As science is done in big teams I'd say it would take a lot to cover someones ass. You usually have a professor, some scientists and a heckload of students of all levels, all interested in finding out the truth (more or less I guess).
I'm in climate research (as a Bachelor candidate, granted) and I wouldn't let anything like that slip, and I know I'm not the only one.
Like I have pointed out several times - there is no (known) truth. There are only hypotheses. You can easily cover up bad data, but you can't easily prove that someone covered up that data.
You say it's easy. I don't think so, for several reasons:
- If you collect raw data, everyone in the group knows. You can claim instrument failure, of course, but then you'd have everyone around trying to help.
- From the raw data alone it's hard to say what to dismiss. So you'd have to do the long-term analysis first and then decide what to ditch - harder then because your data will be on backup, and you will probably have talked about things with people - because it's a group.
- You're most probably not the only person using this data. As I said, there are students doing some work for you, other scientists in your group using your data.. and everyone would be happy to find something new.
- Additionally, I cannot think of a single professor who wouldn't love to find that global warming is not true.. but I could have misjudged here, of course.
demallien has already pointed out what we know and what we think, so I won't go into this.
I'm in climate research (as a Bachelor candidate, granted) and I wouldn't let anything like that slip, and I know I'm not the only one.