And the poster's comments:
When global warming first reached public attention in the 1980's I was not particularly concerned about it for a simple reason. Although temperature had risen since 1900, and CO2 had definitely risen over the same interval, most of the rise in temperature had occurred during the first half of this period while most of the CO2 accumulation had occurred in the second half. What this told me was that CO2 was not the dominant factor affecting global temperature during the first 80 years of the 20th century. Some other factor was operative.
Solar factors are a logical choice. Obviously solar intensity directly affects climate through the Stefan Boltzmann relation, but the magnitude of fluctuations in solar intensity with sunspots are simply too small for this mechanism to be producing the observed effects.
There is a theory that attributes temperature changes to the behavior of the solar magnetosphere: as the magnetosphere expands and contracts, Earth is alternately shielded from and exposed to cosmic rays, and these rays are supposed to be key to cloud formation. Increased cloudiness means increased albedo. Increase albedo means lower temperature. Thus one should expect an inverse correlation between fluctuations in cosmic rays and temperature.
The negative correlation between sunspot number and cosmic rays and the hypothesized negative correlation between cosmic rays and temperature suggests a positive correlation should exist between sunspot number and temperature.
The lowest level of sunspots and temperature occurred in the 1900-1910 period. Both temperature and sunspot number rose for about five decades after these minima. Since around 1950 sunspot number has been roughly flat. Temperature was likewise flat from around 1940 until the mid 1970’s. This correlation suggests that my conclusion in the 1980’s was correct. This cosmic ray effect may have been the factor responsible for rising temperature in the early 20th century (when CO2 rise was small). The lack of temperature rise in the four decades before 1980 might be explained by the lack of rising in solar activity (as indicated by sunspots) over most of this period.
However, in the two decades since the 1980’s there has been substantial warming. Sunspot number has continued to be flat in the decades after 1980 as it was in the decade before. Thus, the solar cosmic ray mechanism isn’t a good candidate for rising temperatures since 1975. CO2 is a good candidate because CO2 levels in recent decades have risen to their highest levels in 800,000 years. Enough of a rise in CO2 has occurred to account for about 0.7 C of warming since the late 19th century. More warming has occurred, but some of this reflects the higher average level of solar activity in recent decades than a century previously.