Anomaly up for both satellites.
Immediately after I posted UAH’s September anomaly for the lower troposphere, RSS posted theirs. RSS’s lower is also up for the month: It rose from 0.146 C to 0.211C. The average of the two reporting services rose from 0.070C to 0.186C.
Here’s the plot since Jan 2001. To emphasize the point that the temperature is up, up, up, I circled the September values. 
Figure 1: Lower Troposphere Temperatures Since 2001.
Any one want to take bets on whether NOAA, GISS and HadCrut will be report up for September? Ok.That’s too easy.
After all, even though some visitors like to constantly tell us the the lower troposphere is not the surface (which we all know) — and let us know just very close to the surface — we do know the satellite observations of the lower troposphere and surface observations track each other closely over all.
So, since the TLT measurements are up in September, NOAA, GISS and HadCrut will probably be up.
Here’s a more daring bet!
When reported in September, GISSTemp values for Jan-August read: 14, 25, 60, 42, 40, 28, 50, 39 in C*100. So, the August GISSTemp anomaly is currently reported at “39″. Anyone want to bet what August’s temperature will read when the table is updated later in October? Just for grins, I’ll bet the August reading will rise to 42. My method is: WAG.
Comments
lucia
(Comment#5682)
October 7th, 2008 at 8:44 am
Thanks rafa. I’d even noticed that, fixed the figure and uploaded it. Then, I must have deleted the wrong bit of html and didn’t notice that! Whoops!
Fred Nieuwenhuis
(Comment#5684)
October 7th, 2008 at 10:47 am
Treading carefully here. Per your link, GISTemp looks to be updated. Things that make you go “hmmmm”, August from 39 to 50?????
2008 14 25 62 36 40 29 53 50 49
Year Jan Feb Mar Apr May Jun Jul Aug Sep
Jorge (Comment#5685)
October 7th, 2008 at 10:52 am
Lucia, It may be my new glasses but your circled points on the graph seem to be at a minus anomaly. Is the Y scale out of kilter?
lucia
(Comment#5686)
October 7th, 2008 at 10:56 am
Fred– 2008 14 25 60 42 40 28 50 39 (Posted in Sept.) I’m going to have a monthly bet to see who can predict what Sept.’s GISS temp will be when it’s updated in Nov.
Looks like I underestimated the jump in August temperatures! Looks like last July and August both got warmer during September. April got colder though.
2008 14 25 62 36 40 29 53 50 49 (Posted in Oct.)
Fred Nieuwenhuis
(Comment#5687)
October 7th, 2008 at 11:09 am
No bet, but a guess of Sept’s number in November would be in the 60 +/- 10 range
Alan S. Blue (Comment#5688)
October 7th, 2008 at 11:18 am
“Anomaly up for both satellites” is far too calm and rational. Take 1: “Largest Four Month temperature increase on record! This month marks the fourth straight month of large temperature increases, a rise of nearly 3.5 K when adjusted for seasonal variation.” Take 2: “Longest cold streak on record! Thirteen consecutive months of seasonally adjusted average temperatures being below the long-term average.” Both of those are by eyeball, and don’t go back to 1978. But if I recall correctly is has been a long darn time since we’ve had thirteen consecutive months below the average.
lucia
(Comment#5689)
October 7th, 2008 at 11:46 am
Jorge– I rebaseline to set the average of the period to zero. That makes it easier to compare the how the two changed and how they relate to the average. It also makes it easy to see if the current temperature is below average compared to the period since 2001. However, it does mean I have a shifting baseline– unique to all climate bloggers.
Bob B (Comment#5690)
October 7th, 2008 at 12:04 pm
-1.5C per century just does seem consistent with GCM models and IPCC projections to me.
Jorge (Comment#5691)
October 7th, 2008 at 12:11 pm
Thanks Lucia, I had read your earlier explanation of this procedure but had forgotten. The new glasses are fine. Now I just need a new memory stick. 
AndyW (Comment#5727)
October 9th, 2008 at 12:02 am
Why pick 2001 as a start point?
AndyW (Comment#5728)
October 9th, 2008 at 12:10 am
Ok, found where you describe why you selected 2001. If the satellite data is also available before 2001 then you should use that as well, as long as the data source is consistent then you can happily use the entire period without worrying about having to pick a start point which might put a human spin on it. Regards Andy
lucia
(Comment#5730)
October 9th, 2008 at 5:46 am
AndyW– I can’t use the satellite data with the AR(1)+white noise model until I figure out a few new things. The reason is the current method uses observations from 1914-1944 to estimate the parameters describing “weather noise”. That doesn’t exist for the satellites. Since other people may want to know why pick 2001, I discussed that most recently at the end of this post:
http://rankexploits.com/musing.....esis-test/
Comments Closed: If you would like them re-opened, Contact Lucia
rafa (Comment#5681) October 7th, 2008 at 8:25 am
Just a minor typo, I pressume caption at Figure 1 should read October 7. Thanks for sharing.
best