CME's, active regions and high energy flares
- Thread starter Michael Mozina
- Start date
Michael Mozina
Banned
- Joined
- Feb 10, 2009
- Messages
- 9,361
Oh look, another preeeeety C class flare......
Michael Mozina
Banned
- Joined
- Feb 10, 2009
- Messages
- 9,361
For the record:
11121 Axx 2(1) 0(1) 0(0)
11122 Axx 2(1) 0(1) 0(0)
11123 Dro 17(40) 5(5) 0(0)
11124 Dro 17(10) 5(5) 0(0)
It looks like my predictions are *BETTER THAN* NOAA and NOAA finally gave my new active region friend a number of it's very own. I guess this means it does exist after all, eh RC?
Of course they sort blew that 40 percent figure since it's produced 3 C class flares already, and it could generate more today. Of course we all know that those 'predictions' will be replaced with "observations" at 22:00, so don't be surprised to see that 40 percent figure jump to 99% at 22:00.
11121 Axx 2(1) 0(1) 0(0)
11122 Axx 2(1) 0(1) 0(0)
11123 Dro 17(40) 5(5) 0(0)
11124 Dro 17(10) 5(5) 0(0)
It looks like my predictions are *BETTER THAN* NOAA and NOAA finally gave my new active region friend a number of it's very own. I guess this means it does exist after all, eh RC?
Of course they sort blew that 40 percent figure since it's produced 3 C class flares already, and it could generate more today. Of course we all know that those 'predictions' will be replaced with "observations" at 22:00, so don't be surprised to see that 40 percent figure jump to 99% at 22:00.
Reality Check
Penultimate Amazing
Flares do not only happen in active regions.
It is not an active region according to NOAA.
Reality Check
Penultimate Amazing
Michael Mozina wrongly accuses SolarMonitor of faking data again
That is idiotic, MM, since the numbers are always the NOAA prediction (but for the wrong day)!
What *EXACTLY* what solarmonitor is *DOING* MM:
They are updating the actual daily NOAA prediction to the PREDICTION for that day that is posted at 22:00 on that day! They literally overwrite the "prediction" for the preceding 24 hours with the "prediction" for the next 24 hours.
The NOAA predictions listed on solarmonitor *ARE REAL*. Their timing issue means that the *REAL* prediction for the next day is presented incorrectly on the web page as starting on the previous day.
That is not what you libeled them with. It is libelous to wrongly accuse anyone of changing predictions to fit the data as you did in the following post and made clear in this post.
It is the NOAA numbers that are changing on the SolarMonitor web page which implies that you meant NOAA. You now seem to be saying that you meant SolarMonitor.
In either case you owe someone an apology. You do not need to make one since it is unlikely that SolarMonitor or NOAA are reading this forum.
In either case you need to make it clear that you retract this libel, especially for the sake of your reputation in this forum.
It is that simple MM.
You are accusing solarmonitor of taking observations (from NOAA?), constructing fake predictions and updating their web page with them.
That is idiotic, MM, since the numbers are always the NOAA prediction (but for the wrong day)!
What *EXACTLY* what solarmonitor is *DOING* MM:
They are updating the actual daily NOAA prediction to the PREDICTION for that day that is posted at 22:00 on that day! They literally overwrite the "prediction" for the preceding 24 hours with the "prediction" for the next 24 hours.
More ignorant libel, MM.
The NOAA predictions listed on solarmonitor *ARE REAL*. Their timing issue means that the *REAL* prediction for the next day is presented incorrectly on the web page as starting on the previous day.
Duh - I have been telling you that.
No! I explained to you that you are simply the libeller. This has nothing to do with the timing issue with the web page.
That is not what you libeled them with. It is libelous to wrongly accuse anyone of changing predictions to fit the data as you did in the following post and made clear in this post.
You are accusing someone (the people running "that system") of *AFTER* the fact, changing numbers (*POSTDICTED*) to fit the facts. That is faking the prediction.Originally Posted by Michael Mozina
Now how much credibility is there in that system anyway if everything gets *POSTDICTED* to fit *AFTER* the fact?
It is the NOAA numbers that are changing on the SolarMonitor web page which implies that you meant NOAA. You now seem to be saying that you meant SolarMonitor.
In either case you owe someone an apology. You do not need to make one since it is unlikely that SolarMonitor or NOAA are reading this forum.
In either case you need to make it clear that you retract this libel, especially for the sake of your reputation in this forum.
Reality Check
Penultimate Amazing
Why are you citing the *FAKE* (according to you) data from SolarMonitor
?That is your opionion. They are not because only an idiot would use a sample size of 1 when calculating statistics from a sample - do you know why (it is basic statistics)?
Yes it is now an official NOAA active region - not your fantasyand NOAA finally gave my new active region friend a number of it's very own. I guess this means it does exist after all, eh RC?
.Only an idiot would use a sample size of 1 when calculating statistics from a sample - do you know why (it is basic statistics)?
Of course we all know that you are mistaken about thisOf course we all know that those 'predictions' will be replaced with "observations" at 22:00, so don't be surprised to see that 40 percent figure jump to 99% at 22:00.
.Those NOAA predictions will be replaced with the next day's predictions at 22:00, so don't be surprised to see that 40 percent figure jump to 99% at 22:00.
Reality Check
Penultimate Amazing
Since you seem to be unable to work this out for yourself, maybe this will help (I doubt it though).
SolarMonitor for Nov 10
11121 Eao 31(1) 9(1) 0(0)
11122 Bxo 5(1) 1(1) 0(0)
NOAA Prediction (no observations!) Issued: 2010 Nov 10 2200 UTC 11122 Bxo 5(1) 1(1) 0(0)
1121 1 1 0 0
1122 1 1 0 0
1123 40 5 0 0
1124 10 5 0 0
The NOAA prediction numbers were issued at 22:00 10 Nov for the next 24 hours.1122 1 1 0 0
1123 40 5 0 0
1124 10 5 0 0
Notice that they appear in the SolarMonitor web page for 10 Nov against the 2 active regions that SolarMonitor made predictions for.
Can you see what is happening?
I can:
At the start of the day, SolarMonitor gets a list of active regions from NOAA. They create a list of predictions for those active regions. They create a table listing the active regions and their events.
Every 30 minutes SolarMonitor updates their table of active regions with the current (and yesterdays) NOAA data.
Every 30 minutes SolarMonitor updates the list of predictions with the current NOAA predictions. That works correctly until 22:00 when NOAA replaces the current forecast file with a new forecast file for the next 24 hours.
This means that from 22:00 SolarMonitor, includes the next day's NOAA forecast (not observations) on their web page.
When SolarMonitor creates an archive for their current forecast, it contains the numbers from the next day's NOAA forecast (not observations) and they retain the link to the current NOAA forecast.
Pointing out the timing issue is good.
You though have accused them of faking the data to fit the observations. That accusation is libel since you have no evidence that the web page contains faked data. A retraction is in order.
Perpetual Student
Illuminator
- Joined
- Jul 8, 2008
- Messages
- 4,852
Michael Mozina
Banned
- Joined
- Feb 10, 2009
- Messages
- 9,361
But RC, I do have that evidence and I have pointed it out to you now at least a half dozen times. If the 4th and or 5th contained *ACCURATE* information related to the *REAL* predictions made by NOAA, all of your statements would be true. Since that is not the case, the only thing I can say about the solarmonitor data is that it is in fact "fake" data and does not *NECESSARILY* (sometimes it does) accurately reflect the *REAL* predictions made by NOAA. As long as you continue to ignore that 4th, 5th fiasco, we are at an impasse on this issue.
I understand your concern about me calling it "fake" data that is retrofitted to the observation, but unfortunately that is the case. The presentation of information is both false and misleading as your earlier calculations demonstrated. You *ASSUMED* it was accurate info as would anyone. It's not. That's not acceptable IMO.
The information is in fact presented in a manner that makes it entirely *IMPOSSIBLE* to 'grade' the effectiveness of of NOAA's predictions because none of the actual daily data can be trusted to be an accurate representation of NOAA's original predictions as the missing information on the 4th now demonstrates. Their presentation of data is not accurate and therefore not useful. In fact it is downright confusing. NASA typically does a *MUCH* better job presenting data and I know that they understand how to do it correctly, and to do it accurately. Until I see some evidence that the 4th or 5th accurately represents the predictions made by NOAA, I can only call that data "fake" data that was retrofitted to fit observations at 22:00 every single day.
Michael Mozina
Banned
- Joined
- Feb 10, 2009
- Messages
- 9,361
Why are you citing the *FAKE* (according to you) data from SolarMonitor
It only becomes "fake" at 22:00.
For 22:00 hours (and only 22 hours) it's actually accurate. After that, it's anyone's guess.I know that statistically speaking your beloved quantified methods have been unusually *inaccurate*. IMO that is due to the fact that their methods are based on sunspot activity and often information comes "too late" to be of much use in real "predictions" related to EM flares. That isn't always the case, and to be honest, even I would admit that my EM flare prediction technique is not well quantified. Too much of it is based on "experience" I cannot properly quantify. I would have to admit that for the time being their "method" is still the only game in town, but I'm pretty dismayed at it's usefulness over the past week or so. It seems to be "behind the curve" by almost a day. That's not good for a method that is based on daily predictions.
Yes it is now an official NOAA active region - not your fantasy
It was "active" before it was "official" RC.
Your NOAA folks blew it on that active region. 40 percent? Between 1121, and 1123, they've blown their top now at least 4 times. 1121 was also "downgraded" far more severely than was warranted based on the iron ion images. I'm sure they are congruent with sunspot categorization techniques, but IMO that simply points out the weakness in their method.I've been following along very closely now in real time for a whole week, starting on about the 4th. That why I've noticed the discrepancies between the actual NOAA predictions and the reported predictions on solarmonitor. From the standpoint of statistics, NOAA's methods have been only modestly accurate. Most days they've simply been "inaccurate" to say the least.
IMO their "technique" is likely to be more "accurate" as we approach the active cycle of the sun. In that scenario, I think the odds probably work in their favor. Most of the cycle however, their methods seem to be behind the curve by nearly 18-24 hours.
Of course we all know that you are mistaken about this
Those NOAA predictions will be replaced with the next day's predictions at 22:00, so don't be surprised to see that 40 percent figure jump to 99% at 22:00.
If the 4th contained accurate information about the 5th, I'd let it go. Since it does not, it's clear that there are "problems" with the software at solarmonitor.org.
Michael Mozina
Banned
- Joined
- Feb 10, 2009
- Messages
- 9,361
That is pure denial on your part RC! How many times do I have to point this out to you?
http://www.swpc.noaa.gov/ftpdir/forecasts/daypre/1104daypre.txt
http://www.solarmonitor.org/forecast.php?date=20101104&indexnum=1
NOAA:
1120 10 1 0 0
1121 3 0 0 0
Solarmonitor:
11120 Cao 16(10) 2(1) 0(0)
The information related to 1121 is simply *GONE*! The data is not accurate RC, it is *FAKE* data.
Michael Mozina
Banned
- Joined
- Feb 10, 2009
- Messages
- 9,361
In terms of EM x-ray flare prediction (as opposed to filament eruption prediction), I can't honestly say that I have a well quantified approach. Too much of it is based on "personal experience" that could not be easily automated. I'm still learning a lot from the SDO images in terms of EM flare prediction.
The only advantage I have over NOAA's approach is that I am using iron ion images to 'guess' with, whereas NOAA is using sunspot formation information which is often a much "slower" process to develop than "activity' seen in say 193A or 94A.
Even I admit that NOAA's quantified approach is the only game in town at the moment, but IMO it will eventually be replaced with a prediction method based on iron line images and magnetic field orientations (as is current the case with NOAA's method actually).
IMO NOAA's sunspot information looks to be about 18 hours behind the curve in terms of when activity is first visible in say 335A and when sunspots actually appear in that same region. Often areas are *EXTREMELY* active in the iron ion images, but sunspot activity is only minor at best. That seems to be the scenario where NOAA's "methods" become less accurate and unfortunately that has been the case most of the past week.
At the moment I can "quantify" *WHEN* events are likely to occur, but determining a percentage of likelihood would require a similar approach to the methods NOAA is currently using in terms of classification of active regions in iron line images, and that will take time. Nobody is likely to accept *MY PERSONAL* classification system.
NOAA's methods are in fact based on many years of observed activity, and any replacement method is likely to require that same sort of effort.
Last edited:
Michael Mozina
Banned
- Joined
- Feb 10, 2009
- Messages
- 9,361
Wow!
Wow, 5 C class flares so far today, and no real signs that 1121 and 1123 are going to quiet down anytime soon!
The active region interaction process seems to actually be increasing at the moment rather than decreasing. We could even see LARGER flares than we've seen so far out of those two interacting regions. The amount of energy that concentrated into those two active areas is amazing IMO. Due to the rotation of the sun we're also directly in the line of fire. It's definitely going to be an interesting day on the sun.
Wow, 5 C class flares so far today, and no real signs that 1121 and 1123 are going to quiet down anytime soon!
The active region interaction process seems to actually be increasing at the moment rather than decreasing. We could even see LARGER flares than we've seen so far out of those two interacting regions. The amount of energy that concentrated into those two active areas is amazing IMO. Due to the rotation of the sun we're also directly in the line of fire. It's definitely going to be an interesting day on the sun.
Reality Check
Penultimate Amazing
But MM, you do not have any evidence and I have have pointed it out to you now at least a half dozen times.
You are accusing solarmonitor of faking forecasts to fit observations.
What is happening is a timing issue where the NOAA prediction for the next day is displayed on their web page for the 2 hours from 22:00.
The NOAA data is not faked by solarmonitor to fit the observations. You need be honest and explicitly retract the accusations. Just one clear post MM, e.g.
"I realize now that my statement that SolarMonitor changed their or NOAA's predictions to fit the observations was wrong."
If you want to continue to harp on about the timing issue then call it what it is: a timing issue.
Reality Check
Penultimate Amazing
That is pure ignorance and defamy on your part MM!
We both know that the NOAA predictions have been replaced by the next days predictions at 22:00. Only an idiot would call that *FAKE* data. It is *REAL* data for the next day.
What is happening is a timing issue where the NOAA prediction for the next day is displayed on SolarMonitor web page for the 2 hours from 22:00.
But this has nothing to do with the libel that you are continuing to repeat and not to retract:
You stated several times that solarmonitor are faking the predictions to fit the observations.
How many times do I have to point this out to you?
Perpetual Student
Illuminator
- Joined
- Jul 8, 2008
- Messages
- 4,852
To those who fight the never ending battle against crackpots like Sol88, Mozina, etc.:
Your patience and endurance are admirable. Without your efforts these people would likely influence some of the uninformed audience to believe their nonsense. The more the lay public understands the nature of true science, the better off is our whole society.
Thank you for your contributions to this forum.
Your patience and endurance are admirable. Without your efforts these people would likely influence some of the uninformed audience to believe their nonsense. The more the lay public understands the nature of true science, the better off is our whole society.
Thank you for your contributions to this forum.
Reality Check
Penultimate Amazing
Actually the NOAA forecasts are the standard "game in town". There are several other "games iin town" according to the literature that I have seen. There are even the SolarMonitor predictions!
As you state the NOAA forecasts are aready based on solar images and "magnetic fileld orientations". The "iron line" bit is wrong and irrelevant.
I suspect that you are thinking of the simplistic technique of calculating the McIntosh classification, calculating flare activity statistics for the active area classes and calculating the probability.
That is only the start of the NOAA method.
A Statistical Solar Flare Forecast Method
So NOAA start with a large database of existing quantitative data, extracts a large amount of quantitative data from observations and calculates predictions that include input from people who actually know the subject. The result is daily probabilistic quantitative forecasts of C, M and X x-ray flares.
What do we have from you?
You look at the images. You decide what is happening. You estimate (should be guess but lets be charitable) that there will be some activity.
You cannot explain what you do other then the simple and trivial prediction of activity from highly active regions.
You cannot give us probablilities of C, M or X x-ray flares.
FYI:The NOAA method is expected to be made more accurate soon by the includion of more magnetic topology: NOAA Scientist Finds Clue to Predicting Solar Flares
Reality Check
Penultimate Amazing
NOAA Forecast Verification
Michael Mozina has a rather checkered record of understanding physics and mathematics. Thus his opinion that the NOAA forecasts are inaccurate when compared to his trivial predictions is very suspect. This is especially true since:
Th data shows that NOAA tended to overpredict M and X flares (C flares are not analysed) between 1986 and 2006. IMO this is a good thing - it is better to be on the high side rather than predict lower activity than happens.
Michael Mozina has a rather checkered record of understanding physics and mathematics. Thus his opinion that the NOAA forecasts are inaccurate when compared to his trivial predictions is very suspect. This is especially true since:
- He insists on looking at one daily forecast at a time.
That is quite ignorant. The forecasts give probablilties of at least one flare in a day. It is a probabliity. There will not be X% of a flare on a day. The proper procedure is to compare days with the same probability and see how many days produced at least 1 flare. - He tends to say that the NOAA forecasts fails X% of the time when there is a flare. This betrays an ignorance of the fact that the prediction is for at least one flare.
- He tends to dismiss the prediction when there are multiple flares. This betrays an ignorance of the fact that the prediction is for at least one flare.
Th data shows that NOAA tended to overpredict M and X flares (C flares are not analysed) between 1986 and 2006. IMO this is a good thing - it is better to be on the high side rather than predict lower activity than happens.
Michael Mozina
Banned
- Joined
- Feb 10, 2009
- Messages
- 9,361
No RC, check it out:
Here are NOAA's predictions made on the 10th for the 11th:
http://www.swpc.noaa.gov/ftpdir/forecasts/daypre/1110daypre.txt
# Region Flare Probabilities for 2010 Nov 11
# Region Class C M X P
:Reg_Prob: 2010 Nov 10
1121 1 1 0 0
1122 1 1 0 0
1123 40 5 0 0
1124 10 5 0 0
Here is now what is LEFT on solarmonitor.org:
http://www.solarmonitor.org/forecast.php?date=20101110&indexnum=1
11121 Eao 31(1) 9(1) 0(0)
11122 Bxo 5(1) 1(1) 0(0)
Notice that the data for 123 and 124 is gone?
Last edited: