Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > sci.physics > #507732
| From | R Kym Horsell <kym@kymhorsell.com> |
|---|---|
| Newsgroups | sci.physics |
| Subject | Re: The Maunder Minimum is back! (Maybe. And we probably won?t notice) |
| Date | 2015-07-17 22:38 +0000 |
| Organization | SDF Public Access UNIX System, Est. 1987 |
| Message-ID | <moc062$4qt$1@odin.sdf-eu.org> (permalink) |
| References | (2 earlier) <N5OdnXN5GOdA9jXInZ2dnUU7-RUAAAAA@giganews.com> <8pcm7c-j83.ln1@mail.specsol.com> <cY2dnTCXQOl6bjXInZ2dnUU7-QMAAAAA@giganews.com> <6mtn7c-l99.ln1@mail.specsol.com> <tKadnVBQFt0Q4jTInZ2dnUU7-fednZ2d@giganews.com> |
Sam Wormley <swormley1@gmail.com> wrote: > On 7/17/15 1:21 PM, jimp@specsol.spam.sux.com wrote: >> >> My point exactly, so what is the point of your constant, off topic >> posts full of speculation and FUD, ass hole? >> >> > > The physics (meteorological and climatology branches) community is > doing a lot of research in this area. Noting to be afraid of jimp. > > The Maunder Minimum is back! (Maybe. And we probably won?t notice) >> http://arstechnica.com/science/2015/07/the-maunder-minimum-is-back-maybe-and-we-probably-wont-notice/ > > > A new ice age is not actually coming. But you wouldn't know that > > based on headlines from places like The Telegraph. The "reasoning" > > behind the headlines is that a new study suggests the Sun is headed > > for a severe low in its activity, and the last time that happened > > coincided with a cold period called the Little Ice Age. Ergo, another > > one is on the way. > > > > Pretty much every aspect of this reasoning is wrong. > > > > To understand why, we have to understand sunspots. Sunspots are the > > product of magnetic activity inside the Sun, and they appear as > > darker, cooler areas on the star's surface. You might think that they > > would cause less light to reach Earth, but the areas around the > > sunspot brighten up, more than compensating for the dark areas. As a > > result, there's a simple relationship: the more sunspots, the more > > solar output reaches Earth. > > > > Normally, sunspots come and go with the 11-year long solar cycle?more > > during peak activity, less during the Sun's quieter times. But > > superimposed on that cycle are longer-term trends in activity that we > > don't understand very well. For example, solar activity has been > > trending downward since the 1980s, and it's not entirely clear why. > > But it's still much more active than it was during a roughly 70-year > > period around 1700, when the Sun experienced what's called the > > Maunder Minimum, with very few sunspots visible the entire time. > > > > That period happened to coincide with the Little Ice Age, which saw > > global temperatures dip and glaciers advance, with Europe > > experiencing notably colder weather. The lack of solar activity > > provided a nice explanation for this, so the two were assumed to be > > connected. > > > > Why is this topic back in the news? Blame can be assigned to a press > > release put out by the Royal Astronomical Society. The release > > described a talk that would be taking place at the organization's > > National Astronomy Meeting, focusing on a new model of solar > > activity. The model, based on a more complex version of the Sun's > > internal magnetic dynamo, suggests that there are two cycles going on > > at the same time. When they're in alignment, the solar cycle results > > in dramatic peaks of activity. When they oppose each other, you get > > greatly reduced activity?something like a Maunder Minimum. > > The more I look at this paper, the more I suspect something is off. The authors have used historical WSO data to derrive some principal components of the solar mag field. Then they extracted these components from the last 30 y of data and put them through Neutonian's equation-finding package. Basically, the s/w tries billions (it says in the literature) of possible equations matching up parts of the data, spitting out those that come close to holding true. THEN the trends so found are projected forward. Without field data for the future they "predict" the trend of the next 3 cycles will (by accident?) more or less follow those in the last 3 cycles which was the training data used for the equation-finder. There are a number of possible problems. I have some experience in "normal" statistics, but I'm starting to believe finding confidence intervals for an equation you find by brute force search may not be so easy to pin down. I'm sure the package can figure out it shouldn't come up with the uncountable number of functions like sin2(x)+cos2(x) == sin2(y)+cos2(y) as a "find", but how can it do the even less decidable problem of not coming up with false hits (slow-moving functions; or selecting variables that don't have much of a range in the data; or spitting out formulas like "corr(x,y) approx eq 0") that are "almost true by definition"? And then to put a "prob this is not chance" on the rest? Sounds ripe for over-fitting. The package sounds fine for culling through mountains of numbers looking for "laws" that someone can later go back to the lab and verify. (The literature has a few examples for the package in questions -- most with the worrying note "we've found this relation to be true; it was previously unsuspected; no known mechanism is know why it should hold"). If anyone has run research data through some of the neat neural net learners that abound these days they may have had the same experience. Fitting e.g. sunspot data with a NN can come up with models that look perfect. (I did a little exercise just on the strength of the paper in questions, and I had to look twice to make sure I hadn't accidentally put the training set in as the test set as well). Until you try to apply such models to new data and the predictions are not so warm. (In my trial the training error was near 0; in the test it was nearer 100%). Anyway. All very interesting. -- And like most of your fallacies, you got the premise wrong. The last 15 years CO2 rose and mean global temperature did not. -- Melvin on Mars (aka Varney@CO), 22 May 2012 22:54:02 -0500 Plot of co2, BEST prelim, and Artic ice x -1: <http://www.woodfortrees.org/plot/best/mean:132/from:1940/normalise/plot/esrl-co 2/mean:12/normalise/plot/nsidc-seaice-n/mean:60/normalise/scale:-1>
Back to sci.physics | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
The Maunder Minimum is back! (Maybe. And we probably won’t notice) Sam Wormley <swormley1@gmail.com> - 2015-07-16 15:38 -0500
Re: The Maunder Minimum is back! (Maybe. And we probably won’t notice) jimp@specsol.spam.sux.com - 2015-07-17 04:26 +0000
Re: The Maunder Minimum is back! (Maybe. And we probably won’t notice) Sam Wormley <swormley1@gmail.com> - 2015-07-17 07:24 -0500
Re: The Maunder Minimum is back! (Maybe. And we probably won’t notice) Sam Wormley <swormley1@gmail.com> - 2015-07-17 17:19 -0500
Re: The Maunder Minimum is back! (Maybe. And we probably won?t notice) R Kym Horsell <kym@kymhorsell.com> - 2015-07-17 22:38 +0000
ah, interesting noTthaTguY <abu.kuanysh05@gmail.com> - 2015-07-17 16:07 -0700
Re: The Maunder Minimum is back! (Maybe. And we probably won't notice) Mahipal <mahipal7638@gmail.com> - 2015-07-17 16:35 -0700
Re: The Maunder Minimum is back! (Maybe. And we probably won’t notice) "Robert Clark" <rgregoryclark@gmSPAMBLOACKail.com> - 2015-07-30 11:12 -0400
Re: The Maunder Minimum is back! (Maybe. And we probably won’t notice) jacob navia <jacob@jacob.remcomp.fr> - 2015-08-02 12:05 +0200
Re: The Maunder Minimum is back! (Maybe. And we probably won’t notice) HVAC <Mr.HVAC@gmail.com> - 2015-08-02 06:58 -0400
Re: The Maunder Minimum is back! (Maybe. And we probably won’t notice) dsr@mail.lns.cornell.edu (Daniel S. Riley) - 2015-08-02 12:53 -0400
Re: BURNING forests and the talk goes on jacob navia <jacob@jacob.remcomp.fr> - 2015-08-03 00:12 +0200
Re: BURNING forests and the talk goes on Wally W. <ww84wa@aim.com> - 2015-08-02 19:57 -0400
Re: BURNING forests and the talk goes on jacob navia <jacob@jacob.remcomp.fr> - 2015-08-03 10:40 +0200
Re: BURNING forests and the talk goes on Wally W. <ww84wa@aim.com> - 2015-08-03 08:03 -0400
Re: BURNING forests and the talk goes on jacob navia <jacob@jacob.remcomp.fr> - 2015-08-03 14:37 +0200
Re: BURNING forests and the talk goes on benj <nobody@gmail.com> - 2015-08-03 13:28 -0400
Re: BURNING forests and the talk goes on jacob navia <jacob@jacob.remcomp.fr> - 2015-08-03 14:41 +0200
Re: BURNING forests and the talk goes on Wally W. <ww84wa@aim.com> - 2015-08-03 22:40 -0400
Re: BURNING forests and the talk goes on HVAC <Mr.HVAC@gmail.com> - 2015-08-04 06:48 -0400
Re: BURNING forests and the talk goes on jacob navia <jacob@jacob.remcomp.fr> - 2015-08-03 14:51 +0200
Re: BURNING forests and the talk goes on Fred J. McCall <fjmccall@gmail.com> - 2015-08-03 08:18 -0700
Re: BURNING forests and the talk goes on benj <nobody@gmail.com> - 2015-08-03 13:35 -0400
Re: BURNING forests and the talk goes on jacob navia <jacob@jacob.remcomp.fr> - 2015-08-03 20:37 +0200
Re: BURNING forests and the talk goes on Wally W. <ww84wa@aim.com> - 2015-08-03 22:40 -0400
Re: BURNING forests and the talk goes on benj <nobody@gmail.com> - 2015-08-04 02:51 -0400
Re: BURNING forests and the talk goes on benj <nobody@gmail.com> - 2015-08-04 02:23 -0400
Re: BURNING forests and the talk goes on Snidely <snidely.too@gmail.com> - 2015-08-04 00:01 -0700
Re: BURNING forests and the talk goes on benj <nobody@gmail.com> - 2015-08-04 13:35 -0400
Re: BURNING forests and the talk goes on Snidely <snidely.too@gmail.com> - 2015-08-05 03:43 -0700
Re: BURNING forests and the talk goes on benj <nobody@gmail.com> - 2015-08-07 01:21 -0400
Re: BURNING forests and the talk goes on jacob navia <jacob@jacob.remcomp.fr> - 2015-08-07 12:50 +0200
Re: BURNING forests and the talk goes on jacob navia <jacob@jacob.remcomp.fr> - 2015-08-07 13:23 +0200
Re: BURNING forests and the talk goes on Snidely <snidely.too@gmail.com> - 2015-08-04 00:43 -0700
Re: BURNING forests and the talk goes on Wally W. <ww84wa@aim.com> - 2015-08-04 08:30 -0400
Re: BURNING forests and the talk goes on benj <nobody@gmail.com> - 2015-08-04 14:07 -0400
Re: BURNING forests and the talk goes on Snidely <snidely.too@gmail.com> - 2015-08-05 03:46 -0700
Re: BURNING forests and the talk goes on Wally W. <ww84wa@aim.com> - 2015-08-05 08:51 -0400
Re: BURNING forests and the talk goes on jacob navia <jacob@jacob.remcomp.fr> - 2015-08-05 20:23 +0200
Re: BURNING forests and the talk goes on Wally W. <ww84wa@aim.com> - 2015-08-06 02:04 -0400
Re: BURNING forests and the talk goes on Snidely <snidely.too@gmail.com> - 2015-08-05 23:42 -0700
Re: BURNING forests and the talk goes on Wally W. <ww84wa@aim.com> - 2015-08-08 09:25 -0400
Re: BURNING forests and the talk goes on Snidely <snidely.too@gmail.com> - 2015-08-10 02:22 -0700
Re: BURNING forests and the talk goes on Wally W. <ww84wa@aim.com> - 2015-08-10 09:34 -0400
Re: BURNING forests and the talk goes on benj <nobody@gmail.com> - 2015-08-10 10:37 -0400
Re: BURNING forests and the talk goes on HVAC <Mr.HVAC@gmail.com> - 2015-08-10 13:55 -0400
Re: BURNING forests and the talk goes on Snidely <snidely.too@gmail.com> - 2015-08-12 01:05 -0700
Re: BURNING forests and the talk goes on jacob navia <jacob@jacob.remcomp.fr> - 2015-08-12 11:17 +0200
Re: BURNING forests and the talk goes on Wally W. <ww84wa@aim.com> - 2015-08-12 05:44 -0400
Re: BURNING forests and the talk goes on "kquirici@yahoo.com" <kquirici@yahoo.com> - 2015-08-12 05:19 -0700
Re: BURNING forests and the talk goes on Sam Wormley <swormley1@gmail.com> - 2015-08-12 08:05 -0500
Re: BURNING forests and the talk goes on Wally W. <ww84wa@aim.com> - 2015-08-13 02:34 -0400
Re: BURNING forests and the talk goes on Snidely <snidely.too@gmail.com> - 2015-08-05 22:23 -0700
Re: BURNING forests and the talk goes on Snidely <snidely.too@gmail.com> - 2015-08-04 00:45 -0700
Re: BURNING forests and the talk goes on jacob navia <jacob@jacob.remcomp.fr> - 2015-08-04 09:52 +0200
Re: BURNING forests and the talk goes on Wally W. <ww84wa@aim.com> - 2015-08-04 08:55 -0400
Re: BURNING forests and the talk goes on Alain Fournier <alain245@videotron.ca> - 2015-08-04 19:28 -0400
Re: The Maunder Minimum is back! (Maybe. And we probably won’t notice) "Robert Clark" <rgregoryclark@gmSPAMBLOACKail.com> - 2015-08-08 10:56 -0400
Re: The Maunder Minimum is back! (Maybe. And we probably won?t notice) R Kym Horsell <kym@kymhorsell.com> - 2015-08-08 15:19 +0000
Re: The Maunder Minimum is back! (Maybe. And we probably won?t notice) Sam Wormley <swormley1@gmail.com> - 2015-08-08 12:07 -0500
csiph-web