I don't think S+C have officially released their new "corrected" MSU LT data, but there are files on their ftp site called _5.2 with a date of 30th June. So I'm assuming thats it. I need to write this post in a hurry cos James A challenged me to a blogging race to be first to the data. I've had to read them in, calc the global means, and then the trends myself. So this is an ever-so-slightly dodgy process. But, I'm using the same code as I used for vn5.1, which appeared to work, so likely this does too. It gets about the right answer (trend 0.116 oC/decade over 1979/01 to 2004/12) so thats nice too.
Here are some pics:
The black is the 5.2 series; the red is the 5.1; the blue is RSS. Trend lines are drawn on.
That pic is a bit hard to read: you might find:
a bit easier. That is an 11-month running mean to remove some of the wiggles (except at the end points...). Some interesting features appear in that, which are clearer in a difference plot:
Black is now S+C 5.2-5.1; blue is 5.2-RSS (and the RSS is vn2.1, I think).
So... there is a jump between 5.2/5.1 around 1992/3 (who knows when the satellites changed?); and it goes all wiggly after 2001. Interesting.
Well, there you are. Don't get too excited because I by no means guarantee that this is (a) correct or (b) that the 5.2 files are indeed the correction that S+C promised. Time will tell.