This graph shows the global temperaure as a red curve, and a number of trends as straight lines of different lengths. These different trends are for various periods: 1850-1865, 1865-1875, 1875-1915, 1915-1920, 1920-1925, 1925-1930, 1930-1935, 1935-1980, 1980-1988, 1988-1995, 1995-june 1997, july 1997-2003, and 2003-2010. And every single one of them is negative. These trends together cover the entire period from 1850 to 2010, that is 160 years, and each of the trends shows a cooling. In many cases, it is a matter of the very strong cooling, and in a few cases a more modest cooling. This means that this entire period of 160 years it has been cooling.
Looking at the red curve, it might appear to the casual observer that in spite of all these negative trends, the global temperature has actually gone up. However, that is but an illusion created by the coincidence that each trend segment starts at a somewhat higher point than the previous trend segment. Hence, the apparent warming does not really exist. It is just a trick of the eye, which disappears when the data is subject to rigorous statistical analysis.
Some warmists may object and say that we should have computed the trend for the entire period. To that I respond: that is a really big cherry you are trying to pick there, Mister! Anyhow, such a long trend has no scientific meaning - it is but a line drawn by a ruler, a trick to hide the decline. And when I say that, I'm being awfully charitable.
So increasing levels of CO2 don’t cause increasing temperatures. How many negative counterexamples of a theory does it take to falsify it?
Parturient montes, nascetur ridiculus mus.