Hello,
I'm doing the "Correlation of Coronal Mass Ejections and Solar Sunspot Cycle" project (https://www.sciencebuddies.org/science- ... y&from=TSW) and I'm currently conducting the experiment. I am having some difficulty understanding the experimental procedure, I understand everything before step number 4, but there is a section that says to determine the range of the monthly CME data. What I don't know is that if I should find the range for each month of each year, or, the range of each year, or, the range of each month for all the years, or, the range for the whole data. Also, I would like to know where I could download graphing programs to do this project.
If I read the writeup correctly, when you scan the range of CMEs in step 4 you are looking for the range of CMEs per month in the whole data set. Your purpose is really to determine the range for theY axid for your plot of CMEs vs. time to compare with the sunspot plot.
As far as finding free downloadable software, the writeup mentions SigmaPlot, Origin, and one or two others. If you do a Google search on these product names you will find multiple sites from which you can download. These are programs which are commercially sold, so to use them over the long term ethically you need to buy them. Some may offer a trial version that you can use for thirty days or so to see if they suite your needs. That should be long enough to do your project. You just need to be sure to uninstall them within the trial period and you shoud be fine.
Hi,
Just wanted to add as far as the software goes that Open Office Calc would probably suffice for this project. It is completely free, with no trial period, and can make plots with secondary Y axes as suggested in this project. The other Open Office programs that come with Calc are also very good. http://www.openoffice.org/
For other free options, try a search for "open source graphing software" and keep an eye out for gnuplot, R, and others. These can be less user friendly than the open office or commercial programs.
I would add a note of caution about using expensive graphing programs with trial periods. Many or all of these programs will automatically lock after the trial period. After that, you will be unable to even open your files on the same computer, even if you uninstall and then reinstall the program. There is potential for some serious headaches.
Hi
Well thank you all for the help with my science fair project, but I still have a question. The graph for the solar sunspot cycles should start in 1964 or earlier and the CME data begins in 1996. So, why should the graph start on 1964 if the curve for CMEs starts until 1996? Or, should another x-axis be added to the graph to show different years for the coronal mass ejection activity? I would really appreciate it if you could reply back tomorrow(02/22/09).
Determining the correlation of two variables which are a function of time is actually quite tricky.
Typically, you want to start by putting time on the x-axis and use two different colors or line styles to show the two variables on the Y-axis. If you do this, the line for the CME data should start in 1996; however, getting your graphing program to not show any data prior to 1996 can be tricky. The best some programs maybe able to do is to show the CME data for earlier months as zero. A different approach would be to simply start at 1996 and stop at the last data point you have. Note: There are some months where no CME data is available and you will have to deal with that problem of how to graph it so that it doesn't imply anything visually besides that absense of data.
Another problem is scaling the amplitude of the two variables so that they both show their variation. One way to do this is to determine what the maximum value for the sunspots and the CME and assign scales to those variables so that they line up close to the same height near top of the graph.
Note: If there is any correlation between sun spot activity and CME data, there maybe a time skew involved. Sophisticated methods of determining mathmatically if there is correlation require convolution integrals using calculas which is well above your grade level. A simplified way of doing this would be to plot one of the two curves on a transparency so that you can place it over the other graph and slide it along to get either the peaks to best allign with the minimum time skew or to get a peaks on one to line up with valleys on the other with the minimum time skew.
A simpler approach would simply to plot CME data and see what the period is to any cyclical pattern and then to determine if the same period exists in the sunspot data. If the CME data isn't cyclical or doesn't have the same period as the sunspot data, you can probably conclude that they aren't correlated.
I just wanted to be sure that I created my data correctly, and this confirms it. I turned in my project today and everything seems to be going as I expected.
/-/https/www.sciencebuddies.org/cdn/Files/21464/5/sb-logo.png){kind=logo}
/-/https/www.sciencebuddies.org/cdn/Files/21464/5/sb-logo.png){kind=logo}
