and I talked about.  Slide 10 is still causing some problems with the videos but other than that I think I'm done.  I want to meet with David on Monday sometime to run through the actual presentation and show him the changes I've made.  Wednesday is my actual presentation so I've got some time to fix any last minute problems. 

Monday August 4

This morning I started looking at my presentation again.  It's pretty much done, but I have not figured out what I'm going to say yet.  I know more or less what I'll say but if I don't think about it ahead of time my presentation will be horrible; therefore, I spent most of the morning trying to figure that out.  Sabrina also sent me an e-mail saying she wants to see what I have so far for my presentation sometime today.  After I met with Sabrina, I had a lot of changes to make.  Most of my videos were not high quality so we tried to resolve that problem.  I spent most of the evening rearranging slides, replacing videos, etc. 

Tuesday August 5

I spent this morning sitting through two of the REU students' presentations.  Last night I finished up most of my presentation.  I just sent Sabrina an e-mail asking her if she would look through it again because David won't be in until I present tomorrow.  This afternoon I will be sitting through three more presentations.