In this case with the sunspots I would like to know how exactly it can be determined that there will be a 50-year cooling.
By crackpot science only.
Thanks for your answers. What I find interesting is how quickly global warming deniers are to completely pooh-pooh any models which seek to predict the climate unless the ideas agree with theirs, in which case very tiny data-points can with 100 per cent certainty determine the climate for the next 50 years!
Indeed. As others have pointed out:
1. Solar scientists cannot consistently give accurate predictions of the next solar cycle never mind 50 years ahead. Any such long term prediction would be exceedingly unreliable.
2. Even if Sun was to go into a Maunder Minimum mode it would not stop the global warming caused by greenhouse gases, not even if we allow by assumption a strong effect for it.
(Sidenote: Additionally we don't know if the Maunder Minimum in 1645-1715 sunspot records actually is the explanation for the co-incident cooling of the period. On it's surface it may sound like a reasonable assumption to make but it's lacking in details and I'm not convinced at all. Solar contribution could as well have been very small. For example, the MM could have been a manifestation of the possible Livingston&Penn effect.)
I'll add to those this very important tidbit:
3. The claim that we have an accurate 400-year record of sunspot numbers is incorrect (as claimed in the thread by batvette, IIRC).
Let me elaborate on that.
How do we know solar activity?
Definition: solar activity is Sun's
magnetic activity.
It's manifested in numerous solar phenomena such as: solar wind, Sun's magnetic field strength, aurorae, solar corona, radio flux, sunspots, cosmogenic isotopes modulated by solar activity etc. Observations of those phenomena give us different
indexes (or proxies) of solar activity.
It's important to note that those indexes capture different phenomena of the magnetic activity and also different physics behind them. It's also very important to note that
none of the indexes of solar activity capture actual sunshine or Total Solar Irradiance (TSI) but may correlate with it.
Here's an analysis how different indexes of solar activity are linked with TSI (=composites in red):
Can the Total Solar Irradiance be reconstructed from solar activity proxies?
(okay, maybe it's not as impossible as it may sound)
How do we know
past solar activity?
Different indexes of solar activity cover different periods, for example:
Cosmogenic isotopes C-14 and Be-10 as proxies (~10000 years but new techniques will likely extend it)
Observations of aurorae (BCE)
Observations of sunspots (1610- but there are much older random naked eye observations back to BCE)
Observations of solar corona during eclipses (at least 1715-)
Geomagnetic measurements with solar wind signal (18/19th cent.)
Disk photographs (19th cent.)
CaK photographs (1890-)
Cosmic rays (1925-)
10.7 cm Radio flux (1945-)
etc.
Of those indexes the sunspot record gets often used as a measure of past solar activity because it provides a long series of observations with the Relative Sunspot Number (aka R or SSN). Sunspots have been observed more or less systematically since 1610, though there are also some random older observations. The problem with historical sunspot records is that the observations vary from one observer to another, and the further back in time we go the less there's archived observations available.
The most used sunspot series are:
1. The International Sunspot Number (ISN, Ri, Rz; starts year 1700) kept by
SIDC.
2. The Group Sunspot Number (GSN, Rg; starts year 1610) compiled by
Hoyt&Schatten in 1998.
They get used a lot for example when trying and predicting upcoming solar cycle (esp. statistical methods), when inferring past solar activity, when calibrating C-14 and Be-10 paleorecord to more recent levels of solar activity, and when making estimates of past TSI.
And now, finally, on to the point. An important heads up about the historical sunspot series:
The most used historical sunspot series are not accurate. They have some serious errors in them.
1. The International Sunspot Number has at least one error of about 20%:
All numbers after 1945 are about 20% too high relative to the preceeding ones in the series. The reason is an undocumented 1945 change in the spot counting method discovered only recently.
2. The Group Sunspot Number has at least one 50% error:
All numbers before ~1885 in the series are about 50% too low. The reason is a set of inaccuracies when the series was first compiled by Hoyt&Schatten in 1998.
These errors have been discovered fairly recently (2011-), they have been confirmed to exist, their sources have been identified, the researchers have been able to quantify them, and the researches responsible for those series have admitted to them (or those still alive that is).
So what does that mean?
The first implication is that solar activity has actually varied much less than previously thought based on those series. For example this GSN record in the Wikipedia is
wrong, and there likely wasn't any 'Modern Maximum' at all:
Here's a preliminary graph of the corrected sunspot record beginning from 1700:
(
source)
The second implication will be that the arguments for solar contribution to the observed global warming get even weaker than before because there actually has been even less variation than thought (e.g. no 'Modern Maximum').
The third implication is that many papers regarding past solar activity will end up into the trash because they have used inaccurate records which exaggerate solar variation as their calibration backbone. For example, some papers claiming that we have had the most active period of solar activity in thousands of years have used the GSN which has at least 50% error in it. Additionally different estimates of past Total Solar Irradiance (TSI) have had huge variability depending on which sunspot series they have used as their backbone, and the estimates cannot all be correct. Just one example of the wide range marked in blue:
The sunspot series themselves are not yet updated to reflect those errors. That is because it takes time to vet them for further errors. There's an ongoing effort to comb through all available historical observations to reconcile the series, and to form a more accurate historical sunspot series with relevant error bars and tools to use it. Group of prominent solar scientists around the world are working on it, and their results are expected to be published next year as a series of papers in a special topical issue of Solar Physics.
The project home page has more information:
SSN-Workshop
For the layman here's a recent easy-to-read New Scientist article about the issue:
Spot of bother
Relevant paper by Cliver et al. explains it in more detail:
RECALIBRATING THE SUNSPOT NUMBER (SSN): THE SSN WORKSHOPS
Science's interesting, innit?