Making a start ... version 2

[Averton]

This is the second attempt at posting this information.
After we posted the first time, our post showed up on the forum but when we checked again today it was gone. Apparently this has to do with taking too long with writing the information which would be quite true in this case. The first time was early this morning so the likelihood of the contents being the same this time round is very small, but here goes.

After the meeting yesterday, we sat down to try and clarify where we were up to with classifying prominences. Prior to yesterday's meeting we had largely used this article from DayStar as our guide for internal (for us) classification of prominences and filaments.
http://www.company7.com/library/daystar ... intro.html
We had found it adequate for our own purposes. As we said in the meeting we deliberately haven't publicly classified our images of prominences due to our perception of a lack of a definitive classification system.

Trying to get a better understanding of the classification issue, yesterday afternoon we did a quick on-line search for previous work in the area. (We're sure that a more thorough search would uncover a lot more information.)

This article gives a brief history of prominence research up to 1998.
https://articles.adsabs.harvard.edu/pdf ... .150...11T

We found this next article, a useful start.
https://articles.adsabs.harvard.edu/cgi ... etype=.pdf
The earliest classifications appeared to be done by Secchi after the development of the spectroscope around 1868, with only two categories, quiescent and eruptive.
With the development of the spectroheliogram, Edwin Pettit (circa 1930)created a scheme of classification that had six main classes.
The classes were Active, Eruptive, Spot, Tornados, Quiescent and Coronal. The categories in this system don't look to be a good starting point as they are unclear and inadequate.

The next system was the Menzel-Evans scheme of classification (1953). Their classifications were made using motion pictures.
This is interesting considering the comments made at the meeting regarding the benefit that animations give to identifying what is actually
going on in a given prominence. This system used just 3 letters to classify any prominence. The first character was for the motion of the material, the second was for whether it was associated with a sun spot and the third was the form of the prominence.
Maybe the categories need a revision but the elegance of the lettering system is appealing. Such a system could be a protoype to consider.

The article finishes with the work of F Shirley Jones, whose study was to measure and classify prominences with both motion pictures and still frames. The conclusion was that still frames did provide a matching classification to the motion picture version in 78% of the data.

A final article we looked at had an interesting quote regarding the challenges of classifying prominences/filaments (she uses the terms interchangeably).
https://link.springer.com/content/pdf/1 ... pdf=button
"Decades of study have shown that it is very difficult to characterize filament properties. Filaments
show differences in morphology, lifetime, position on the solar disk, complexity of their magnetic
field environments, etc. They are not uniform in shape and show a fine, dynamic structure at the
limit of the instrumental resolution. This high variability makes their classification difficult and also
results in a wide range of physical conditions deduced from observations that poorly constrain the
models of prominence formation and disappearance (Vial, 1998)."

Finally we found a book titled "Solar Prominences" by Oddbjørn Engvold, with a chapter Description and Classification of Prominences.
https://link.springer.com/chapter/10.10 ... -10416-4_2
Does anyone have this book?
Is it any good?
Does this chapter have any relevance to the project of classifying prominences?

We are not sure if this information is pertinent or helpful but we found it interesting none the less.

[Montana]

Fabulous work I agree totally.

The image from Engvold is included in the January solar challenge if you want a quick peek, it was the only snippet available

Alexandra

[rigel123]

Great research! I have always used the Daystar basis of commenting on any of my prom shots and of course since I mainly do animations I am fairly confident (but wrong at times!) as to whether a prom is actually lifting off or a rather active quiescent prom and if there is any rotation. I really like the idea of a Tree that Alexandra alluded to during our meeting with probably two trees. The trunk of one being Quiescent and the other Eruptive, then the branches are the basic morphology, Trunk, Arch, Hedgerow, "Slug", "Comb Over", etc for quiescent and Surge, Spray, Spindle, Loop, etc. You could even have at the branches such as for separating Surge and Spray comments such as "Plasma rises and falls back in place" for Surge and "Plasma rises and escapes" for Spray so folks would know that they may need to view over a period of time and maybe include a link to an animation of each.

Just thinking out loud and some of my old college courses that we would use Keys for flowers, mushrooms, trees, etc are coming out!

[Montana]

You guys are probably right, perhaps we need to produce a key diagram first and try to decide on a vague classification chart which we can then adapt along the way.

I'm a bit bogged down with work this week, but if we can all try to sketch a chart (basic don't spend too much time), we can combine, adapt and thrash out a basic outline. We will always forget something and remember it later, or someone else will spot a missing bit. But a working plan probably needs to be done first.

Alexandra

[Averton]

You guys are probably right, perhaps we need to produce a key diagram first and try to decide on a vague classification chart which we can then adapt along the way.

I'm a bit bogged down with work this week, but if we can all try to sketch a chart (basic don't spend too much time), we can combine, adapt and thrash out a basic outline. We will always forget something and remember it later, or someone else will spot a missing bit. But a working plan probably needs to be done first.

Alexandra

We are happy to try and rough out something but we still wish to clarify a couple of things before starting.
1. Is the categorisation system going to be designed from the outset to do both motion images and still images?
2. We feel the need to clarify at the most basic level terms like quiescent due to technological changes that have altered the meaning in this context, see another post.

[rigel123]

So I started just sketching out a Key that came to mind and quickly saw that it could be quite convoluted quickly! But here was the example that I started just to help stimulate better lines of thought as to how something could be laid out.

Prom-Classification.jpg (277.74 KiB) Viewed 2760 times

[Averton]

So I started just sketching out a Key that came to mind and quickly saw that it could be quite convoluted quickly! But here was the example that I started just to help stimulate better lines of thought as to how something could be laid out.

Prom-Classification.jpg

Very helpful Warren. You only realise just how difficult this is when you try to map it out in this way and you are right, it gets convoluted very quickly.
As you go further down the tree, the decisions become less cut and dried. Considering this and our recent research we are beginning to wonder whether a flow chart/tree/key approach is only applicable to the top levels of classification and then after that, a selection list (tick boxes) of features that could apply. This gives the possibility that a given image has an overall categorisation plus one or more sub-categorisations.
Good work.

[rigel123]

So I started just sketching out a Key that came to mind and quickly saw that it could be quite convoluted quickly! But here was the example that I started just to help stimulate better lines of thought as to how something could be laid out.

Prom-Classification.jpg

Very helpful Warren. You only realise just how difficult this is when you try to map it out in this way and you are right, it gets convoluted very quickly.
As you go further down the tree, the decisions become less cut and dried. Considering this and our recent research we are beginning to wonder whether a flow chart/tree/key approach is only applicable to the top levels of classification and then after that, a selection list (tick boxes) of features that could apply. This gives the possibility that a given image has an overall categorisation plus one or more sub-categorisations.
Good work.

I like that thinking. Maybe start at a high level with Quiescent and Active with specific criteria to separate them and then go to your tick boxes such as shape, number of attachment points, movement over time, etc. This will be fun….and challenging!

[Averton]

So I started just sketching out a Key that came to mind and quickly saw that it could be quite convoluted quickly! But here was the example that I started just to help stimulate better lines of thought as to how something could be laid out.

Prom-Classification.jpg

Very helpful Warren. You only realise just how difficult this is when you try to map it out in this way and you are right, it gets convoluted very quickly.
As you go further down the tree, the decisions become less cut and dried. Considering this and our recent research we are beginning to wonder whether a flow chart/tree/key approach is only applicable to the top levels of classification and then after that, a selection list (tick boxes) of features that could apply. This gives the possibility that a given image has an overall categorisation plus one or more sub-categorisations.
Good work.

I like that thinking. Maybe start at a high level with Quiescent and Active with specific criteria to separate them and then go to your tick boxes such as shape, number of attachment points, movement over time, etc. This will be fun….and challenging!

We'll try and put something together shortly.

[Montana]

This is really great! I have lots of ideas too, I only wish I had some free lunchtimes. However, I have a student who is in such a bad way I am having to work every hour to help as he is running out of time. Maybe the weekend I can have an attempt too. Great work in finding the prominence chapter by the way

Alexandra

[Averton]

We had a go at trying to draw up a classification chart from the information gathered thus far. It is just a draft for discussion.
The categories above the line, which apply to both filaments and prominences, seem to be agreed on in most documents, so it probably only requires a decision as to which is the best term to use for clarity.
The other question is whether Quiescent should be broken in two so you just have three categories. Also in the MNL Type C category, from what we can find there are only polar crown proms but their form/shape still would need to be classified and appears to be the same as the MNL Type B.
Below the line is just a collection of terms that have been used in various documents and at different times. No reason why we can't add slug and comb-over
Of interest is that it appears that Surges and Loops are no longer considered as prominences at all so it leaves very few Active/Eruptive proms, if any? Not sure whether that means there should be a separate upper level category for "Active Region jet phenomena and post flare loops". This would mean the categorisation system would be more than just prominences.
Maybe chromospheric features?
The overall idea is that you would just select the top level category and then tick/select from the shape form descriptions to match your observation.
We can't see any reason why you couldn't tick or select multiple shapes.
Hope this kind of makes some sense.

Solar Prominence categorisation draft V0.1.pdf

(57.04 KiB) Downloaded 71 times

[rigel123]

This is a great start, and I’m so glad they pulled surge proms and other eruptive events away from prominence categories. I never really felt as if I was viewing a prom but rather a jet or eruption of plasma! So do all prominences form outside of active regions?

You might find this paper interesting as well. Of course they are utilizing wavelengths we can only access from the SDO site but we know these events can be captured in Ha as well.

https://iopscience.iop.org/article/10.3 ... 357/aa7945

[Montana]

Mmmm I definitely feel we need to include surges and loops as they are so often observed on the limb. They are equivalent to arch filament loops within active regions. However, we could have two groups on the active side, general active region phenomena and flare related phenomena. During flaring we get ejection of large amounts of material as in spray, loops. Whereas general active region phenomena are arch filament loops, surges. We also need a category explaining mottles/fibrils and spicules.

Can you remind me what the MNL stands for? is it magnetic neutral lines?

One thing I think I would prefer is the use of actual names, letters and codes are hard to remember especially for a beginner whereas names are a bit easier and less intimidating, what do you think?
I need to go and get a piece of paper and jot down my ideas too.

Alexandra

[Montana]

I had a doodle attempt this evening, although I ran out of room on both sides and had to extend the left and curl the right downwards (apologies). I am sure there is plenty I have forgotten

DSC03960 crop.jpg (930.86 KiB) Viewed 2697 times

Alexandra

[rigel123]

That looks great Alexandra, kind of what I was trying to put together. I think spicules and maybe mottles would be moved from Quiescent since spicules only last minutes but not sure where you might want to put them on the chart. Great base to work from! And do we want to include filaprom somewhere or not even worry about that and just classify them under Slugs and Combovers?

[Averton]

I had a doodle attempt this evening, although I ran out of room on both sides and had to extend the left and curl the right downwards (apologies). I am sure there is plenty I have forgotten

DSC03960 crop.jpg

Alexandra

Great work Alexandra! This really clarifies for us what you were aiming to achieve in the first place. We have definitely been barking up the wrong tree. The only thing we would question is calling the project a Prominence/Filament Categorisation. It would be good to think that the work was inherently scientifically correct. Particularly when you look at Engvold's quote below.

"Any prominent solar feature seen above the rim of the Sun was historically associated with the term prominence. Increased awareness of related phenomena led to a rich "zoo" of solar features that were subsequently classified under a prominence umbrella. This included features like flare loops, surges, coronal loops or arches, various types of mass ejections, large spicules, coronal rain and coronal cloud prominences. Surges and Loops, which occur in conjunction with flare activity, are now recognized as active region jet phenomena and post flare loops, respectively, and regarded quite different from solar prominences. Furthermore, spicules constitute rather the main structure of the chromosphere. In the following chapters Coronal Rain and Coronal Cloud Prominences remain as a significantly different types of prominence."

This looks to be more like a General Observers/Imagers Guide to All Solar Features. It reminds us of books like "A field guide to birds/trees/flora" etc. This means anything can be included and as you are not classifying it in a technical sense it can't be wrong.

Additionally we love your sketches, they are excellent! They raise a question as to whether the guide should actually be illustrated by sketches rather than by images. There is a good rationale behind this as a lot of signage uses graphics not images because they have been found to be more recognisable. In a sketch, you can choose to only incorporate the core features and not have background or distracting "noise". Sketches allow for the accentuating/caricaturing of the essential feature greatly increasing recognition.

Excellent work, this is the way to go!!

[Montana]

Interesting thoughts!!

Warren, you are probably right about the spicules, they classed under the 'quiet sun' heading, hence I put them on a branch of the quiet Sun, but, like Clare & Peter commented in modern days we can see they only last 10 minutes (fast Sun) and the reason they thought they were unchanging was because they didn't do timelapses. This is a really interesting observation.

Clare & Peter, I do feel we need to keep scientific (keep respect in that field) and if they say that flare related phenomena are not prominences then fair enough I think we need to be clear on the labelling, however observers do see them on the limb and wonder what they are and this is really important too. I think this is the beauty of the tree diagram, they become an arm on their own to describe. Perhaps both, this group and the spicule group need to have their branches more clearly defined and described, what do you think?

Once we get a clear consensus on this we could properly draw out the diagram, I think sketches do help for the diagram, but then we need a good page of images of each type to show the variability in real life, it is much harder in real life. Maybe if we could agree where the branches could go (add any missing ones) we could then set up a thread for each type, then if someone gets an image or has one already they could add it to the thread. This way we would have the images sorted and tagged

Alexandra

[Carbon60]

This is great work and all very interesting.

Just a few thoughts from me, thinking specifically about prominences as seen above the limb.

I think we need some metrics:

1) Position (angle around the disk). Maybe a little subjective with some proms, but we could aim for the middle of the prom as the datum. This value can be measured using, for example, Tilting Sun.

Tilting_Sun_Disk.jpg (200.05 KiB) Viewed 2663 times

2) Width of the prom. This could be in km, or the angle subtended from the solar centre.
3) Height of the prom in km

A solar scale tool such as this one could be used for this purpose:

Solar Scale_Inverted.jpg (613.99 KiB) Viewed 2663 times

4) Activity (quiescent or active). This is a little more tricky, but we could propose that we look at the prom after a period of time (say 1 hour, or 2 hours...or whatever) and then note the difference. I've played around with difference imaging in the past and maybe this could be applied in this situation. Basically, two images are compared by subtracting one from the other to see where things have changed. Quiescent proms will show minimal change and active ones will show a lot of change. I've used a couple of images from one of Warren's animations (hope this is okay, Warren). One image was grabbed from the start of the animation sequence and the second was grabbed towards the end. Subtracting one from the other easily shows which of the two proms is active and which is quiescent, since active proms show more contrast changes as a result of their motion which is enhanced by image processing. Using the difference image alone we can easily convey the dynamics of the prominence.

The images are shown below.

1st still image:

Warren1.png (788.86 KiB) Viewed 2663 times

2nd still image:

Warren2.png (750.12 KiB) Viewed 2663 times

Difference image:

Warren_Velocity.png (525.88 KiB) Viewed 2663 times

Clearly, in this example, the lower prom is active, whilst the upper prom is quiescent.

Obviously we would also need to describe the form of prom, much as is done today...arch, hedgerow.......etc and here we might suggest further categories depending on what we see from our dataset of prom images.

Stu.

[rigel123]

Good points Stu, no problem using my images! I like identifying position and maybe for visual folks we could suggest quadrants on the sun such as NE, SE, NW, SW, etc. Of course for many we need a simple rule as to how to determine what is North and South to begin with!

That particular event was pretty violent and watching the entire animation I would definitely categorize it as an eruptive jet or spray.

[Carbon60]

That particular event was pretty violent and watching the entire animation I would definitely categorize it as an eruptive jet or spray.

It sure was active and the image with both proms is great example, in my view, demonstrating the difference between quiescent and dynamic features. One would never know from a regular single still image.

Thanks for letting me use your images

BTW, this works for full disk images too, easily showing where the main activity is really happening.

Stu.

[rigel123]

Thinking about Clare and Peter's comments about using the sketches to identify the prom type is a good idea. Then we could include images of "typical" types, such as Hedgerow, Fan, Tree, etc. Possibly include some animations with elapsed time to show how quiescent can vary. An example is a typical Hedgerow shows gentle sometimes imperceptible movement, while a Tree or an Arch can show a lot of activity in a short time, but still lasts hours and would be classified under quiescent. And of course the "Active" or whatever term we decide upon change and erupt in a matter of minutes although the aftermath can last up to an hour or more. Spicules and Mottles could be a separate category of short term, non-eruptive events.

[Montana]

I've had another attempt today and changed the branches with regards to some of the comments.

I've changed to make a separate branch of quiet sun vs active sun, to indicate the zones. In the active sun this is broken down into two 'Quiescent features' and active region features. I think quiescent filaments are still quite active and are in the active region zone. Active region associated features still show up on disc and limb and leave an observer wondering what they are.

Stuart's use of tilting sun is great to work out where the active region zone is and whether a filament is diagonal or horizontal.

Is this getting any closer? any changes needed or forgotten something?

DSC03962 crop.jpg (1.25 MiB) Viewed 2565 times

Alexandra

[rigel123]

That looks great Alexandra!

[oriolestation]

I had a doodle attempt this evening, although I ran out of room on both sides and had to extend the left and curl the right downwards (apologies). I am sure there is plenty I have forgotten

DSC03960 crop.jpg

Alexandra

Where's Waldo? ......(sorry, couldn't resist )