I have changed the graphs to psym=10, included /ystyle and change the flare rate to 24 hour periods. I have also started printing to postscript files instead of jpegs. WordPress does not allow eps files however so i had to convert back to jpeg and some quality was lost.
Graphs have been cosine corrected and updated to include the flare times (for area) and flaring rate per 12 hours (for flux and flux emergence rate). It can be seen that flaring is most frequent when area and flux are large. In the flux plot we see the flaring rate increase as the flux increases and we then see the flaring rate die off again as flux disappears.
From the flux emergence rate we can see that there is a rapid flux emergence prior to flaring. As the flaring rate peaks, the flux emergence rate begins to fall away.
Video of AR 11861 tracked slowly showing a ‘+’ where the gradient is at maximum for both the normal magnetogram and the gradient magnetogram maps. (Make sure to watch in HD, also the horizontal black lines which sometimes occur in magnetogram data seem to be caused be graphics issue on my laptop)
The highest gradients are nearly always found along neutral lines between nearby emergent flux regions of opposing polarity. During these times there is lots of flaring. Towards the second half of the video flares are more sparodic and the maximum is sometimes found in uni-polar features. The maximums themselves are lower as the flare rate dies off at the end as well which can be seen in the figure below. The figure also clearly shows that when the maximum gradient is high for the field of view there is more flaring activity.
Just a quick update about what I’ve been doing this week. I have been trying to run YAFTA (yet another region tracking algorithm) and mpole to look for flux emergence and plot field extrapolations. The code takes a long time to run and I have been running into some errors before it outputs the final data cubes but I’m hoping to have it running successfully by the end of the week.
Below are some images of the gradient of the magnetic field. The first is the gradient clipped at 0.25 Gauss/km and 1 Gauss/km to remove noisy low gradient features. Circular features due to the gradient within the sunspots are still produced.
Using a mask and taking only values near 0 Gauss produces a map of the gradient along the neutral line. In the first image (below) the data is masked for values of magnetic field greater than +/- 80 gauss and the gradient is clipped at 0.03 Gauss/km and 1 Gauss/km.
In the second image, the data is masked for values of magnetic field greater than +/- 200 Gauss and the gradient is clipped at 0.15 Gauss/km and 1 Gauss/km.
One other thing worth noting is that the gradient data appears to be about 4 or 5 times that of what is found in your method mentioned in your paper here
The improved resolution of the magnetogram data may account for this difference