The articles I read claiming the 97% consensus are really bad science. One (Anderegg) was a questionnaire sent to scientists with three very generic questions.

The articles I read claiming the 97% consensus are really bad science. One (Anderegg) was a questionnaire sent to scientists with three very generic questions. IIRC, a number of scientists didn't reply, and the 97% figure even comes from ignoring those who replied two the first two questions but not the third one. The other (Cook et al. 2013) takes incredibly generic mentions of climate change in papers' abstracts to infer a consensus (stuff like: "co2, a major responsible if climate change, also plays an important role in (something completely different)." Turns out that a lot of times scientists mention climate change in their abstracts as pure click bait.