Southern field close to flipping?

[Philalethes]

1 hour ago, Patrick P.A. Geryl said:

Lower 13 month SSN… The previous 6 cycles were in a 5.9 percent range🤔

No other theory was even close to these calculations… The Astrophysical Journal almost published it… Will try again when they see it is right…

(PDF) Calculating the Exact Strength of Solar Cycle 25 using 365 Days Smoothing

https://www.researchgate.net/publication/356556167_Calculating_the_Exact_Strength_of_Solar_Cycle_25_using_365_Days_Smoothing

Could you please tell us why you're intentionally avoiding addressing what we're pointing out, and also avoiding explaining yourself in clear language?

Also, I already did go through that writing of yours. What stood out to me as rather shocking there is this:

Quote

We tried dozens of formulas and found one that has a fault off less than 5.9 percent.

That sounds an awful lot like cherry-picking to me, and sounds like it would inevitably lead to overfitting. That's when your model is tailored to get within a certain error bound of previous data, without actually having much predictive value to speak of, if any at all. And that's of course on top of the small sample size of cycles.

I still see nothing here that gives credence to your prediction or your claim that we're currently at maximum (and you didn't even bother to address what we mentioned earlier about the SSN and F10.7 maxima being at the second peak of SC23 when you tried to cherry-pick that as an example, just as you've glossed over most other concerns people have had).

To me it doesn't sound like you're doing science at all, but rather some self-congratulatory travesty of science. I'd love for you to prove me wrong and start explaining what you mean more clearly and address the above points, but until then I must admit it's hard for me to take you seriously.

Edited February 3, 2023 by Philalethes Bythos

[Aten]

I'm pretty new here, so I don't know if this has been pointed out before, but it appears our friend Patrick has a long history of making batshit crazy claims. If you think his solar cycle predictions are wild, you should check out some of his old predictions for the end of the world in 2012.

Here is a (silly) preview:

 

[Patrick P.A. Geryl]

1 uur terug, Philalethes Bythos zei:

Could you please tell us why you're intentionally avoiding addressing what we're pointing out, and also avoiding explaining yourself in clear language?

Also, I already did go through that writing of yours. What stood out to me as rather shocking there is this:

That sounds an awful lot like cherry-picking to me, and sounds like it would inevitably lead to overfitting. That's when your model is tailored to get within a certain error bound of previous data, without actually having much predictive value to speak of, if any at all. And that's of course on top of the small sample size of cycles.

I still see nothing here that gives credence to your prediction or your claim that we're currently at maximum (and you didn't even bother to address what we mentioned earlier about the SSN and F10.7 maxima being at the second peak of SC23 when you tried to cherry-pick that as an example, just as you've glossed over most other concerns people have had).

To me it doesn't sound like you're doing science at all, but rather some self-congratulatory travesty of science. I'd love for you to prove me wrong and start explaining what you mean more clearly and address the above points, but until then I must admit it's hard for me to take you seriously.

Ok. One by one. The calculations are key. Where can I publish the excel file so that anybody can see it?

if you can make it better then 5 percent, you may be co-author.

The calculations are automatically done. So no cherry picking🤔😅

Math skills needed from a 12 year old🤣

Who can beat NASA?

Maybe other findings possible…

[Jesterface23]

4 hours ago, Patrick P.A. Geryl said:

I forgot another indicator… Said to Jan months ago that when 1K crosses the line with the solar flux… the high is in… There is a 6 month time lag… so this is now🤣

Mind: lower cycle then cycle 24 calculated… So can be used. That are at least 4 indicators…

Common sense. Remember to get the correct smoothed ISN number for right now we need to wait 6 more months. No one can say what will happen until we wait 6 more months, everything else is a full on guess. In reality we will need to wait even longer to get the smoothed maximum ISN.

1 hour ago, Patrick P.A. Geryl said:

Ok. One by one. The calculations are key. Where can I publish the excel file so that anybody can see it?

if you can make it better then 5 percent, you may be co-author.

The calculations are automatically done. So no cherry picking🤔😅

Math skills needed from a 12 year old🤣

Who can beat NASA?

Maybe other findings possible…

Excel documents can be uploaded on Google Drive.

If you have a common goal similar to someone it is fine to try to better them. If you want to conduct research with a purpose, you need to be with them, not against them. (attempting to say that in the right way)

[Philalethes]

3 hours ago, Patrick P.A. Geryl said:

if you can make it better then 5 percent, you may be co-author.

The calculations are automatically done. So no cherry picking🤔😅

You literally said you picked by hand one function that matched the best out of dozens. And nothing to prevent overfitting has been done. You could also use virtually any machine learning algorithm to "automatically" fit the given parameters to the function you want even more perfectly, but this would still be overfitting and have zero predictive value whatsoever. That is why such algorithms typically reserve a significant fraction of the data as a test set (hold-out), and make the model from the rest. In this case the sample size is so low anyway that even this would be of questionable validity.

At this point I'm getting more and more convinced that you have very little understanding of data science, and would urge you to contain your claims to this thread, at least until you're prepared to address the seeming lack of proper use of statistical methods.

I hope I'm wrong, and that you're just way smarter than everyone else (without doing much to prove that, though), but it's getting harder and harder to believe that narrative.

Edited February 3, 2023 by Philalethes Bythos

[Patrick P.A. Geryl]

21 uren geleden, Jesterface23 zei:

Common sense. Remember to get the correct smoothed ISN number for right now we need to wait 6 more months. No one can say what will happen until we wait 6 more months, everything else is a full on guess. In reality we will need to wait even longer to get the smoothed maximum ISN.

Excel documents can be uploaded on Google Drive.

If you have a common goal similar to someone it is fine to try to better them. If you want to conduct research with a purpose, you need to be with them, not against them. (attempting to say that in the right way)

CHALLENGE: PREDICT THE STRENGTH OF CYCLE 25
 

Ok … I uploaded it… anybody can download the file

You need to find a higher accuracy then 5 percent🤔🤩

https://docs.google.com/spreadsheets/d/1ZPmTn8K0PQiZof-C5LmRHrXliteLckSB/edit?usp=sharing&ouid=111616500414055206487&rtpof=true&sd=true

 

here is the theoretical explanation

(PDF) Calculating the Exact Strength of Solar Cycle 25 using 365 Days Smoothing


https://www.researchgate.net/publication/356556167_Calculating_the_Exact_Strength_of_Solar_Cycle_25_using_365_Days_Smoothing
 

Edited February 4, 2023 by Patrick P.A. Geryl

[Philalethes]

4 hours ago, Patrick P.A. Geryl said:

You need to find a higher accuracy then 5 percent

You don't have an accuracy of 5% for the future; what you have is an accuracy of 5% on your cherry-picked overfit model.

[Patrick P.A. Geryl]

1 uur terug, Philalethes Bythos zei:

You don't have an accuracy of 5% for the future; what you have is an accuracy of 5% on your cherry-picked overfit model.

How do you think astronomers do it?

[Philalethes]

12 minutes ago, Patrick P.A. Geryl said:

How do you think astronomers do it?

Not even remotely like that. The fact that you think so is extremely alarming. Astronomers don't use overfit models, but models where test data can easily be predicted by modeling non-test data. That's not even close to what you've done here.

[Patrick P.A. Geryl]

7 minuten geleden, Philalethes Bythos zei:

Not even remotely like that. The fact that you think so is extremely alarming. Astronomers don't use overfit models, but models where test data can easily be predicted by modeling non-test data. That's not even close to what you've done here.

That is the reason they fail… Used the same principle for the prediction of the low and that worked out fine. We will know shortly if we are right.

[Archmonoth]

3 hours ago, Patrick P.A. Geryl said:

How do you think astronomers do it?

Well there are many ways... here is a newer one. 

Astronomers develop new method for predicting explosive solar flares | Astronomy.com

"The team tested their so-called "kappa scheme" on roughly 200 large groups of sunspots from 2008 to 2019, seven of which resulted in large flares. The model successfully predicted six of the seven flares, giving early warning time of several hours to 24 hours. In each of those cases, the flare occurred at or near the site predicted by their model. Of the 198 sunspot groups that didn't produce flares, the model predicted only three false positives."

 

It seems like 3 false positives in roughly 200 groups from 2008-2019 is better than 5%. I know these aren't the same kind of predictions, but astronomers predict all sorts of things. 

 

How many false positives have your predictions produced? 

 

Here is a comparison paper on different prediction methods: A Comparison of Flare Forecasting Methods. II. Benchmarks, Metrics, and Performance Results for Operational Solar Flare Forecasting Systems - IOPscience

"For the presentation here, we select a representative array of dichotomous-based and probability-based metrics, with accompanying graphical evaluation tools, to try and provide as complete a picture as possible."

 

What dichotomous or probability-based metrics are you using? 

I didn't see anything like that from your excel sheet, but perhaps you can point it out?

 

Edited February 4, 2023 by Archmonoth

[Philalethes]

Quote

The simulations play a significant role in the Three Body Problem series and serve to highlight the limitations of human understanding and the dangers of overconfidence in scientific models. The repeated failures of the simulations to accurately predict the behavior of the Three Body system are used to demonstrate the inherent uncertainty and unpredictability of the universe, and the need for caution and humility in our scientific endeavors. The simulations also serve as a metaphor for the ways in which humanity has historically approached the unknown, often with a false sense of certainty and control, only to be proven wrong time and time again. Through these simulations, the series offers a cautionary tale about the limitations of human knowledge and the importance of embracing uncertainty and unpredictability.

[novaxformepal]

Curious what you mean by a 3 body system?  Not as smart as you guys and just catching up....Feels like I'm learning 10 different things at once.

[Archmonoth]

12 minutes ago, novaxformepal said:

Curious what you mean by a 3 body system?  Not as smart as you guys and just catching up....Feels like I'm learning 10 different things at once.

Here is the wiki on the 3-body problem. Three-body problem - Wikipedia 

In 1687 Newton tried but was only able to define formulas for 2 bodies interacting. 

 

The reference it the 3-body problem is a good reminder that the Unknown is perhaps its own boundary, which is to say we can't know the completeness of our knowledge, and any absolute claim of deterministic knowledge is a metaphysical claim.

 

Likewise, being humble or agnostic in the face of the Unknown allows us to explore the Unknown, rather than define it with our crude claims and blind narration.

 

A less philosophical view might be the idea of 3-Value logic: Many-valued logic - Wikipedia 

This means that besides true/false dichotomies, there may be additional systemic knowledge, or values with spectrums of true/false. 

 

Rather than a prediction being absolutely true or false, it can be both, or neither, or something else entirely. Once this thinking is practiced, strict deterministic certainty might seem like low resolution vision (8bit versus 64bit) 

 

This is off topic, but I hope this answers your question. 

Edited February 4, 2023 by Archmonoth

[novaxformepal]

Thanks, I'll go read!

[Philalethes]

1 hour ago, novaxformepal said:

Curious what you mean by a 3 body system?  Not as smart as you guys and just catching up....Feels like I'm learning 10 different things at once.

It's a reference to a book (the title being a reference to the classic problem Archmonoth posted about above), first part of a trilogy. It's not really on the topic of this thread, but that specific reference to it is meant to illustrate the importance of remaining humble when making scientific predictions, rather than loudly proclaiming yourself to be right before sufficiently testing the model against future events.

Edited February 4, 2023 by Philalethes Bythos

[Patrick P.A. Geryl]

Correct me if I am wrong

1. The northern and southern polar field flipped BEFORE the average in cycle 21, 22 and 25

2. The highs of cycle 21 and 22 where reached BEFORE the average flipped.

High cycle 21 1979/9

High cycle 22 1989/6

 

[Jesterface23]

monthly_sn != solar_maximum

[Patrick P.A. Geryl]

9 uren geleden, Jesterface23 zei:

monthly_sn != solar_maximum

Is this a valid explanation?

 

1. The southern polar field flipped in August 2022.

2. The northern polar field flipped in January 2023

3. The northern and southern polar field flipped in cycle 21 & 22 before the average. In both cycles the high was reached before the average flipped. Above that in cycle 21 the high was reached when the southern field flipped and in cycle 22 even before the southern flipped.

Cycle	Northern polar field flip	Southern polar field flip	Average  polar field flip	Solar Max
21	1979/02/25	1979/09/03	1979/12/02	1979/09
22	1989/01/03	1989/08/21	1989/12/19	1989/06

4. In January 2023 the SSN was just slightly lower than the highest month of cycle 24. The 10.7 Solar flux was slightly higher. 

5.  From 1, 2, 3 and 4 we conclude that the high of cycle 25 fell in January 2023.

I made a fault and corrected it. The northern field flipped in January 2023

Edited February 14, 2023 by Patrick P.A. Geryl
Correction flip northern field

[Sam Warfel]

Sounds to me like according to your average from a very small sample size of 2 SCs, solar maximum “ought” to have occurred in mid 2022. Obviously, it didn’t. 

[Patrick P.A. Geryl]

3 uren geleden, Sam Warfel zei:

Sounds to me like according to your average from a very small sample size of 2 SCs, solar maximum “ought” to have occurred in mid 2022. Obviously, it didn’t. 

You need to take the last flip of the northern or southern

Solar Max fell in cycle 21 on the month the southern polar field flipped.

This would be January 2023

Solar Max fell in cycle 22 two months before the southern polar flipped.

This would be November 2022.

No cycle is the same. So January would be a good fit. I have pre knowledge that the average flipped in February

In our published article about the start of solar cycle 25 we found a range between August 2019 and January 2020. Afterwards we narrowed the range.

I made a fault and corrected it. The flip from the northern field was in January…

Edited February 14, 2023 by Patrick P.A. Geryl
Flip was in January

[Jesterface23]

11 hours ago, Patrick P.A. Geryl said:

Is this a valid explanation?

There is the point, there aren't many data points to go off of. You do understand you've been basically saying solar maximum rather than monthly SN maximum?

Aaand scrap this part for now, to be determined from Bean's reply.
If you want to go daily, we officially reached the highest daily SN of this cycle a few days ago.

Edited February 14, 2023 by Jesterface23

[Bean]

Just a point to Jesterface23 , the highest daily Sunspot Number for SC25 (so far ) was recorded on the 19/01/2023 at a number of 206, any number for February is provisional at the moment and will not be a matter of record until SILSO release the monthly bulletin >    https://www.sidc.be/sunspots/bulletins/monthly/monthlybull202301.pdf

[Patrick P.A. Geryl]

3 uren geleden, Jesterface23 zei:

There is the point, there aren't many data points to go off of. You do understand you've been basically saying solar maximum rather than monthly SN maximum?

Aaand scrap this part for now, to be determined from Bean's reply.
If you want to go daily, we officially reached the highest daily SN of this cycle a few days ago.

isn’t Solar max the month with the highest SN?
You have the 13 month smoothed average also. That could be March 2023. The 365 day smoothed average for the 10.7 Solar flux will be a few months later.

[Patrick P.A. Geryl]

Op 4/2/2023 om 14:54, Philalethes Bythos zei:

You don't have an accuracy of 5% for the future; what you have is an accuracy of 5% on your cherry-picked overfit model.

If you open the document, you find an easy formula with an accuracy of 10 percent.

No complicated math needed to find that. Show me a model from astronomers that can give that accuracy for the last 6 cycles.