A new survey of over 12,000 peer-reviewed climate science papers by our citizen science team at Skeptical Science has found a 97% consensus in the peer-reviewed literature that humans are causing global warming.That's all well and good (and not surprising), and you can examine their methodology if you want to.
I'm not very keen on these kinds of numbers -- they are made for lazy journalists who don't want to examine the complexity of the science, reporters who just want a number that quickly and easily supports their position.
You'd never hear a scientist use such a number, because they know there is a wide spectrum of opinions about the statement "humans are causing global warming." Some find some parts of the science more convincing than other parts. They may see parts that aren't known well at all, and parts that are very well known. They know that some papers are good and some not so good and they aren't all weighted equally, nor are authors. (About physics, Luis Alvarez said, "There is no democracy in physics. We can't say that some second-rate guy has as much right to opinion as Fermi.")
The simple statement doesn't address questions like how much warming? What kind of warming (where)? How much are humans causing? How are they causing it? How well is this knowledge known? How good is the data? What are the consequences?
People using these kinds of numbers aren't so much interested in these kinds of questions or these kinds of debates as they are in avoiding questions and ending debates. They're for activists, not for those who want to really understand what's going on.