UAH Global Temperature Update for February 2012: -0.12 deg. C
By Dr. Roy Spencer
The global average lower tropospheric temperature anomaly cooled a little more in February, 2012, again not unexpected for the current La Nina conditions in the tropical Pacific Ocean (click on the image for the full-size version):
The 3rd order polynomial fit to the data (courtesy of Excel) is for entertainment purposes only, and should not be construed as having any predictive value whatsoever.
Here are the monthly stats:
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 +0.327 +0.321 +0.332 +0.155
2011 9 +0.289 +0.304 +0.274 +0.178
2011 10 +0.116 +0.169 +0.062 -0.054
2011 11 +0.123 +0.075 +0.170 +0.024
2011 12 +0.126 +0.197 +0.055 +0.041
2012 01 -0.090 -0.057 -0.123 -0.138
2012 02 -0.116 -0.014 -0.217 -0.281
Progress continues on Version 6 of our global temperature dataset, which will have a better adjustment for drift of the satellites through the diurnal cycle, and an improved calibration procedure for the older MSU instruments (pre-1998).
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
“”””” braddles says:
March 2, 2012 at 1:20 pm
The funniest thing about that 3rd-order polynomial fit is that historically, by chance, it does have predictive power. “””””
Well of course it does; if you run the time clock backwards from today, the cubic does a fairly good job of predicting the past events.
To those expecting a rebound into El Nino territory: That was what the models were predicting at this time last spring, but it didn’t happen. Then, last fall, some of the same models were predicting this La Nina would plunge to record levels, but that didn’t happen either. I tend to look at the Enso models “for entertainment purposes only,” and to feel they “should not be construed as having any predictive value whatsoever.”
For example, click onto the “ENSO/SST Page” tab at the right hand side of this site, and scroll down to the second graph, “CFS forecast Nino3.4 Forecast anomalies.” Compare the most recent model runs (blue lines) with the earliest model runs (red lines). The blue lines suggest that, after flirting with neutral, we sink back to the third dip of a triple dip La Nina. However the red lines suggest we warm up into El Nino territory. (This is in the graph updated Sat Mar 3 2012, involving “intial conditions: 21Feb2012 — 1Mar2012.”)
In other words, there is such variety in the model runs you can pretty much pick and chose one you prefer, if you squint hard enough into the spaghetti.
Then check back in a couple days. There will be a whole new batch of model runs to be dumbfounded by.
I’m adopting a wait-and-see policy. Triple-dip La Ninas are not that common, but I can spot at least three, looking back to 1950. The 1954-1957 one interests me, as that was at the start of a cool PDO cycle, I think. However the sun wasn’t as “quiet.”
In the end all I’m sure of is that an El Nino will warm things up, and a La Nina will cool things down. And I’m not 100% sure of even that!
Dinostratus says:
Is too.
Nope.
Minimally acceptable asymptotic analysis requires at least a passing glance towards the grouping of terms and the type of expansion.
Isn’t that, either. It is a perfectly acceptable third order polynomial trend line, fit to the subject data. You like even numbers? Use a fourth order. The shape and fit are the same.
As is the entertainment value, which is the designated purpose to which that curve is directed. Both third and fourth order polynomial fits cause guys like you to screw yourself into the ground, with little smoke plumes curly cuing from your ears. Then there’s the back flips you perform when you sooooooo desperately want to use a first order polynomial trend line, over your own goofily stated objections. Of course your desire for those is presently waning, as statistically significant timeframes are pointing those toward the state where – how did you put that again? Oh yeah – “…in the future the zeroth law of thermodynamics will be violated and we will have a negative absolute temperature.”
Who needs HD with 5 channel sound, when that is on tap?
George E. Smith; says:
March 3, 2012 at 10:11 am
Several questions come to mind.
1/ Are you suggesting that the “big jump early in the month” was a fiction and did not really happen ? IF that “big jump early in the month” really happened then of course the “number” would NOT “have been much lower” ; it would have been exactly what it was. If so why did you even mention it ?
2/ so “all” the evidence “”””” is pointing towards GHGs working as tiny little thermostats “””””.
What evidence is that since you cite none. I don’t disagree that “albedo changes” can affect Temperatures; but to the extent that such albedo changes might be cloud and by inference water related; there is also a direct absorption of solar spectrum energy, that permanently reduces the total solar energy captured by earth. That might affect Temperatures also.
If your data disagrees with Dr Spencer’s, then let’s see it; otherwise, why wouldn’t we accept what Roy said it was ?
1) No, I’m not saying it was a fiction. I’m saying the warming bump may have been due to a low probability event. Because of the timing of that event it hides the fact that February would have been cooler than what we ended up seeing. Electroscavening supposedly works to reduce cloud formation. Without the clouds the temperature increases until the effect wears off.
2) The evidence I’m looking at is the ERBE data shown in David Evans article last week. It supports a view of GHGs having two effects. The first is the GHE. The second gets no mention at all from the warmists. The GHGs radiate most of the energy the Earth receives back to space. If you add more GHGs to the atmosphere there are more parallel paths for that energy to take. The result is that GHGs may provide a cooling effect in addition to the GHE. The balance of these two effects may work like a thermostat. At low concentrations of GHGs the GHE dominates. As more GHGs enters the atmosphere the GHE starts to saturate and effect of the added radiation paths may just cancel out the warming at some point. If we’ve reached that point, I would expect to see a pattern in the ERBE data exactly as David Evans showed.
The data will need to be adjusted “faster than expected”.
“It is a perfectly acceptable third order polynomial trend line, fit to the subject data…..
You like even numbers? Use a fourth order……”
Again, the choice is arbitrary. Why throw an arbitrarily chosen shape across the data? Let the data stand by itself. It’s power is lessened by superimposing a shape that is for “entertainment value”.
Joshua, I would also direct you to http://discover.itsc.uah.edu/amsutemps/
Choose 14000 feet AQUA Ch 5 since this is the primary source of Lower Tropospheric temperature used for UAH.
It will give you an opportunity to compare the current collection data source by day since the current source started in 2002. Warning, although the label says 14000 feet, this channel is mislabeled. It is influenced by temperatures from other heights — including the stratospere. UAH uses a procedure to remove the influences of other heights before releasing its monthly estimate.
Dinostratus says:
Again, the choice is arbitrary.
Pay attention. The choice between a third order polynomial trend fit to these data (which you objected to) and a fourth oder polynomial fit to these data (which you suggested) is not arbitrary. It is irrelevant. They both produce the same shape, with the same fit statistics. You were all up on even order polynomial fits, when you very mistakenly thought that they would show something different – something more to your preconceived liking. Now you run away.
Why throw an arbitrarily chosen shape across the data? Let the data stand by itself.
If that is a genuine question, you will find a perfectly good answer in any high school mathematics program. Perhaps you should take one.
If that is a rhetorical question, then please, do publish your finding that the practice of fitting trend lines to data is wrong. It would seem that a Nobel prize awaits your revolutionary finding, Dr. Dino. Gvient the way the Nobel comittee hands those things out these days, you’re probably a shoe in.
It’s power is lessened by superimposing a shape that is for “entertainment value”.
He says, in glaring proof to the contrary. That simple trend line is the gift that keeps on giving.
A El Nino can only get so warm and so frequent enough to peak global temperatures. After that the only way is back down towards a less intense period or a period of negative PDO with more La Nina’s than El Nino’s. The previous decade will likely be the peak one until at least mid-century. Should be around May when global temperatures start responding from very recent changes in ENSO conditions towards neutral.
“The choice between a third order polynomial trend fit to these data (which you objected to) and a fourth oder polynomial fit to these data (which you suggested) is not arbitrary. It is irrelevant. They both produce the same shape, with the same fit statistics. You were all up on even order polynomial fits, when you very mistakenly thought that they would show something different – something more to your preconceived liking. Now you run away.”
No, a fourth order polynomial and a third order are not the same shape. That is a silly assertion.
However, you inadvertently make a good point, “something more to your preconceived liking”, is the reason why throwing a third order polynomial across the data is a bad idea. The third order polynomial is to Dr. Scencer’s liking and showing a liking across good hard data lessens the impact of the data. It is better for the data for it to stand alone. Embellishments distract from its truth..
Dinostratus says:
No, a fourth order polynomial and a third order are not the same shape. That is a silly assertion.
It is not a silly assertion. It is the simple truth. Fitted thru the UAH dataset, a third order polynomial trend line and a fourth order polynomial trend line are indistinguishable. You’ve been talking out of the wrong orifice since you started.
However, you inadvertently make a good point, “something more to your preconceived liking”, is the reason why throwing a third order polynomial across the data is a bad idea.
Yes, it is your desire to see a specific outcome “implied” that is the sole source of your objection. That is why you only object to a trend line fitted to these data and not all data, why you initially objected to a third order polynomial trend line while suggesting others, and why you run from those suggestions when they turn out to not behave they way you thought they would.
On the other hand, there are objective reasons why, of the readily available choices, a third order polynomial trend line is the best pick for these data, for purposes yet more rigorous than the stated entertainment value. And it is entertaining, watching the data bounce along a progression of very similar and better fitting third order polynomial curves month after month.
The third order polynomial is to Dr. Scencer’s liking and showing a liking across good hard data lessens the impact of the data.
Showing a trendline fitted to data does not lessen the impact of the data. That is why the whole wide world, not just Roy Spencer, uses trendlines in presentation of timeline data. Given that you think trendlines are an illegitimate distraction, why dont you take your one man crusade to rid the world of trendlines on the road? Once again, every field that trades on numeric data uses trend lines fitted to the data. You’ll make quite a name for yourself if you demonstrate that this practice is an incorrect “embellishment”.
“That is why the whole wide world, not just Roy Spencer, uses trendlines in presentation of timeline data.”
Congratulations on knowing all that is in the whole wide world but you must have skipped at least a part of it. Observing my small part of the whole wide world, I can say if someone put up a polynomial fit to data and said it was for entertainment value, they would be thrown out of the room. Now my small part of the whole wide world may be populated with black swans but if a physical reason for a third order fit is not given, the researcher would be told to get his GD fit line out of the way so the serious people could pay attention to the data without any distractions. The data is entrainment enough.