Over the last million years there have been ice ages with inter-glacials such as we are experiencing now about every hundred thousand years. The amount of CO2 in the atmosphere has moved in step with the temperature oscillations to a large extent, although this pattern has been broken recently. The historical record shows that temperature generally leads CO2 changes. Therefore CO2 is not the cause but the result of climate change.
However with the industrial society we have seen truly dramatic changes in CO2 levels which may well have taken us into new territory as far as cause and effect are concerned. With CO2 having never been outside the range 180 to 290 ppm (parts per million) for some 800,000 years, in just the last century or so it has raced up to 370 ppm. It seems that this change must be due to human action rather than being part of the natural ice age cycles.
The question then remains as to whether CO2 increase really can cause dramatic temperature changes or whether it only works the other way?
Various climatologists have tried to examine this question from a variety of points of view, from the expected physical effects of CO2 in the atmosphere through to whether temperature changes or CO2 changes are leading indicators of the other. Using some of their data I have assembled the following graph.

The upper line in the mauve colour shows the rate of change of CO2 concentration. I have used the change over a 12 month period and plotted it at the centre of that 12 month period. This means that it represents a 12 month centred moving average of the rate of change.
The lower line shows the temperature anomaly. This is the fluctuation of temperature from its average level. It shows what changes are happening in global temperatures from year to year.
It is clear that both graphs have a cycle of about 3.6 years in them. I have marked a regular 3.6 year cycle with green arrows and the two series both peak at the same time as this cycle. We can also see that individual cycles have unique characters which are common between the two series (CO2 and temperature), with for example a much higher sharper spike about 2/3 of the way through, so that we can safely say that the fluctuations are causally connected.
Incidentally, this period is rather close to 1/3 of the Sunspot cycle and appears to be a reflection of the SOI or Southern Oscillation Index, or El Nino cycle.
But which one of CO2 change and temperature fluctuation is causing the other?
If the temperature anomaly is causing increased CO2 production, that would make sense in terms of a chemical reaction formula.
If CO2 were to be affecting temperature, then we would expect the amount of CO2 to affect temperature or the rate of temperature increase, rather than the rate of change of CO2 being related. It makes no sense that when CO2 is increasing the temperature is higher, but when CO2 has reached a peak the temperature is already dropping again. So we can safely say that in the short term, CO2 is not the cause of temperature fluctuations.
Even so, this relationship is a bit mysterious. It makes sense that temperature can affect the rate of CO2 production, but in such a case we might expect that when temperature falls back to the same place the level of CO2 would also fall back to the same place. It does not. There is a continuous addition to CO2 in addition to the fluctuations due to temperature.
Perhaps the continuous extra production of CO2 is due to humans and the fluctuations in step with temperature are due to nature. This seems the likely conclusion of the observed relationship.
The Russian astronomer V A Kotov made a study of 160 minute oscillations in the Sun and found that this period of 160 minutes showed up in many places within and outside the solar system. The outer planets lie on the nodes of a 160 minute standing wave centred on the Sun. Planets, asteroids and binary stars more often rotate and orbit in periods that are multiples or fractions of 160 minutes. Kotov devised a general method for statistically checking such commensurabilities.
At one time, the solar oscillation was doubted because its period was so close to 160.0 minutes that it was though to be an artifact of the Earth's rotation as 160 minutes is exactly 1/9 of a day. However further research showed that the period was 160.010 minutes and this was confirmed at different observatories in Crimea, Stanford and Antartica with common phase measurements from all three. There can be no doubt that it is a real oscillation.
Links to Kotov related papers on the 160 minute solar oscillation.
I have mentioned before how difficult it is to get cycles related articles to stay in Wikipedia. There are a group of self-appointed people who go about trying to get cycles material deleted or at least made unintelligible or in some way made to seem inferior. The article on the 160-minute solar cycle is another such case. It states:
History
The birth of helioseismology occurred in 1976 with the publications of papers from Brookes, Isaak and van der Raay [1] and Severny, Kotov and Tsap [2], both of which reported upon the observation of a 160 minute solar oscillation with an amplitude of approximately two metres per second.
It was rapidly realised that this frequency corresponded to one-ninth of a day, and therefore the authenticity of this signal was in some doubt. If a non-sinusoidal oscillation is present in a time-series then power will be seen in a periodogram at not only the frequency of the oscillation, but also harmonics at integer multiples of this frequency. A re-analysis of data obtained over the period of 1974-1976 by Brookes et al. [3] showed that the evidence for a stable, phase-coherent 160 minute oscillation at a constant amplitude was far from conclusive. Although the signal could be detected the amplitude appeared variable and was lower than first reported.
A re-affirmation of the 160 minute signal was obtained by analysis of data from groups in Crimea and Stanford over a long period of time. It was found that the phase showed a steady drift, indicative that the frequency being used in analysis differed slightly from that in the data. This implied that period of 160.01 minutes [4] produced a better fit to the data. Evidence also emerged that multiple sets of observations were phase-coherent. These facts contributed to impressions that the origin of the observed signal was stellar and not terrestrial in origin.
In 1989 as higher-quality multiple-year datasets from a single site became available it was shown by Elsworth et al. that the period of the 160 minute signal was indeed 160.00 minutes, and the amplitude was dependent upon both the length and quality of data obtained in an season, with the signal more prominent at time where atmospheric condition were worse. The group were able to demonstrate that the signal may be simulated by a slightly distorted diurnal sine-wave such as may be obtained by differential atmospheric extinction.[5]
However claims of the presence of a 160 minute period in the Sun were still being presented by Kotov et al. in 1990 [6], and 1991[7], however the mainstream scientific establishment had moved on.
According to the "mainstream" or "establishment" or "prevailing view" there is no 160 minute oscillation in the Sun. Here is some further evidence that shows how silly their view is:
Here is a graph of solar flares and x-rays from July 2000:

This is from the GOES Solar x-ray satellite. The material in green was added by me. It shows regular increases in x-ray flux every 160 minutes, and how solar flares happen sometimes at these peaks. Quite clearly the Sun has a 160 minute oscillation and it is relevant to the production of Solar Flares.
Here is another graph from GOES data for a period in April 2004:

This time the data in red was added by me. It shows clearly that increases in solar x-ray output happen quite often at regular 160 minute intervals. These increases are typically 1000% and may be as much as 8000%. Save and display the picture for a clearer view as it is too big for my blog format.
I do not know why so many scientists do not want to look at cycles information. Cycles is the key to understanding so much about the world around us. The Sun has so many cycles in it and such a rich structure of oscillations and changes ... but that is another article.
Blog of Ray Tomes research on cycles, news reports on cycles, my original research on the Harmonics Theory and discussion of these matters.
There are cycles in everything. There are cycles in the weather, the economy, the sun, wars, geological formations, atomic vibrations, climate, human moods, the motions of the planets, populations of animals, the occurrence of diseases, the prices of commodities and shares and the large scale structure of the universe. None of these are independent of each other.
Research shows that very different disciplines often find the same cycle periods in their data. The inter-relatedness of all things is an idea who's time has come. The study of cycles is an excellent way to understand this because the periods of cycles are as easy to recognise as fingerprints or DNA sequences.
"The universe, believe it or not, is nothing other than a giant musical instrument with a very special but predictable pattern of harmonically related oscillations which determine the structure of everything from galactic clusters to subatomic particles and we are just parts of the various vibration modes."
The single axiom of the Harmonics Theory is that:
The Universe consists of a standing wave which develops harmonically related standing waves and each of these does the same.