Articles like this give the impression that climate change doubters and believers talk from two legitimate but different scientific frames of reference.
Roy Spencer is one of the main proponents paid as an academic scientist who claims that the human effect on climate is negligible. Barry Bickmore of Brigham Young University critically analyses Spencer’s position. In a 2008 paper:
Spencer was only able to obtain this result by assuming unrealistic values for various model parameters. If realistic values are used, the effect Spencer described is negligible.
Then in later work Spencer “claims to show with his simple climate model that, not only is climate sensitivity low, but most of the global warming in the 20th century can be explained by a natural cycle called the Pacific Decadal Oscillation.”
I took apart Spencer’s climate model, programmed it into my computer, and showed that, once again, he was only able to come to his conclusions because he was willing to use absurd values for some of his model parameters. Furthermore, he used a bizarre statistical technique that he apparently just made up, because it was capable of giving him nearly any answer he wanted.