Solar Cycle Precursor Prediction using Maximum Monthly Sunspot Number

[Drax Spacex]

Observation:  A previous solar cycle with low activity in its last 3 years is predictive of a weak current solar cycle.  A previous solar cycle with higher activity in its last 3 years with a more jagged tail and sharper fall-off with some final months of higher activity is predictive of a strong current solar cycle.

Hypothesis:  The maximum monthly mean sunspot number in the last 3 years of a previous solar cycle is positively correlated to and predictive of the maximum 13-month smoothed sunspot number of the current solar cycle.

Data:
SC, Solar Min Date YYYYMM, SC-1 maximum monthly mean sunspot number from last 3 years of cycle, SC 13-month smoothed sunspot number maximum

18,194404,111,219
19,195404,153,285
20,196410,73,157
21,197603,84,233
22,198609,117,213
23,199605,85,180
24,200812,59,116
25,201912,44,SC25Max

(111,219), (153,285), (73,157), (84,233), (117,213), (85,180), (59,116)

Linear Regression fit of above points:

https://ibb.co/kh0t9rN

y=1.48x+52

where 

x=SC-1 Maximum Monthly Mean for Last 3 Years of Cycle
y=SC Predicted 13-Month Smoothed Maximum

Predicted SC25MAX = 1.48*44+52 = 117

Data Source: WDC-SILSO, Royal Observatory of Belgium, Brussels

Edited February 19, 2023 by Drax Spacex
Added Data Source attribution

[braintic]

Seriously? Why did you stop at cycle 18? Let me cherry pick the previous 6 cycles instead:

[braintic]

Or perhaps we could compare the maximum for the FIRST 25 months of the cycle to its maximum. (There is nothing special about multiples of 12, right?)

Pretty tight fit right? That gives a projection of 142 for cycle 25.

How many different arbitrary data associations do you think we could explore? Especially ones that involve fickle extreme values?

Is it possible you had an idea in advance of what what sort of projection you wanted, then fiddled until you got what you wanted?

[Drax Spacex]

5 hours ago, braintic said:

Is it possible you had an idea in advance of what what sort of projection you wanted, then fiddled until you got what you wanted?

Absolutely!  It's called the scientific method.  Observation led to Hypothesis led to Testing led to Results. Iterate.

5 hours ago, braintic said:

Or perhaps we could compare the maximum for the FIRST 25 months of the cycle to its maximum. (There is nothing special about multiples of 12, right?)

Pretty tight fit right? That gives a projection of 142 for cycle 25.

Yes that works too.  That would be more of an in-situ prediction (using data from the current cycle to make a later prediction within that cycle) rather than a precursor prediction using only data that preceded the current cycle.

 

[braintic]

No that is NOT "the scientific method".

You didn't "test" your hypothesis. You FORCED it.