RHESSI Nugget #384

Wednesday 7th October
Article: RHESSI Nugget 384
I was asked by Hugh Hudson to do a Nugget for the RHESSI people despite it being a little out of field. I of course accepted and because I have simultaneously been working on the research group website, I had the urge to write it here, the most inconvenient format for such a thing. The text was edited by Hugh Hudson for the release on the RHESSI site, so presented below is the original version using my words.

Differential Rotation within a Sunspot Umbra

Sunspots have long been associated with solar flare activity and are — as manifestations of the solar magnetic field — a key aspect of solar physics. In fact, flares and sunspots are so intimately linked that there have been several studies trying to associate the configuration and motion of sunspots with the occurrence and strength of flares. They were for the most part successful; as it was indeed found that the more complicated spot groups tend to have larger and more frequent flares. However these were not the only factors contributing to the flaring chance of a region, as evidenced by the several outliers that produce large flares despite being smaller and with fewer polarity inversion lines.

Finding out why there are these outliers and defining what exactly causes one region to flare and another one not to is unfortunately not going to be answered here. What is investigated is one particular event that occurred in NOAA active region 12158 on 10th September 2014, using a new technique that allows us to study the motion of the sunspot group in greater detail than before. In this particular case my technique, Multi-Layer Thresholding (MLT), tracks the rotation rate of sunspots in a sunspot group to build up a picture of the evolution of the region before, during, and after a large solar flare.

Tracking the rotation of sunspots can be done using ellipse fitting. In this technique an ellipse is fit to the edge of the sunspot umbra, and the angle of the major axis of the ellipse to the normal is tracked over multiple images to build up a rotation profile of the sunspot. MLT is an extension of this technique designed to be adaptable to different configurations of sunspot groups, and to lessen the bias in determining where the edge of the umbra is. In MLT the ellipse fitting procedure is not just applied to the edge of the umbra, but at all the key locations in the sunspot group. This is achieved by combining several iterations of thresholding.

Figure 1. Example of how MLT builds up a picture of a sunspot using layers.

A sunspot image has one iteration of thresholding applied to it at a percentage of the average quiet Sun intensity (e.g. 55%). All pixels darker than this threshold are found and then grouped into clusters using the OPTICS clustering algorithm. These two steps are repeated for a number of different threshold levels. At the end of the process the clusters in each layer are matched with other clusters in the layers below them, and suddenly we have a powerful tool for sunspot analysis (Figure 2).

Figure 2. NOAA 12158 with outlines of the clusters found by MLT. The green (55% quiet sun intensity) and blue (35%) ellipses treat the sunspot umbra as a whole, whereas the lower layers of purple and pink (20% and 15% respectively) show that it is made of two separate regions.

As an example of MLT in action and of how it can be useful to us as scientists, we applied the method to the sunspot in NOAA active region 12158. On 10th September 2014, an X1.6 flare occurred above the lower region of this sunspot (Figure 2). During this time the sunspot also underwent a dramatic change in rotation rate - to the point that the sunspot was seen to reverse the direction of rotation after flaring.

Applying MLT to this region reveals it's complexity; not only does it have a ring of smaller sunspots (pores) surrounding the main umbra, but the main umbra itself is split into two regions. The top part of the spot is a much smaller region that is distinctly different from the lower region — as shown by the rotation profile — which is completely independent of the lower region's rotation. The lower region (Figure 2) is a much larger and darker region, hence it is made up of more layers. It also exhibits a sharp rotation across all layers that changes the rotation of the spot by approximately -7 degrees an hour. This sharp change in the rotation coincides with the X1.6 flare, originating above this region of the sunspot.

Figure 3. Velocity against distance from the centre of the umbra. Points on the left are closer to the centre than points on the right.

The fact that there was a rotation during the flare is nothing new - others have seen the same thing at this same sunspot. What MLT found was a discrepancy between the rotation rates of the innermost clusters and the clusters at the edge of the umbra. The clusters in the centre of the umbra appeared to be rotating faster, getting progressively slower to the edge of the umbra. (Figure 3) shows the velocities of clusters against distance from the centre of the umbra. The red circles show the velocities of the clusters in the 4 hours before the flare, which have a consistent median velocity of around 1 degree per hour anticlockwise. When the flare occurs there is a sudden change in all the velocities, as shown by the black/white crosses. These points show a significant displacement in the clockwise direction, and more so there is a gradient in the median velocities with distance to the centre. This indicates that the points on the outer edge of the umbra are on average rotating slower than the clusters in the middle.

The discrepancy in rotation could be due to a number of factors; there have been studies showing that solar flares can seem to cause a rotation in small sunspots, which could be the case here. The X-class flare is known to have originated approximately in the centre of the umbra in question, hence we could be seeing the result of this eruption propagating along the spot. Or perhaps we are seeing some unusual structure within the emerging twisted flux rope, giving the appearance of differential rotation as it rises from the interior. It's also possible that this is due to a more complicated set of interactions that we have yet to consider. In the end, we reach the only too common conclusion in science: we need to study more cases.

This website uses cookies. By using the website, you agree with storing cookies on your computer. Also you acknowledge that you have read and understand our Privacy Policy. If you do not agree leave the website. More information about cookies


Enter your comment. Wiki syntax is allowed:

physics/blog/1_rhessi_nugget_384.txt · Last modified: 2021/04/05 16:56 by richard