As readers of my reviews know, I am skeptical about the theory that civilization is in danger because CO2 emissions are causing catastrophic levels of global warming. Of the books that I have read, the skeptical books are of higher quality than the pro-global warming books. For some time, I have been looking for a solid statement of the science that supports the Al Gore theory. At first, this book looked like it provided the science that Gore does not. John Houghton is a qualified scientist, and he knows this field. The book is filled with cool charts, graphs and photos. It contains a great deal of information. It is grounded in the scientific literature. Despite this, however, this book is not science. It is a "briefing" in the sense that lawyers use the word "briefs." It is a sustained argument for one point of view. It never takes the other side seriously. It is an advocacy piece. Let us be clear what the argument is about. No one disputes that a lot of CO2 is going into the air. No one disputes that CO2 is a greenhouse gas. The questions are: (1) is the globe, in fact, warming; and (2) if so, is this warming caused by CO2? Beneath all of the charts, graphs and photos, Houghton has two pieces of evidence, and two pieces of evidence only, for the Al Gore theory. First, global temperature rose from 1970 to 2000. Second, scientists have constructed computer models which suggest that, if CO2 levels keep rising, the globe will get warmer. That is it; that is ALL the evidence that Houghton has. The rise in temperature from 1970 to 2000 is suggestive, but, in itself, proves little. Temperature fell from 1940 to 1970. Temperature has stayed about the same since 2000. In short, the trend lines are all over the place. There is, it is true, a long-term rise in temperature since the 19th century, but given that the 19th century was the end of a cooling period -- called the "Little Ice Age" -- one would expect temperature to rise since then. As for the computer models, Houghton makes no effort to prove that they work. Instead, he simply asserts that they work. He says, "I am a big deal scientist. I say the models work. Accept it." This is an appeal to authority. This is not an appeal to reason. Here is what an appeal to reason would look like. If someone like Houghton seriously wanted to persuade a non-believer of the accuracy of the computer models, he would give us examples of the models working. For example, take one of the models, as it exists today. Feed into it raw data from the 1980s. See if it is able to predict the weather of the 1990s. The advantage of this approach is that, since we know what happened in the 1990s, such an approach tests whether the model actually works or not. If the model can accurately "predict" the past, then the model has some credibility There are other ways to show that the models work. Houghton uses none of them. I assume that, if he had any arguments like these, he would use them. He is, after all, an advocate. I trust him to find all of the arguments for his position. If he has nothing to show that the models work, except pounding on his chest and telling us to believe them, then I assume there is no evidence that the models work.