January 2008 – 4 sources say "globally cooler" in the past 12 months

January 2008 was an exceptional month for our planet, with a significant cooling, especially since January 2007 started out well above normal.

January 2008 capped a 12 month period of global temperature drops on all of the major well respected indicators. I have reported in the past two weeks that HadCRUT, RSS, UAH, and GISS global temperature sets all show sharp drops in the last year.

Also see the recent post on what the last 10 years looks like with the same four metrics – 3 of four show a flat trendline.

Here are the 4 major temperature metrics compared top to bottom, with the most recently released at the top:

UK’s Hadley Climate Research Unit Temperature anomaly (HadCRUT) Dr. Phil Jones:hadcrut-jan08

Reference: above data is HadCRUT3 column 2 which can be found here

description of the HadCRUT3 data file columns is here

NASA Goddard Institute for Space Studies (GISS) Dr. James Hansen:GISS January Land-Sea Anomaly

Reference: GISS dataset temperature index data

University of Alabama, Huntsville (UAH) Dr. John Christy:UAH-monthly-anomaly-zoomed

Reference: UAH lower troposphere data

Remote Sensing Systems of Santa Rosa, CA (RSS):rss-msu-2007-2008-delta520.png

Reference: RSS data here (RSS Data Version 3.1)

The purpose of this summary is to make it easy for everyone to compare the last 4 postings I’ve made on this subject.

I realize that not all the graphs are of the same scale, so my next task will be to run a combined graphic of all the data-sets on identical amplitude and time scales to show the agreements or differences such a graph would illustrate.

UPDATE: that comparison has been done here

Here is a quick comparison and average of ∆T for all metrics shown above:

Source: Global ∆T °C

– 0.595

GISS – 0.750
UAH – 0.588
RSS – 0.629
Average: – 0.6405°C

For all four metrics the global average ∆T for January 2007 to January 2008 is: – 0.6405°C

This represents an average between the two lower troposphere satellite metrics (RSS and UAH) and the two land-ocean metrics (GISS and HadCRUT). While some may argue that they are not compatible data-sets, since they are derived by different methods (Satellite -Microwave Sounder Unit and direct surface temperature measurements) I would argue that the average of these four metrics is a measure of temperature, nearest where we live, the surface and near surface atmosphere.


The website DailyTech has an article citing this blog entry as a reference, and their story got picked up by the Drudge report, resulting in a wide distribution. In the DailyTech article there is a paragraph:

“Anthony Watts compiled the results of all the sources. The total amount of cooling ranges from 0.65C up to 0.75C — a value large enough to erase nearly all the global warming recorded over the past 100 years. All in one year time. For all sources, it’s the single fastest temperature change ever recorded, either up or down.”

I wish to state for the record, that this statement is not mine: “–a value large enough to erase nearly all the global warming recorded over the past 100 years”

There has been no “erasure”. This is an anomaly with a large magnitude, and it coincides with other anecdotal weather evidence. It is curious, it is unusual, it is large, it is unexpected, but it does not “erase” anything. I suggested a correction to DailyTech and they have graciously complied.

UPDATE #2 see this post from Dr. John R. Christy on the issue.

UPDATE #3 see the post on what the last 10 years looks like with the same four metrics – 3 of four show a flat trendline. 

Sponsored IT training links:

Get professional help for your HP0-J33 exam! Download the 650-575 test questions for practice and pass 117-201 exam on first try.


newest oldest most voted
Notify of
Jeff in Seattle

Seems like GISS is the odd man out and should be discarded as an “adjustment”.


Anthony – if you’re going to use a common reference period, use a 30-year period that ends in 2000 (the latest decade). This is at the recommendation of the WMO.
I think that would be 71-2000 (which should still cover the “modern global warming” period).
Remember, GISS consistantly shows as the “hottest” because of the reference period (51-80, I believe), while HadCRU and NCDC use later time periods.


Why in the world should 2000 be used. we’re already almost a decade past that and inconveniently, showing, a dramatic decline in temps. This past 7 years has already given back all of the gain of the past 150 years. Shows how vulnerable the purported changes are to real results.


speaking of real results, did you know it snowed in San Diego last week?


This drop in temperature is certainly very unusual. The fall of 0.595 degrees since Jan 2007 is the largest January-to-January drop in HAD CRU3 since 1875, and the biggest drop for any 12-month interval since -0.681 degrees in February 1974.
The January temperature is the lowest for any month since 1994, and the lowest for a month unaffected by volcanic eruptions in 20 years.
January 2008 was cooler than January 1932, even after all the downwards adjustments they have applied to the old data.

Nice to see that the data with the highest decline is the GISS, which has relied on Polar Amplification to reinforce its warming position. There’s way too much noise to pluck anything of value from the N Pole data, but the MSU data shows the Extratropics have dropped 0.94 deg C in the last 12 months. That’s a chunk, to use technical language.
NOAA is still showing an elevated SOI (1.9 for January, 1.8 for December, 0.9 for November). Dependent on the time lag between the SOI and the La Nina (2 to six months typically) we’re in for some chillier weather for a couple more months, assuming one believes ENSO has an impact on global temp. It’s tough to miss.
How long can the trend continue? How long is the lag between solar and global temperature? We won’t be seeing the rest of the recent drop in TSI for a few more years now. There go a couple more tenths.
It’ll be fun to watch.

Joe Black

So how’s that correlation of the temperature anomaly with the atmospheric CO2 change going? /rhet


Does anyone have an early-access look at UAH or RSS data? Can we see were we stand 20 days into February?
REPLY: No. This is why I keep saying I’m not going to analyze the surfacestation data until the survey is complete. Don’t rush the science.

So, the oceans have turned cooler or at the very least stayed the same over the last few years, and the temperature of the atmosphere have cooled significantly over the last decade. The idea of “committed warming”, i.e. a strong radiative imbalance is rather hard to reconcile with the fact that the climate system has lost heat.
It should be time to
1) use a relatively short response time to perturbations for the climate system when calculating the sensitivity and
2) admit that either the sun or possibly ocean currents play a much bigger role then admitted by the IPCC.

Patrick Hadley

Can I as an AGW agnostic sound a note of caution: that sudden changes in temperature are usually quickly reversed. If that happens we could soon be back in record warm temperatures.
On the other hand it is pleasant to see the low temperatures on the graph. Thank you for doing all this work.

Bill in Vigo

I surely wish that we could have unadjusted records over the peroid of record. I am not a scientists but am having trouble with the current adjustments They seem to be designed to make the recent trend warmer. The raw data dosen’t seem to be doing that.
I just wonder if GISS is trying to adjust the older records up to match current UHI levels? If so they are surely being disengenious. I will not say they are deliberately changing the record, but I do have trouble in that they do seem to have personal bias toward warming and their adjustments seem to show that.
We need to get back to raw data and eliminate the microbias stations from the USHCN and start using good methodology.
From the layman’s point of view I want climate science that I can trust. If we are going to use proxies, use ones that are conclusive (not tree rings that might be moisture related, CO2 related, or temperature related).
Thanks Anthony for such a good job..


Drop about another .5C, and we’re basically back to where we were in the mid-late 1800s when the Little Ice Age ended.
I’m really, really anxious to see Feb data now.


“Seems like GISS is the odd man out and should be discarded as an “adjustment”.”
That is *exactly* what I thought the instant I saw their data. Looks like they knew they were overestimating for a long time, and this was a great time to adjust their reporting to be in line with everyone else.
REPLY: Soon, after examination of the algorithm, they may be singing “Thats the nights that the lights went out in Georgia….” There is a flaw in the adjustment, see my post on Cedarville…urban adjustment applied to a stable rural station. Nightlights is flawed.


“Lights=0” = 0?


I wonder if there will be a scientific consensus that this temperature drop was not caused by humans?

Gaudenz Mischol

Please keep in mind that GISS takes 1950-1980 as their reference period and not 1960-1990 as the others do. This may explain the higher values.


Hmm. I think that would only affect their baseline (i.e., “what is zero”), but not the actual raw amount of the drop (0.75C).
And this is odd, condsidering their recent adjustments to NOAA data are nil–they don’t push current temps UP (any more than the NOAA), they have been pushing past temperatures DOWN.
“I wonder if there will be a scientific consensus that this temperature drop was not caused by humans?”
No comment on the subject would probably translate to “not human”. OTOH, there are those who say “it’s Global Warming” (Dr. Pielke points this out in his blog.)


BTW, what happens when one overlays the 100-year record of NASA and NOAA (metadata) with the NOAA data (raw, with only outliers removed)?
REPLY: So many questions, not enough time.


You can actually follow daily temp UAH here
by clicking on the graphs you could roughly calculate the temps for feb up to current but as Anthony says it wont be relaible as it’s going up and down ect but its still below (that’s the lower troposphere data)
REPLY: I’ve visited this resource and I’ve seen some errors in the way javascript presents the data on that website, so take it with some caveats that it may not accurately represent the result due to it being an interactive web application.


What stands out to me isn’t the cooling, but the discrepancies in the numbers. There are plenty of variables that could alter the absolute temperature for each study. But if these numbers truly represented a global average temperature, there should be a very high degree of agreement over the delta T. Particularly in the most recent, and presumably most reliable, numbers.
When the data for the single most precisely measured year, shows a disagreement equal to fully 1/4 of the claimed total change, the system simply isn’t knowable at a relevant level. Assuming all the numbers are gathered and adjusted in good faith, that data can only get less precise in preceding years.


GISS includes the poles – the other temperature series do not. A larger drop in GISS simply indicates that the poles got colder faster than the rest of the globe. I would not read too much into the difference between them.
That said, I am not convinced that GISS polar measurement estimates have any connection with reality but that is a seperate issue.

Wondering Aloud

Patrick is absolutely right …except for the part about it being nice to see low temperatures on the graph. I am freezing.

steven mosher

Raven, GISS ESTIMATE the Poles , hadcru does not.


More data for the mix.
El Nino/La Nina effects on ocean temps (+/- 5 degrees from equator, 120-170 degrees West. (Celsius), 1950 to 2007

Obsessive Ponderer

When you redo the data for the temperatures to a common reference, could you make the raw data available on your website? I have some neat statistical software I am playing with, find it frustrating using all the difference data sets and don’t have a really good idea how to change it all to a common reference.
REPLY: There are links to the raw data below each graph, maybe you missed them?

Jeff C.

“You can actually follow daily temp UAH here
by clicking on the graphs you could roughly calculate the temps for feb up to current but as Anthony says it wont be relaible as it’s going up and down ect but its still below (that’s the lower troposphere data)”
As Anthony mentioned the page is quirky and doesn’t seem to plot properly. However, you can select show data as text and get a nice data dump for the past 10 years and then play with it in Excel. I selected 3300 ft. altitude (lower toposphere) and performed a few quick calculations.
Dec 2007 average = 270.209 Kelvin
Jan 2008 average = 269.950 deg Kelvin
Feb 2008 average (through 2/18) = 270.007 Kelvin
February to date is roughly a 0.06 deg C increase from January. However, take it with a grain of salt and please check it out yourself as this was real quick and dirty calculation.


Take a look at NOAA’s breakdown of global temps. NH land is showing a negative anomaly and is a huge 2.4C cooler than last January (no mistake in that number).
Interestingly, Jan 2008 was the warmest Jan on record in Australia. February 2008 looks like being substantially cooler. So Feb 2008 will probably see the SH ‘catching up’ with the NH.


Thanks for the average. Now somebody with time on their hands should average out the four time series over the whole satellite era.
Maybe there could be a consensus that, everybody should use this average of averages as a common reference base when talking about recent global temperature changes. Without an aggregate, the debate always becomes pointless as parties get to pick their favorite temperature series.

Gary Gulrud

While differing baselines could displace one graph from another linearly it seems that three graphs are congruent, the other not.
I would say as a result that the work done on the fourth is not confirmed, and barring justification of the kind Raven and Steven M. are sorting out I might even call it invalid. Perhaps I’m anal, though.

steven mosher

I Am toying with a statistical test.
Giss say that they know the global temp per year within +-.05C
Hadcru say the same. how close should they be? how far apart

I’ve started “Rurudyne’s Daily Global Cooling Watch” and would like any feedback or suggestion on the article I’ve assembled. The following link is to a nice, safe, non-political forum … yes, it’s a Transformers Fan Forum. –.^

[…] online sources (e.g. Watts Up With That) point to other sources providing similar messages, including that the year-to-year change in […]


“February to date is roughly a 0.06 deg C increase from January. ”
But that’s an increase over Jan.’s temperature, not Feb.’s zero-anomaly, right?
REPLY: lets not speculate on future datasets please. Let them be released first then we can all squabble over the numbers. Don’t rush Science.


#Re Jeff C: February to date is roughly a 0.06 deg C increase from January. However, take it with a grain of salt and please check it out yourself as this was real quick and dirty calculation.
Thanks for that! great calculation…A re-check at the end of month should confirm whether we can use this for approximations at least.


The whole idea of man made global warming is so ridiculous. Will the liberal establsihment of pop-culture ideas admit their fault? That will be interesting to see! I think it will take some time. Whichever way earth’s current climate trend goes, there is little or nothing we can do about it. We can NOT PANIC and adapt as needed.


A.) Okay, okay, I’ll send back my cattle prod.
B.) The IPCC followed me home. Can’t I keep it, huh?
C.) But Unca’ Rev, all the OTHER kids get to rush the science! THE’VE already GOT a hunderd years! I don’t get to have ANY fun!

[…] In fact, Watts Up has four such charts. […]

Stan Needham

Let them be released first then we can all squabble over the numbers. Don’t rush Science.
But, but, but, Algore said we only have 10 years left, and that was nearly 2 years ago. Geez, if we can’t rush the science, we’re DOOMED!! Floods, earthquakes, tornadoes, hurricanes, forest fires, droughts, rising sea level, oh the horror of it.
Sorry, Anthony, that just seemed to fit this thread.


Sam Said: (20:34:11) :
“Why in the world should 2000 be used. we’re already almost a decade past that and inconveniently, showing, a dramatic decline in temps. This past 7 years has already given back all of the gain of the past 150 years. Shows how vulnerable the purported changes are to real results.”
And that’s the point. Why does GISS continue to use 51-80 for their reference period, almost 40 years in the past (and in a cooler period, before the “modern global warming” period.)
If you compare surface temps to a walk-in freezer, you WOULD get higher anomalies.


“Will the liberal establsihment of pop-culture ideas admit their fault?”
The jury is still out. (Consensus OFF. Debate ON.)
It seems probable to me that man has had some measurable effect on the temperature. However, there is preliminary evidence that temperature has been poorly measured and very badly adjusted by those responsible. Data is in the process of being gathered regarding this. Results still pending.
There is also evidence of stronger correlations to factors other than Greenhouse gasses, particularly the PDO/AMO cycle, which may well account for warming of the the 19-teens-’30s, cooling from the ’40s-70s, and warming since 1980. (The degree of warming since 1980 being particularly called into serious question). The temperature trend of the next five years may well provide (most of) the answer.
Mere correlation is not proof. It is a mere starting point for further empirical observation. Lack of correlation, however, may be considered likely to be disproof.
I think that puts it pretty fairly.
And what if those of the pop-culture establishment (which I do not consider to be “liberal” in any real sense of the definition) do, in the end, turn out to be wrong or mostly wrong, then will they admit their error?
As a liberal apostate (and continuing bleeding-heart liberal), I think I can answer that question by consultatiion of the the historical record:
Every bit as much as they admitted error on demographics.
Every bit as much as they admitted error on resource depletion.
Every bit as much as they admitted the “surge” was a success.
Every bit as much as they admitted that poverty has been reduced.
Every bit as much as they admitted that the W.H.O. is right about DDT.
Every bit as much as they have admitted Paul Ehrlich was wrong.
Every bit as much as they have admitted Dennis Meadows was wrong.
Every bit as much as they have admitted Herman Kahn was right.
Every bit as much as they have admitted tax cuts result in greater revenues.
Every bit as much as they have admitted that putting bad guys in jail reduces the crime rate.
Every bit as much as they have admitted that Reagan caused the fall of the Soviet Union.
I find your lack of faith disturbing.
Apology accepted. (*insert appropriate ghastly sound effects*)

“Will the liberal establishment of pop-culture ideas admit their fault?”
Evan … At the risk of citing a ‘liberal’ I must wonder at the notion that the luminaries of pop-culture have any ideas at all…. Well, good ideas.
Since the 1970s we have so many things in our popular culture that they just couldn’t have imagined and didn’t imagine. In their most cynical moment of mad brilliance all they could come up with, all they could place on the stage with Howard Beale ( http://www.youtube.com/watch?v=NfGdbFh6cSI ), can’t hold a candle to the sheer mendacity and creative incompetence which these luminaries themselves have actually foisted on this nation.
They had Beale mourn the unreality of TV and yet their peers now give us “Reality TV.” Beale said “Go to God! Go to your gurus. … Go to yourselves!” but now they invite us to come to Oprah or any of the countless “People’s Court” clones.
It’s amazing to realize how unimaginative they were about themselves and their peers, what THEY were capably incapable of.
Could they make Network today? What would they say to us today? Not “Turn off your TVs!” because how would that play with the people paying for product placement ads in the movie? Maybe a Howard Beale in 2008 would be watered down to talk about the evils of … um … well …hmmmmmm?
Maybe I just lack the imagination to think of anything they might do that they haven’t already done. Go figure?


Wasn’t ’98 “the hotest year on record” due to a super El Nino? How come GISS missed it ?

John Willit

The 30 day temp anomaly from the NCDC (Jan 20 to Feb 18) is here.
The map shows the extreme cold conditions in southern Asia, north-west North America, parts of Arctic and Antarctica over the past 30 days. Europe and Russia are above average.
The 7 day anomaly shows there has been some moderation in all the anomalies.
So, February to date might be showing some moderation in temps from January.
And thanks Anthony for a great report and great website (people should print this out and show it to their global warming friends).
But the biggest cause of the temperature decline this year is La Nina. The latest sea surface temperature map shows La Nina is still going strong but ther may be some weakening. Historically, global temperature lag the La Ninas / El Ninos by 3 to 6 months so there should not be any significant increase in temps for at least 6 months.

Bruce Cobb

I hate to say it but, any drop in temperature if and when recognized will be attributed to AGCC. C02 will still be the villain. It’s the whole “tipping point” thing, and because we’ve upset nature’s “delicate balance”. Global Warming is so Yesterday.


Just to make sure I understand this, you are comparing January 2007 with January 2008, and drawing your conclusions from those numbers alone?
REPLY: No, but you seem to be. This is a very large anomaly, largest seen in years, coming on the heels of predictions of increasing temperaturs, so worth pointing out. Besides the fact that 4 metrocs are in agreement, what is also worth noting is that this large anomaly coincides with a large La Nina, a PDO shift, and a deep solar minimum. See previous posts on the subject.

Gary Gulrud

Roger: They’ve moved on to bigger and better things, no dwelling on their legacy of achievement–commendable in a way.

steven mosher

henry and others, changing the anomaly peroid just sifts the line vertically, it doesnt change the trend and the trend is the issue

Harold Vance

Roger, technically 1934 was the hottest year on record.
It is impossible to tell which periods were warm relative to others by looking at the GISS chart of anomalies. For that, you need to look at a chart of the actual temperatures (preferably unadjusted). Frankly, I don’t have a clue what the GISS chart represents. There is no sign of cooling from the 1940’s to the 1970’s, and there are no significant anomalies during the 1930’s.
I don’t really have any opinion about global warming, but I do have a strong opinion about how the data is being collected and how it is being “adjusted.”
PS. If you want to read a defense of Hansen, check out Mark Bowen’s Censoring Science, which was published in December. There isn’t a single mention of any of Hansen’s critics, their reverse engineering of his mistakes or Hansen’s stonewalling of requests for information. When critics are mentioned, they are labeled “deniers.”


Does anyone have a handy link to one (or more) of the Big Four monthly anomalies? (I.e., the “average” for each month from 1979-1998 (or whatever time period .)
“and yet their peers now give us “Reality TV.” ”
Actually, I think FOX started that.
“Could they make Network today?”
Maybe as a South Park episode.


Never Mind!
For purposes of comparison, here are the NOAA global anomolies 1900-2000
J: 2.8
F: 3.2
M: 5.0
A: 8.1
M: 11.1
J: 13.3
J: 14.3
A: 13.8
S: 12.0
O: 9.3
N: 5.9
D: 3.7
Mean: 8.5
J: 12.0
F: 12.1
M: 12.7
A: 13.7
M: 14.8
J: 15.5
J: 15.8
A: 15.6
S: 15.0
O: 14.0
N: 12.9
D: 12.2
Mean: 13.9
For those jumping the gun, note that Feb. is normally a warmer month that Jan.