CME's, active regions and high energy flares

[Michael Mozina]

Things could get interesting in about 12-24 hours.

http://sdo.gsfc.nasa.gov/assets/img/latest/latest_1024_0211.mpg

There are two regions that present major flare potentials over the next couple of days, 11121 has a new active region neighbor springing up right next to it in the southern hemisphere. That new active region right next to 11121 will grow in geographic size and result in all sorts of "circuit reconnection" events over the next few days.

The image is also interesting for another reason. If you look closely at about the 11:00 position, you can see the dark filament that results in the largest "flare" from the sun in the last 24 hours. There is also a large dark filament forming at about the 10:00 position that is rotating in our direction. If it begins to "erupt", it too could spew materials in our direction.

The sun is worth keeping on eye on today, because it has the potential to generate both types of flares, EM flares and filament eruption flares. It also shows a filament eruption earlier in the video at about the 11:00 position. If you look closely, you can see the dark filament rise into the atmosphere and "disappear", just before the arcade appears in the corona, and just before (a few hours before) mass flow appears in Lasco.

 

[GeeMack]

http://sdo.gsfc.nasa.gov/assets/img/latest/latest_1024_0211.mpg

There are two regions that present major flare potentials over the next couple of days, 11121 has a new active region neighbor springing up right next to it in the southern hemisphere. That new active region right next to 11121 will grow in geographic size and result in all sorts of "circuit reconnection" events over the next few days.

More sciency talk including a ridiculous inference that there will be "all sorts of circuit reconnection events", but still no description of your supposed quantitative objective method for "predicting" CMEs. No explanation of the claim that dark filament eruptions cause CMEs. In fact, not one single thing in this post addresses any question that you've been asked in this entire thread. Is there some particular reason you refuse to provide the description of your supposed "prediction" method, substantiate your claim that dark filament eruptions cause CMEs, or admit that your claims are unsupportable?

 

[Perpetual Student]

I know why:

 

[Michael Mozina]

Once again you are wrong. They did not change after the fact. You did not know what you wre looking at.

No RC, I know what I saw and read (for hours) on their website.

But at least you can copy and paste correctly:

I think I will make that a habit from now on just in case the site changes again.

FYI: Michael Mozina, 90% means 1 chance in 10 of being wrong. It does not mean 100% chance of being right. Only a truly ignorant person attempts to do statistics with a sample size of 1.

Actually RC, I've been "following along" very closely now for most of the past week. So far their (actual) batting average is actually closer to 50% (without allowing for after the fact modifications to the website during the day). That's not "bad", in fact in terms of EM flare prediction that "pretty good" IMO. It's not perfect of course, but it's not bad. Unfortunately they missed the first two M class flares entirely.

I'm more curious how they do for today and then for the rest of the week. They don't even have a 'number' for the small active region that I now have my eye on. They'll just probably classify the entire region as 11121, including that new small active region forming a little to the left and down from 11121.

I think however their 90 percent figures were a tad "optimistic" in terms of what 11121 was doing that particular day. Today they could luck out because of the active region forming below and left of 11121. That will eventually generate flares IMO, but I'm not sure it will happen before the end of the 9th UT. They could end up being 90% wrong again today. I'll have to wait and see. I'm also curious what they will do with tomorrows "prediction" because today looked to be pretty quiet until I saw that small active region start to form, but tomorrow looks like it will get interesting. Will they raise or lower that 90% figure? Inquiring minds want to know.

 

[GeeMack]

Actually RC, [...]

Given your ignorance of the requests to support your claims, it looks like you're abandoning them... again. Yes or no?

 

[Michael Mozina]

We would be interested in hearing your explanation of your claim that dark filament eruptions cause CMEs, including fully quantitative descriptions of the relevant data and citations to actual qualified scientific material that supports your claim.

Well, if I am going to go through all that, I might as well write a paper and organize it properly. I certainly had no intention of turning a public thread into such a "painful process" for your personal amusement. I won't do that in fact.

I think the three of you (mostly PS) have convinced me that it is in fact worth my time putting the information together in a formal format, and I will do so (AS IT WORKS FOR ME, NOT FOR YOU). I have a programming project going that has the potential to double my income next year. I have no desire to waste my valuable time arguing with you and RC over every single issue, every single step, every single citation from every single paper on a public forum. That would be *STUPID* and complete waste of my time.

In the mean time, I will continue to 'predict' flares and CME's using these techniques, because as anyone who's actually followed this thread closely know that it works. That doesn't include PS by the way since I think he's only read a couple of pages of this entire thread whenever he felt like it.

This whole process has taught me what a true waste of time you and RC have turned out to be. Neither of you is actually interested in CME or flare prediction. You aren't interested in whether or not one can use SDO images to 'predict' them. All you're interested in is arguing, and frankly I'm just plain bored silly of you and RC. Neither of you has a clue what your talking about. RC got all upset over the term "dark" filament when in fact that is the 'property' that makes them 'unique' and interesting and makes it possible to track them in the first place. You demonstrated your ignorance the moment we started to talk about mass flow and the direction of mass flow (that's actually nothing new by the way, that's par for the course with you).

 

[Michael Mozina]

Given your ignorance of the requests to support your claims, it looks like you're abandoning them... again. Yes or no?

No.

 

[GeeMack]

Well, if I am going to go through all that, I might as well write a paper and organize it properly. I certainly had no intention of turning a public thread into such a "painful process" for your personal amusement. I won't do that in fact.

You're speaking in future tense about developing a method here. You never did have a quantitative objective method for "predicting" CMEs, otherwise you wouldn't be complaining now about how painful it might be to assemble one. That's as good as admitting your previous claim to already having such a method was simply not true. That claim was not true, Michael. Your argument about having the method you claimed was a lie. Those of us who didn't believe you actually had a method (which would be pretty much everyone who commented in this thread, by the way) were correct.

I think the three of you (mostly PS) have convinced me that it is in fact worth my time putting the information together in a formal format, and I will do so (AS IT WORKS FOR ME, NOT FOR YOU). I have a programming project going that has the potential to double my income next year. I have no desire to waste my valuable time arguing with you and RC over every single issue, every single step, every single citation from every single paper on a public forum. That would be *STUPID* and complete waste of my time.

Yes, I agree it would be colossally stupid and a complete waste of your time, yet oddly enough attempting to argue every single issue is exactly what you've been doing here since you started this thread over 7 weeks ago. You certainly can't fault anyone else for continuing to ask you to support your claims when it was you all along refusing to uphold your responsibility to support them.

In the mean time, I will continue to 'predict' flares and CME's using these techniques, because as anyone who's actually followed this thread closely know that it works. That doesn't include PS by the way since I think he's only read a couple of pages of this entire thread whenever he felt like it.

Since you've acknowledged that you do not presently have a quantitative objective method for "predicting" CMEs, your current technique is guessing. All but one of the participants in this thread already knew that. Now that you admit it, you know, too.

This whole process has taught me what a true waste of time you and RC have turned out to be. Neither of you is actually interested in CME or flare prediction. You aren't interested in whether or not one can use SDO images to 'predict' them.

I've made it very clear since the beginning of this thread what I was interested in. I expected you to either support your claims or admit that you couldn't. Now you've admitted that your claim to already having an objective quantitative method for "predicting" CMEs was untrue. It would be the honest, scientific thing for you to admit that you can't support the claim that dark filament eruptions cause CMEs, too.

All you're interested in is arguing, and frankly I'm just plain bored silly of you and RC. Neither of you has a clue what your talking about.

What we are talking about? We have been asking you to support your claims. We have been pointing out the unequivocal fact that you haven't supported them, don't seem able to support them, and certainly aren't qualified to support them. We do know what we're talking about. You haven't offered any arguments that would give us any reason to see it otherwise.

RC got all upset over the term "dark" filament when in fact that is the 'property' that makes them 'unique' and interesting and makes it possible to track them in the first place.

"Dark" wasn't a number when Reality Check asked you to quantify your description, and "dark" still isn't a number. Your insistence that "dark" has some objective relevance only goes to show you've been guessing all along and that no quantitative objective method for "predicting" CMEs has existed. Your argument has failed and falls deeper into failure with every attempt you make to rescue it.

You demonstrated your ignorance the moment we started to talk about mass flow and the direction of mass flow (that's actually nothing new by the way, that's par for the course with you).

I again note your persistent incivility and desire to make personal attacks rather than address relevant issues or provide valid arguments to support your claims.

 

[Michael Mozina]

You're speaking in future tense about developing a method here.

No, I am speaking future tense about developing a detailed *DOCUMENTATION TRAIL*. Those are entirely different issues.

 

[Michael Mozina]

I've made it very clear since the beginning of this thread what I was interested in. I expected you to either support your claims or admit that you couldn't.

And I supported my claims by correctly "predicting" a number of CME's and flares.

Now you've admitted that your claim to already having an objective quantitative method for "predicting" CMEs was untrue.

No, I have not "admitted" anything of the sort! Do you even know how to tell the truth? What I have said is that am *NOT* guessing, there are "cause/effect" relationships at work, and there's nothing left to chance. What I have said is that I have no desire to go through that quantification process with you on this forum since you and RC are not even REMOTELY interested in an honest discussion on this topic. RC can't even wrap his head around the *FACT* that the filaments in question are "dark', and you have no clue about the mass flow relationships. What would be the point?

Even when I asked you two to commit to the first few things one *MUST* know before a categorization process can even begin, RC's stubbornly dragging his feet claiming that all filaments erupt "up"! You won't commit to *ANYTHING AT ALL*. You two are utterly clueless. All you want to do is argue, argue and argue some more. I'm bored of you two. That's what I said. Did you hear it this time?

 

[Reality Check]

No RC, I know what I saw and read (for hours) on their website.

No MM, I know what they have recorded on their web site and how you managed to fool yourself into the delusions that

1. The McIntosh classification stayed the same.
2. The probability changed from 0 to 70.

You just looked for hours at the web site and did not know what yoiu were looking at !

Originally Posted by Reality Check
This is what it contains:

5 Nov: 11121 Axx 2(99) 0(70) 0(5)
6 Nov: 11121 Cro 9(99) 1(85) 0(20)
7 Nov: 11121 Eai 43(99) 19(85) 1(20)
Note the changing McIntosh classification.

So you are mistaken or lying about the McIntosh classification being the same. It is evaluated and can change whenever SolarMonitor make their daily prediction, i.e. once a day.
​

At no point does a prediction go from 0% to 70%

So you are mistaken or lying about the change in the prediction.
​

I go for mistaken: Look at 5 Nov. See the "0(70)", Micheal Mozina?

5 Nov: 11121 Axx 2(99) 0(70) 0(5) : 0 1 0
6 Nov: 11121 Cro 9(99) 1(85) 0(20) : 2 1 0
7 Nov: 11121 Eai 43(99) 19(85) 1(20) : 1 1 0
8 Nov: 11121 Esi 49(90) 5(50) 0(5) : 0 0 0
9 Nov: 11121 Ero 0(90) 0(50) 0(5) : 0 0 0
​

I added the events for the day at the end.
Oh look - the McIntosh classification changed again!

It's not perfect of course, but it's not bad. Unfortunately they missed the first two M class flares entirely.

ETA: Fixed day to days
Wrong: the first 2 M flares happened on days that there was a prediction of 70% and 85% for at least 1 M flare and at least 1 M flare happened.

ETA2:
My statistics is really rusty. I suspect that to really evaluate the effectiveness of the NOAA predictions that you need to go the whole null hypothesis + chi-squared route.
But here is a very probably wrong technique: Add up the probabilities for each day for at least one flare in the class and treat them as a probability of at least N flares during the period. That gives:

• C: 3 flares for 4.77 expected. 63% accuracy?
• M: 3 flares for 3.40 expected. 90% accuracy?
• X: 0 flares for 0.55 expected. 0% accuracy?

 

[Reality Check]

This whole process has taught me what a true waste of time you and RC have turned out to be. Neither of you is actually interested in CME or flare prediction. You aren't interested in whether or not one can use SDO images to 'predict' them. All you're interested in is arguing, and frankly I'm just plain bored silly of you and RC. Neither of you has a clue what your talking about. RC got all upset over the term "dark" filament when in fact that is the 'property' that makes them 'unique' and interesting and makes it possible to track them in the first place. You demonstrated your ignorance the moment we started to talk about mass flow and the direction of mass flow (that's actually nothing new by the way, that's par for the course with you).

How stupid MM:

1. I am interested in CME and flare prediction now that your thread (not the dumb "predictions" in it) has caused me to research the interesting science involved.
2. There is nothing unique about dark filaments being dark. All dark filaments are bright when viewed from the side. A filament is dark and bright in different wavelengths.

ETA:
The questions that you are stiill ignoring abut what makes dark filaments unique in their role of prediction:
If we see a CME in a LASCO image can we tell what color the filament that erupted was?
IOW: What are the physical propeties of a CME associated with a dark filament that make the CME different from the CME associated with any other kind of filament.
(22 October 2010)
Please give your citations to the darkness of filaments being "relevant" (significant?) in terms of "mass flow prediction"
(22 October 2010)
What is your evidence for a difference in dark/bright filament eruption distribution?
(22 October 2010)


As for "demonstrating ignorance": You have demonstrated your ignorance of the subject from the start of the thread by not being able to cite any literature on the subject until this week.

 

[Reality Check]

More information on the NOAA flare prediction method form a paper on another technique being developed in 2005:
A Statistical Solar Flare Forecast Method

Solar flare prediction remains in its infancy, because of a lack of detailed understanding of the physical processes underlying flares [e.g., Priest and Forbes, 2002]. Existing prediction methods are probabilistic. One popular approach relies on the McIntosh optical classification of sunspots, which divides sunspot groups into 60 classes based on three parameters [McIntosh, 1990; Bornmann and Shaw, 1994]. The historical rate of flaring for a given classification provides an initial estimate for the expected flaring rate of flaring of an observed sunspot group. This approach is the basis for predictions published by the Space Environment Center of the US National Oceanic and Atmospheric Administration (NOAA)2 as well as NASA [Gallagher, Moon and Wang, 2002]3 and IPS. NOAA/SEC uses an ‘expert system’ developed by McIntosh [1990], and adopted in 1987. The associated code begins with the McIntosh classification but also incorporates additional information, including dynamical properties of spot growth, rotation and shear, magnetic topology inferred from sunspot structure, magnetic classification, and previous (large) flare activity. The method involves more than 500 decision rules including ‘rules of thumb’ provided by human experts.

 

[Michael Mozina]

No MM, I know what they have recorded on their web site

Ya, now.

and how you managed to fool yourself into the delusions that

1. The McIntosh classification stayed the same.
2. The probability changed from 0 to 70.

Nobody claimed that the classification system changed. That seems to be your own personal delusion I'm afraid.

The probability changed *AFTER THE FACT*, after two of the M class flares had already occurred. I know exactly what I saw RC.

You are also blatantly ignoring the failures (like yesterday) of your own favorite "guessing" method. Let's review yesterday shall we?

11121 Esi 49(90) 5(50) 0(5)
11122 Cao 16(15) 2(1) 0(1)

They were 90 percent wrong on 11121 on C class, and a whopping *50* percent wrong on the M class flares. WRONG WRONG, *GUESSING* wrong! Not so much as a single *TIME* associated with it any closer than an entire day. It's a "guess". It was wrong. They were also 15 percent wrong on 11122. That's 105 percent wrong for the day. Talk about a wild 'guess" and covering all the bases.

Let's see how today finishes out RC. Like I said, they 'could' get lucky due to that second region forming next to 11121. That hasn't happened yet. I am frankly pessimistic about it growing large enough in the next few hours to trigger anything substantial. It's more likely IMO that the smaller region next to 11121 will flare into 11121 or visa versa than 11121 to actually trigger anything substantial by itself over the next few hours.

So what if it's "quantified" as being 90% wrong plus another 15% wrong? It was still "wrong". It's also a pure "guess" in terms of the actual time of event to within say an hour or two, or the "cause" of the actual event in question. There's no mass flow predictions. It's no less of a glorified mathematical guess than any "guess" based on careful observation and a clear understanding of the cause/effect relationships going on.

You just looked for hours at the web site and did not know what yoiu were looking at !

I know exactly what I was looking at. Unlike you last week wasn't the first time I'd visited the site.

No, I noticed that the probability jumped 70 percent from about 7:30 my time - 09:30 my time when it was 3 percent to somewhere around 10:00 ish, when the probability jumped to 99 percent (from three percent). It went for 0 to 70 percent on the M class flare category as well.

The rest of your statistics are all predicated on our debate over the 5th. Take that one date that seems to be in debate out the statistics and factor in today in say three hours or so and then tell me how your methods of quantified "guessing" are doing, ok?

I can't wait to see if they raise or lower the odds tomorrow because frankly that "could be" (looks to be) a fun day ahead.

 

[Michael Mozina]

How stupid MM:
[*]I am interested in CME and flare prediction now that your thread (not the dumb "predictions" in it) has caused me to research the interesting science involved.

Did you figure out that dark filaments are "dark" yet?

There is nothing unique about dark filaments being dark.

Bzzt.

All dark filaments are bright when viewed from the side.

Compared to WHAT? Space? Compared to the background of the sun in various wavelengths (all the ones I am interested in) they are 'dark', and there is a physical reason they are dark that is also relevant.

A filament is dark and bright in different wavelengths.

A dark filament in H Alpha is also dark in 335A and 211A and 193, etc compared to the rest of the disk! Come on. You're so into "arguing", you don't even care why it's dark, or why it erupts or anything of the sort. I should not have to explain this kind of thing to you "on the fly", as you learn things for the first time. If you're going to argue, at least do so intelligently. There must be a PHYSICAL CAUSE, a PHYSICAL REASON that the filament is 'dark' compared to the rest of the sun in various wavelengths. Aren't you even curious what that "cause" might be?

You evidently haven't learned much of anything. I'm still waiting for you to logically explain what *POSSIBLE* method they might have used to isolate those 'disappearing filaments" with software other than by looking at the pixel intensities of the images?

As for "demonstrating ignorance": You have demonstrated your ignorance of the subject from the start of the thread by not being able to cite any literature on the subject until this week.

That's just ridiculous. I didn't come here to "quote your scripture" for you, I started this thread to *DISCUSS* (not necessarily argue) solar flares and solar CME's. If you wish to quote your scriptures, that's fine by me, but I'm more interested in the physics, the cause effect relationships and the actual processes that make it possible for me to "predict" both a time and directional component of flares and CME's before they become visible in LASCO. You're so fixated on the 'dark" thing that you can't see the forest for all the trees. The filament eruption *PRODUCES* the mass flow you see in LASCO at least *SOME* of the time, maybe *MORE* of the time than you (or I) even realize.

You and GM aren't even willing to cop to the fact that filament eruptions are the "CAUSE" of flares even after it's been shown to you in many satellite images now, including SOHO and SDO images. What's it going to take?

It's one thing to go through a "peer review" process where those that read your paper know a little history about the topic in question and study it a bit *BEFORE* crusading against the idea. It's another thing entirely to sit here in a public forum trying to get you to even learn the basics (like it's relevant that they are "dark" compared to the rest of the disk). I don't have time for this sort of willful stupidity.

 

[GeeMack]

Did you figure out that dark filaments are "dark" yet?

"Dark" isn't a number.

 

[GeeMack]

You and GM aren't even willing to cop to the fact that filament eruptions are the "CAUSE" of flares even after it's been shown to you in many satellite images now, including SOHO and SDO images. What's it going to take?

You do not possess the qualifications to understand what you're seeing in any solar imagery. This has been demonstrated repeatedly to the satisfaction of everyone here except possibly you. Any opinion you might have regarding anything you see in any satellite imagery will be dismissed as unsupported and unqualified, or in other words, a guess.

 

[Michael Mozina]

No MM, I know what they have recorded on their web site and how you managed to fool yourself into the delusions that

1. The McIntosh classification stayed the same.
2. The probability changed from 0 to 70.

You just looked for hours at the web site and did not know what yoiu were looking at !

5 Nov: 11121 Axx 2(99) 0(70) 0(5) : 0 1 0
6 Nov: 11121 Cro 9(99) 1(85) 0(20) : 2 1 0
7 Nov: 11121 Eai 43(99) 19(85) 1(20) : 1 1 0
8 Nov: 11121 Esi 49(90) 5(50) 0(5) : 0 0 0
9 Nov: 11121 Ero 0(90) 0(50) 0(5) : 0 0 0
​

I added the events for the day at the end.
Oh look - the McIntosh classification changed again!


ETA: Fixed day to days
Wrong: the first 2 M flares happened on days that there was a prediction of 70% and 85% for at least 1 M flare and at least 1 M flare happened.

ETA2:
My statistics is really rusty. I suspect that to really evaluate the effectiveness of the NOAA predictions that you need to go the whole null hypothesis + chi-squared route.
But here is a very probably wrong technique: Add up the probabilities for each day for at least one flare in the class and treat them as a probability of at least N flares during the period. That gives:

• C: 3 flares for 4.77 expected. 63% accuracy?
• M: 3 flares for 3.40 expected. 90% accuracy?
• X: 0 flares for 0.55 expected. 0% accuracy?

http://www.solarmonitor.org/forecast.php?date=20101109&indexnum=1

11121 Ero 0(40) 0(10) 0(0)
11122 Dro 17(5) 5(1) 0(0)

Now tell me again that it didn't change *AFTER* the fact? Unless you can't copy and paste correctly *AND* my eyes deceive me, it just happened again.

Now how much credibility is there in that system anyway if everything gets *POSTDICTED* to fit *AFTER* the fact?

 

[Michael Mozina]

Ya know,.........

I'm still really curious to hear how those those percentages go up and down without any change in the classification category. It's like they simply make this up as they go and then they modify it to fit if they blow it the first time.

I thought that the classification was related to the frequency of expected flares? How can the percentages just change up and down like that without changing the category of the sunspot group? It was classified Ero all day long, and yet the numbers changed by 50 percent! That doesn't make any sense to me at all.

 

[Reality Check]

Moot since MM finally produced evidence to back up his claim.
With my help of course !