1. Scientists ruin their careers by repeatedly attempting to discredit theories they cannot find the evidence to discredit (or proving claims they cannot find proof to claim). If a truly compelling set of studies discrediting global warming trends came out, you can be sure that many publications would be eager to publish it. This is how science works, and how it has worked historically. It's a surprisingly stable environment of meritocracy. Politics intrude, as in any human endeavor, but we'll get to that in:
2. Differences in opinion in the peer review process are as old as science. That is why there are multiple publications and multiple standards for entering into those publications. The whole notion of "scientific consensus" is built around assumptions like schisms such as this. If everyone agreed about the exact criterion for publication we'd only have a few (perhaps regional) scientific publications (and then we'd truly be in trouble).
3. Science is about consensus, but that implicitly accepts dissent. People argue for awhile, a consensus is formed, and then things blow over. Seldom are careers ruined. For an example, see Einstein's extremely controversial claims that eventually became the next major stepping stone for physics. People violently opposed his propositions at first, and now we look at him is one of the greatest minds in human history.
The difference here is that there are active sources of misinformation trying, for whatever non-scientific reasons, to drum up controversy around the global warming issue of anthropogenicity. This issue is still under active debate and research, but unlike many other subjects in a similar status (e.g., quantum physics) there are "deniers" who will not accept any positive evidence against their claim and will resort to any means, social or otherwise, to make their point of view supreme.
Some might argue this is warranted because climate science is making predictions that require us to radically restructure our industry and energy infrastructures to the tune of massive sums of money. To be honest, I think this complaint is shortsighted. Scientific predictions, even those in the limbo state between "well-accepted" and "just a theory" that we are calling "consensus", have always dictated the allocation of resources. Consider the exorbitant cost of the LHC, which is one of the single greatest achievements of human engineering and physics.