UAH Global Temperature Update for August, 2011: +0.33 deg. C
By Dr. Roy Spencer
The global average lower tropospheric temperature anomaly for August, 2011 retreated a little, to +0.33 deg. C (click on the image for a LARGE version):
Note that this month I have taken the liberty of adding a 3rd order polynomial fit to the data (courtesy of Excel). This is for entertainment purposes only, and should not be construed as having any predictive value whatsoever.
Here are the stats…we are beginning to see cooling in the tropics from La Nina conditions which are re-emerging there:
YR MON GLOBAL NH SH TROPICS
2011 1 -0.010 -0.055 +0.036 -0.372
2011 2 -0.020 -0.042 +0.002 -0.348
2011 3 -0.101 -0.073 -0.128 -0.342
2011 4 +0.117 +0.195 +0.039 -0.229
2011 5 +0.133 +0.145 +0.121 -0.043
2011 6 +0.315 +0.379 +0.250 +0.233
2011 7 +0.374 +0.344 +0.404 +0.204
2011 8 +0.325 +0.323 +0.327 +0.157
The global sea surface temperatures from AMSR-E through the end of August are shown next. The trend line is, again, for entertainment purposes only:
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
I am indeed entertained.
Thanks Dr Spencer. Another example of your sterling work. Always entertaining, adherent of the scientific method and striving for accuracy.
Busy day, huh?
As long as we are just playing around with the numbers here, what about a Fast-Fourier Transform (FFT)? Nothing better for resolving periodic influences…
Have we reached “peak UHA?”
Looks like a slope of about +0.13C/decade.
I wonder if the small increase in the trace gas has anything to do with it…
PhilM says:
September 2, 2011 at 12:31 pm
=================================
Since 1979 and 2002? yeah, that’s it….
…what is funny is the zero “normal” line…..move it up and we’re below “normal”….move it down and we’re above “normal”…….LOL
That’s 0.335°C since January this year!!!!!
Could be the classic “head and shoulders” shape well known to stock market chartists. Only time will tell of course…
Really Dr. Spencer … whether it’s the 13-month average or a 3rd order polynomial, anyone beyond the 3rd grade ought to know that all that those lines do is indicate in simpler terms, what has happened, not what will happen. Okay, maybe the 4th grade.
Instead of adding a polynomial fit, why don’t you (and everyone else, for that matter) try some Fourier series fits (instead of Taylor series fits)?
And it’s about this time every month that i provide a link to my post about the preliminary Reynolds OI.v2 Sea Surface Temperature data (Global and NINO3.4), this time for August 2011, along with the weekly values for the same datasets:
http://bobtisdale.wordpress.com/2011/08/30/preliminary-august-2011-sst-anomaly-update/
Some Guy says:
September 2, 2011 at 1:25 pm
“Instead of adding a polynomial fit, why don’t you (and everyone else, for that matter) try some Fourier series fits (instead of Taylor series fits)?”
To avoid frequency reflections (Gibbs phenomenon) you would have to use a window function like a Hamming window, for instance; this would necessarily reduce the weight of current data to near zero, so you won’t see the influence of the last few months/years very much. A trend line or especially a moving average is much better to see the development up to now. (You stay in the time domain instead of switching to the frequency domain)
DirkH says:
September 2, 2011 at 2:45 pm
“To avoid frequency reflections (Gibbs phenomenon) ”
Correction: The Gibbs phenomenon describes not frequency reflexions, but “waviness” of the transform. Sorry.
Everyone give a big hello to DeSmog blog’s most prolific commentor, PhilM. Phil is an Aussie and a huge fan of Julia Gillard. If you engage him in conversation you will have no doubt that he considers himself the smartest man in the room. Hey, Phill! The trend line for any meaningful “climate change” legislation here in the states is still 0.0 per decade.
Note that this month I have taken the liberty of adding a 3rd order polynomial fit to the data (courtesy of Excel). This is for entertainment purposes only, and should not be construed as having any predictive value whatsoever.
Thank you for the disclaimer – I go apoplectic when someone tries to project a polynomial fit. I don’t even like them near the end points. Whether or not my apoplexy has any entertainment purpose is unclear.
Gee, a curving fit. You’d never know that, as Ric Werme states in this train, that trying to find a non-linear fit to something that is known in the very long term and the very short term to have sinusoidal-like cycles is a foolish thing and leads to foolish expectations that, over time, rises are followed by falls that are followed by rises, sometimes bigger than others.
Please tell me why we can’t say that there are somewhat curivlinear cycles going on now that went on in the past and probably will do so in the near future. Is it an anal-rhetorical problem with mathematical rules of certainty because we don’t have enough cycles in the box for the statistician to wrap a program around?
Correlations are the first stage of scientific discovery, even if it is only that the causation is two steps back. When my right fingers go numb, I know that it is not a problem with my fingers, but wih the scar damage in my neck vertebrae, but I sure as hell use that correlation to know when it is time to go to chiro.
If the pattern is there and it is useful, use it. Work out the reasons later (like CERN CLOUD).
Mike Mangan:
Hey Mike!
You may be thinking of a different PhilM!
Thank you for the math added just for grins; it confirms my thoughts based on observing your data about once a week. Keep up the good work. TNX
I think I now remember why I stopped putting any trend line on these plots. 🙂
“Ric Werme says:
Note that this month I have taken the liberty of adding a 3rd order polynomial fit to the data (courtesy of Excel). This is for entertainment purposes only, and should not be construed as having any predictive value whatsoever.
Thank you for the disclaimer – I go apoplectic when someone tries to project a polynomial fit”
However, if you want to measure the RATE, then a nth-order polynomial is your best friend.
this is a fit of HadCrut3, using a polynomial, which is information free (and so user bias free). After you do the fit all you do the calculus in you spreadsheet of choice:-
http://i179.photobucket.com/albums/w318/DocMartyn/hadcrut3fit.jpg
Why the reticence about the curve? No theoretical line predicts anything until proven, but a theoretical sine is 99% more likely to match a natural pattern than a theoretical straight line.
Nature doesn’t do lines, so any attempt to fit Nature to a linear trend is automatically illegitimate.
Ric Werme says:
September 2, 2011 at 3:13 pm
“I go apoplectic when someone tries to project a polynomial fit.”
Do you know why?
At which point in mathematical studies do most people stop learning math do you think? :p
PhilM, really? What’re the odds, eh? You are a warmist, no?
It looks like the developing (returning) La Nina is “sticking out its tongue” at global warming:
http://www7320.nrlssc.navy.mil/global_ncom/anims/eqp/sst30d.gif
Furthermore, the BoM Monthly Subsurface Pacific Ocean Equatorial Temperature Anomalies down to 400 Meters show a strongly cooling equatorial Pacific:
http://www.bom.gov.au/climate/enso/sub_surf_mon.gif
On this basis one would expect this month’s global UHA temperature downturn to be continued for a few months more. Plus a chilly NH winter ahead.
Then the curve-fitting fun will really begin!
Some Guy says:
September 2, 2011 at 1:25 pm
Instead of adding a polynomial fit, why don’t you (and everyone else, for that matter) try some Fourier series fits (instead of Taylor series fits)?
-/———
Assuming you are serious and not leg pulling, the answer is without multiple complete cycles to work on the transform result will be junk.