First, let's look at the temperature data we have. One set of data comes from NASA [source] and the other is provided by Met Office Hadley Centre for Climate Change in the UK.[source]
The data don't agree exactly because (1) the NASA data shows the deviation from the 1951-1980 average and the Hadley Centre data shows the deviation from the 1961-1990 average and (2) the calculations were done independently so small differences are expected. We should bear in mind that the older data comes from spottier readings and is less reliable. The Hadley Centre data is shown both raw and smoothed. Now we'll look at the different factors that affect global average temperature, comparing them with the smoothed data.
Sunspots receive plenty of mention in the popular literature and we have more data to look at.[source]
This is promising. Notice that the sunspots are lower in number, almost zero, in the period 1650-1700. There is anecdotal evidence that Europe and China were cooler then.[source] There also is anecdotal evidence of the same thing happening in the early 19th Century[source], although the Tambora volcano could have contributed.
Looking more closely, we see that the low number of sunspots around 1900 fits the lower temperature then, and temperature and sunspot-count both rise thereafter. There was more activity around 1960 that shows up as a temperature bump, and also around 1980-1990, that fits with a slight, stretched-out bump. So it seems clear that sunspot activity affects global temperature. Or, possibly, sunspots affect the irradiance and it's the combination that affects global average temperature. A suggestion under review is that solar activity diminishes cloud formation by influencing the intensity of cosmic rays, as shown in this figure:[source]
On the other hand, the temperature bumps are small compared with the upward temperature trend since 1900. Furthermore, if solar activity was the main driving force, then average temperature should drop after 1990 but instead it keeps going up. That means something else has become a stronger driving force since 1900.
Solar irradiance is the intensity of solar energy striking the earth and its atmosphere in watts/sq meter. It seems to follow sunspot activity, which seems reasonable. But it only matches the temperature changes about as well. [source]
The gas emissions from natural vegetation are an important part of the atmosphere's loading, but the amount of land devoted to it hasn't increased. It's possible that emissions have risen as a result of global warming. Either way, natural vegetation can't be blamed for the temperature rise since 1900.
Volcanoes emit gases, too. We can do a quick calculation that shows volcanoes could never affect the atmosphere's CO2 concentration.
Volcanoes also emit particulates and aerosols, which reflect heat away from the earth and cause more clouds to form, causing further cooling. Data from the Mauna Loa Observatory shows the effects of volcanos since 1958.[source]
We can see that volcanoes reduced solar transmission in 1982 and 1991, but they don't affect the global-average temperature rise by much. The conclusion is that volcanoes don't affect global warming either way.
One way the Earth's core could heat the oceans is by undersea volcanoes. We can do a quick calculation that it would take around a half-million undersea volcanoes equal in size to the one at Mount Saint Helens in 1980 every year to account for the warming the oceans have seen since 1955. Even if the calculations are off by a factor of ten, it would take around five thousand such volcanoes every year just to account for ten per cent of the warming. And, there would have to have been no volcanoes before 1910. So undersea volcanoes aren't a major factor.
Another possibility is the extrusion of magma into the oceans at the edges of separating tectonic plates. But the USGS has found that the rate tectonic plates have been moving hasn't changed in the last thirty years from what it's always been.[source] So magma doesn't explain the recent warmup.
As is the case for natural particulates and aerosols, artificial particulates and aerosols have a cooling effect by reflecting sunlight and by causing clouds to form. The temperature graph shows a sharp drop around 1940 until almost 1950, then a slow rise until 1980 or so, and after that a sharp rise. That fits with our expectations: industrial production increased radically during the war. Virtually no attention was paid to the resulting pollution. The postwar period experienced some relaxation in both production and pollution. About 1970, serious efforts were started to control particulate emissions from fossil-burning power plants, and the temperature graph clearly shows that global warming accelerated.
There are a lot of these that can be important: carbon dioxide, methane, and nitrous oxide are the most dominant. We need to consider their emission rates in order to compare their relative importance in the changing of the global average temperature.
The US Department of Energy has estimated their yearly emission rates and ranks them this way (2005 data [source]). Global data comes from IPCC's report for 2001[source]. All the rates are in million metric tons per year. These numbers are calculated, but show more precision than they should. Nonetheless, they show relative magnitudes.
|US||US (CO2 Equiv)||World||World (CO2 Equiv)|
Clearly, CO2 is the most important artificial greenhouse gas in respect to changing temperature. The present CO2 content of the atmosphere is 3,036,000 MMT, so the emissions amount to almost 1% of what's presently in the atmosphere. The CO2 concentration is rising roughly 0.5% per year, so about half is staying in the atmosphere and the other half is going somewhere else, mostly into the ocean. We have some measured CO2 concentration data taken from ice cores.[source]
This is our smoking gun. The CO2 concentration has risen from less than 300 parts per million all the way up to 383 ppm in 2007. Of all the factors affecting global average temperature, it's the only one that's been increasing since 1980, so it's the only one that can explain the temperature rise during that time.
What is especially troubling is that, before 1850, CO2 concentration has not exceeded 290 ppm in over 400,000 years.[source]
That's not to say that we can ignore the other greenhouse gases, but controlling CO2 emissions is essential to limiting global warming.
The evidence shows that solar activity and aerosols can influence global temperature. Before 1900, when greenhouse-gas concentrations were below 300 ppm, solar activity seems to have been the main driving force. Since then, greenhouse gases have become the main driving force.