Anthropogenic Global Warming is a theory with no scientific merit. For instance: it cannot explain the Medieval Warm Period.
To explain: the climate was approximately 1 degree C warmer in the Medieval Warm Period (c. 950 AD to 1050 AD) than it is today.
If the Medieval Warm Period ocurred due to anthropogenic (man-made) effects, how did this happen at a time when the only heavy industry was Cathedral building? And why (if the MWP was anthropogenic) did the temperature drop and stay lower for the next 800 years?
And if the Medieval Warm Period wasn't caused by anthropogenic (man-made) effects, then what is the basis for stating that our Modern Warm Period is due to anthropogenic forces?
The answer is that climate changes sinusoidally over time. This is due to changes in received solar radiance caused by changes in the Earth's orbital path and angle (the Milankovitch cycle).
There are also decadal cycles in the Sun's magnetosphere that appear to affect low- and high- cloud formation (which affect climate in opposite directions)
These periodic or sinusoidal oscillations in climate can be seen clearly in the following graph which are the collated US Senate Reconstruction of climate as reported in 2005. Note that the temperature was warm in about 1000 AD and is warm again now, and was rather cold in the 1600s.
We have this observed sinusoidal oscillation over ~ 1000 years. There was a warm period from about 50 BC to 50 AD. There was a warm spot in the period 1300 BC to 900 BC. We are in one now, and we had one in ~1000 AD.
So there is a millenial cycle, no doubt about it. But (thinking about it) I don't think we can obtain this observed ~1000 year cycle from the Milankovitch cycles.
Just to expound on this: there are five Milankovitch cycles, each with its own multi-millenial frequency:
Apidal Precession (rotation of the locus of the Earths orbit about the sun) which you wouldn't think would have a large effect.
Shape of Orbit (how close to a circle the orbit is) which should have a large effect on received radiation.
Orbital Inclination (angle of orbit to plane of the ecliptic) - I guess a light effect.
Axial Tilt (North-South angle to earths orbit): large effect.
Axial Precession (North-South angle against sidereal background). I can't distinguish a difference in insolation between this and Axial tilt but maybe there's some factor I'm not seeing.
So there are five Milankovitch cycles, maybe two or three of them are important oscillations. And they vary in frequency from about 100,000 years down to 25,000 years. Adding their effects together would give us an interesting wobbly curve, but the addition of low frequency curves cannot magically give us one with a modulated frequency of ~ a thousand years.
I guess therefore that we have to look for another factor that effects effective insolation (sunshine entering the Earth's energy budget) as a source for the "1000-year itch".
We already have the high frequency 11-year and 121 year temperature cycles. These are discernable from the Senate graphs and are well correlated to sunspot activity, or the lack thereof.
For instance: there was a period of low sunspot activity at the start of the 18th century called the Maunder Minimum: this coincided with Europe's so-called "Little Ice Age".
In passing: there is an interesting piece of history here. 200 years ago, during the "Little Ice Age" the famous astronomer William Herschel was reading Adam Smith's "Wealth of Nations" and noted that grain prices were higher in years with low sunspot activity (meaning that the harvest was poorer in those years). This seemed a crazy correlation at the time, but we can now begin to make sense of it.
Here's the theory: in periods of high Solar magnetic activity ( = loads of sunspots) the Sun's magnetosphere is relatively strong. It acts as a 'shield' against the background sleet of powerful cosmic rays that normally cause cloud-formation at low level -which is a strong cooling effect.
When the sunspot-count is low, more cosmic rays get through, causing low-level clouds and cooling. When the sunspot-count is high less cosmic rays get through. It conceivable (but not remotely proven) that the weakened cosmic rays during high sunspot periods have a role in high atmosphere cloud formation, which would have a strong warming effect.
These magnetic mechanisms have a 11 year and 121 year cycle. It seems perfectly plausible that there is a ~ 1000 year periodicity to the Suns magnetic cycle as well. Or maybe a 1331 year cycle, which would maintain the baffling 11-times-table behaviour of this periodicity :0)
Which is my guess: I believe there is a millennial period to the Sun's magnetosphere, as well as the known 11 and 121 year variations. We don't know enough about the inner workings of the Sun to get a reason for any of these periodicities but a slow thousand-year period seems like a fair hypothesis.
Hope this is helpful.