NCDC updates database for Dec08 – NCDC's own graphic shows decadal cooling trend

One of the best things about WUWT is the number of eyes and minds at work, multiplying the efforts. This is interesting. Now that the 1998 El Nino is disappearing off the 10 year scale, things are looking a bit different

From “crosspatch” in comments:

NCDC now has December 2008 in the database. Annual North American temperature since 1998 (11 years of data) is falling over the period at a rate of 0.78(F)/decade or 7.8(F)per century. At this rate we will be in an ice age within 5 decades. If you can get the graphic, the heavy black like is the average over the century 1901 to 2000.

Here is the graphic from their automated graphics generator linked to their database:

ncdc-december-2008

Source: National Climatic Data Center

While the link he provided is only a result, I’m sure he’ll share the method in comments to this post.

UPDATE: He has indeed, see below. Try your own hand at it. The trend will likely flatten a bit with the removal of 1998 from the 10 year set. Of course you could pick any number of scales/periods and get different results. The point being made here is that the last 10 years hasn’t met with some model expectations.

Also I have corrected in the text the reference to Centigrade when it was actually Fahrenheit, note the (F). NCDC being an arm of the US government operates on the English unit system whereas most other organizations use metric, and thus Centigrade. I’ve made the mistake myself, so has NASA, who famously lost a Mars probe when they botched orbit entry calculations by use of Metric and English units on different science teams.

UPDATE2: Some folks are erroneously thinking that this graph above represents a global trend, it does not. Read on.

It represents US data from NCDC. Also there has been the usual complaint that “10 years isn’t long enough to determine any useful trend”. Perhaps, but when NASA’s James Hansen went before congress in 1988 to declare a “crisis in the making”, there had only been about 10 years of positive trend data since the PDO flip in 1978. It seemed adequate then:

hansen_predictions

In the graph above, note that the GISS station data does follow the Hansen C scenario, but that we are currently well below it.

Yes we really do need longer data periods to determine climate trends, 30 years is the climatic standard,  but you can also learn useful information from examining shorter trends and regional trends.

To generate the graphic I made:

First navigate here

Leave the “Data Type” field at “Mean Temperature”

Select “Annual” from the “Period” field pull down

Select “1998″ as “First Year To Display”

and click the blue “Submit” oval at the below the data entry form.

Advertisements

  Subscribe  
newest oldest most voted
Notify of
VG

and again hathaway is always right as the goalposts are moved and D Archibald wrong
http://www.solarcycle24.com/ (sorry for being sarcastic…). D Archibald will still be much closer I reckon at 40

Robert Wood

Uh oh, temperatures are normal.

ak

Choose an exceptional year and make a trend line from that point – brilliant! How can you generate that graph for a 20 year trend? 30 years? 50 years?
And shouldn’t the trend line be within the data shown? Why is it below the actual temps?

Keith

I’m just waiting for the cries that this is only a regional trend, since this is from the U. S. National Climatic Data Center, not a true Global record rendering. Even more, that the cooling trend is only due to the last two years’ aberration , and ten years do not a climatic trend make.
Gavin? Tamino? I’ve stated the party line for you, and your cue has been said, so please stroll in from the wings now.

Adam Gallon

A Snowy Owl has been seen in Cornwall, first time since 1948.
http://news.bbc.co.uk/1/hi/england/cornwall/7806553.stm
Surely a sign of “Global Cooling”!

Jim G

Where at NCDC is the data located?

crosspatch

To generate the graphic I made:
First navigate here
Leave the “Data Type” field at “Mean Temperature”
Select “Annual” from the “Period” field pull down
Select “1998” as “First Year To Display”
and click the blue “Submit” oval at the below the data entry form.

crosspatch

Note that the narrative at the main page has not been updated yet for December but the December data is in the database. You can do a comparison of all Decembers and notice that it now includes December 2008. Yesterday it went only to 2007. You will note that this December is the 35 coldest since 1895 at about 1C below the 1901-2000 average temperature for North American Decembers.

crosspatch

oops, those temperatures are in F not C. Sorry, I am more used to seeing degrees Celsius when working with climate data and didn’t look closely at the units.

crosspatch

“Choose an exceptional year and make a trend line from that point – brilliant!”
Well, actually it is 11 years, I just did 2008 – 10 wanting to see a decade of data. It should be 1999 which only shows -4.9F/century trend. My apologies.
The point wasn’t in picking any particular year to show things being any worse, I was simply trying to show the past 10.

I wonder about the scorce of this temperature data considering the change in urbanization (rural versus urban) since 1900 and scientific location and maintenance of measuring equipment.

jack mosevich

AK: all of your questions can be answered at the site pointed to by crosspatch. You can do annual as well as specific months.

7.8C per century ???? A continuation like that would prove me to have an extreme talent for modeling future global temperatures. If that rate were to continue then my temperature reconstruction (in response to Mann’s Hockey Stick) would become an extremely accurate bit of work. Should my “HockeyDip” ( The Excluded Data / HockeyDip LINK ) become reality would the entire approach to climate ‘science’ be changed? If you understand my words with the graphs then of course the answer is no…. not really.
For a more serious response. I see several scenarios where we could very likely face conditions similar to the LIA. The next couple of decades should provide science with a natural change to learn from and everyone to experience.

Steve M.

ak,
How can you generate that graph for a 20 year trend? 30 years? 50 years?
I did a graph from that site from 1939 to 2008 for December… -0.01f/decade. So, absolutely no warming or cooling for 70 years. How’s that for a trend?

Edward Morgan

Nice one crosspatch!

Dave

Are global temperatures from December 2008 already available somewhere?

Boris

This again? Really?
REPLY: Cowardly snark again? Really?

Matthew

WOW !!! Even the 1900-2000 trend is FLAT !!!!

George E. Smith

I’m always bothered by these “trend” Lines. It seems quite common to simply join the first point to the last point with a straight line, and call that the trend.
So presumably there is some climate master theory that says climate always progresses linearly from one state to another.
Almost all of the real physical phenomena I am familiar with tend to change from one state to another with an exponential decay transition, for the simple reason that the processes driving the change tend to generate a restoring force that is proportional to the deviation from equilibrium.
It also seems to me that the average value from 1901 to 2000 is quite irrelevent as a reference level for what has been happening since 1998.
The ten year time scale may be unreasonably short in the climate world, given that we have hard observation evidence of definite changes that have taken, and are taking place within the system. I don’t have a problem with Croospatch putting the green line on the long term average as he has. The angular difference is clear to the eye, and it doesn’t distract from the recent data. But if the start point were 20 or 30 years ago, the trend would be somewhat different, and the recent (since 2005) movement might look more ominous; sort of an upside-down Hockey stick.
2009 is likely to be an intriguing climate year as 2008 was.

Steve M.

I thought this was interesting:
http://climvis.ncdc.noaa.gov/cgi-bin/cag3/hr-display3.pl
one cooling cycle (1940 – 1970) and one warming cycle (1970-2000) … again essentially no trend.

L Nettles

“ak (09:31:36) :
Choose an exceptional year and make a trend line from that point – brilliant! How can you generate that graph for a 20 year trend? 30 years? 50 years?”
Ok, then what is the proper starting point, defend your selection

crosspatch

“I wonder about the scorce of this temperature data”
The main page says:

Data used to calculate United States mean temperatures are from the USHCN version 2 data set.

So the data in the database is probably biased warm judging from Mr. Watts’ survey findings to date. The more recent data is probably being biased warmer than the oldest data.

So sprach VG (09:21:41): “…and again hathaway is always right as the goalposts are moved…”
Golly, guys, what’s Hathaway supposed to do, commit Hairy Kerry? If the weatherman predicts 100°F, and it snows instead, does he close up shop? No! He simply predicts something lower for tomorrow. Life goes on. Sooner or later, he’ll get it right. Do you suppose that guy who wrote, “DEWEY DEFEATS TRUMAN!” never wrote another headline? No! You seem to think Hathaway has to get it right every time. That is a misconception. He’s there to make an intelligent guess based on NOW. He’s doing the best he can, under the circumstances. Sooner or later, he’ll get it right. Right?

ak

what a difference a year makes – 1997-2007.
@crosspatch, seeing as how this was given it’s own blog entry, i thought some sort of significance was being placed on this subset of data. anyway, the entry has brought to light (for me at least) this fun tool! thanks for sharing.

Adam Sullivan

While one can set different periods and get different trends from the dataset available, it is unmistakable that the climate models of 7 to 10 years ago which were used to scare the crap out of thinking people worldwide are now proven inaccurate.
Thinking people (a minority on this planet, and especially so among consumers of western media) need to step back and rethink past assumptions. One replaces assumptions with facts as time proceeds (or at least modifies the assumptions). That is what thinking people do.
Problem is we have a self feeding alarmist industry that is now in the fact distortion / rejection business. Meanwhile, the planet’s most vulnerable populations will feel the effects of a cooling cycle much more than of a warming one. But nothing will be done to prepare because too many “scientists” need to save face and preserve a meal ticket.

crosspatch

” jorgekafkazar (11:16:46) : ”
What is going to be very interesting to me are GISS numbers, particularly for stations in North America. When a data value is missing for a particular month in the GISS input data that it gets from NOAA, Hansen apparently calculates an average value that he uses to “fill” the missing value.
When we have a month that is significantly warmer that usual for a given station this new value is used in the calculation for “missing” values of that month in the past so a warmer December 2008 would cause past values for December that were calculated due to missing data to rise. This is why when looking at GISS plots, not only does the current month change, but the past temperatures also change as well. So a warm month can change the past using Hansen’s methods.
When we have a month that is significantly cooler than usual, by the same token, past temperatures that were calculated using average values for that month will be reduced. What we should see, if December is cooler than average is a decrease in GISS temperatures in the past as well as the current month. This means that years that were hyped as showing unprecedented warmth will now be cooled due to a cooling of the average.
Hansen’s method worked well for him when temperatures were above average. The other edge of the sword comes into play when temperatures decline and his method should also exaggerate cooling as it has exaggerated warming. Cooling this month should also cool the recent past.
I am interested to see if Hansen now modifies his algorithm to fill missing data since he is now being faced with a rather inconvenient truth.

The point of making a prediction is not to demonstrate psychic ability, NOR is it to gamble.
In science (remember that?) the idea is to establish a hypothesis or theory, maybe establish a methodology, then have your idea or work generate a prediction. If that prediction proves accurate then your theory or hypothesis is reinforced. If the prediction is gravely wrong then your theory or hypothesis has been disproved. As anyone with a science background should know, there is no way to PROVE a theory. Theories can be accepted as probable, or disproved.
If your prediction about climate or solar activity or whatever doesn’t even remotely match the reality then your theory or hypothesis is wrong. Even if your idea crosses from theory or hypothesis into belief, that does not make it any more correct. Just because you get a few things right once in a while doesn’t mean your ideas are correct or vindicated, since a stopped clock is right twice daily.
If you have strong beliefs in the AGW debate (you know, that debate that doesn’t exist since the “science is settled”) and predictions don’t even remotely match the reality, perhaps you should rethink your position.
Yeah, downward trends for a decade are significant, especially during a time when we are being told over and over that “warming is accelerating” and things are “even worse than predicted” (if things were worse than predicted, we’d all be swimming in the boiling oceans right now).
Remember, AGW alarmists tell us that CO2 should be resulting in increased temperature, increasing trends, and once the heat is captured it has no place else to go. Therefore, ANY cooling (not just a decade’s worth) has disproved the AGW theory. It’s time to let it go, guys. Man up and get a new pet theory.

jorgekafkazar (11:16:46) :
Hathaway’s doing the best he can, under the circumstances.
I agree, Hathaway is a good scientist.

ak

@L Nettles, i would choose time span that didn’t start with the peak of a solar cycle and end at it’s trough. wouldn’t you agree?
@Steve M. not really sure why you choose to selectively graph only the month of december and exclude the other 11 months, or 92% of the year. can you justify? otherwise the trend is 0.18 F/decade.

Ron de Haan

Adam Gallon (09:41:40) :
“A Snowy Owl has been seen in Cornwall, first time since 1948.
http://news.bbc.co.uk/1/hi/england/cornwall/7806553.stm
Surely a sign of “Global Cooling”!
There was a snowy owl spotted in Texel, The Netherlands November last year.
http://www.nd.nl/artikelen/2008/november/17/zeldzame-sneeuwuil-gespot-op-texel.
Maybe it’s the same owl taking advantage of the all time low of the English Pound?
The last time there was a snowy owl spotted in The Netherlands was 2002.

Dave

@ Ron de Haan (11:38:42) and Adam Gallon (09:41:40) :
For the moment, there is also a Snow Owl in Belgium. He is here since November 24 2008 and seems to like it 😛
It’s the 7th time since 1896!!!

George E. Smith (10:48:04) :
I’m always bothered by these “trend” Lines. It seems quite common to simply join the first point to the last point with a straight line, and call that the trend.
Respectfully….. The slope (rate) of decline represented by the green line is created automatically by the software. It was not added by Crosspatch. While there are other methods of visually presenting data “trend lines” are a simple process and meaningful for the intended purpose.
While a decade is not a significant period of geological time in regards to long term trends, neither is 3 decades or 10 decades. While start point, baseline, etc. can all be used to ‘fudge’ or misrepresent they are necessary when you desire a trend for a specific period. If one wants to know the trend from 1927 to 2001 it is inappropriate to employ a start point of 1850 and an end point of 2008. Doing so will produce the ‘wrong’ response since both points are outside of the desired range.
In a recent debate with someone (a Global Warmer who works professionally in the field) regarding the ‘current cooling trend’ their rebuttal was to produce an IPCC graph illustrating the trend from 1964 to 2003. With that the were trying to prove the current cooling did not exist. Granted it was a 40 year span instead of the 8 years I was referencing. However, what in the world does the time span they used have to do with the current trend? NOTHING.
Lee

crosspatch

Please note that NOAA says on the page that displays the graphic:
“Some of the following data are preliminary and have not been quality controlled.”
It will be interesting to see if December is “adjusted” warmer in a week or so.

Mike from Canmore

Source: Nattional Climatic Data Center
Anthony: Small mistake up there but I have to laugh.
I discovered a guy I work with, nice guy, is a true warminist so I asked him if he wanted to discuss the science. I got the usual response, “all these scientists . .. , last 650,000 years, etc. etc. I corrected him on the “all these scientists claims” and got the “yea, there all paid by Exxon Mobile, blah blah blah.”
So I pointed him to your site
The response I got was pointing out the error above and claiming he had trumped me in my sources. I don’t get it. He is a bright intelligent guy who loves the outdoors. When things are held too close, some bad thinking goes on.

Bernie

My initial issue with most of the AGW debate was a sense that there was a false precision almost everywhere I looked, coupled with what appeared to be significant massaging of the data. Crosspatch’s little exercise serves to illustrate that we can readily manipulate the appearance of a trend. The iconic hockey stick by greatly expanding the time period for the trend is extremely powerful support for AGW proponents BUT it is based on what appears to be massaged data and fails to reflect known patterns of past temperatures, i.e., MWP. Bottom-line, the climate models appear to be seriously incomplete – presumably somebody is busy trying to make them more complete.

Patrick Henry
Neo

Newsweek will be happy to know that they were right all along.

Jeff Alberts

Dave (11:57:01) :
For the moment, there is also a Snow Owl in Belgium. He is here since November 24 2008 and seems to like it 😛
It’s the 7th time since 1896!!!

Maybe it took that long to save up enough Frequent Flyer Miles?

RSS is out: 0.174

Jeff Alberts

The response I got was pointing out the error above and claiming he had trumped me in my sources. I don’t get it. He is a bright intelligent guy who loves the outdoors. When things are held too close, some bad thinking goes on.

People don’t like to admit they’ve been duped. It’s understandable. But I can hardly blame the average person, since they have the AGW nonsense spewed at them CONSTANTLY.

Glenn

“If the weatherman predicts 100°F, and it snows instead, does he close up shop?”
Well I doubt his ratings would go up. Take it from there.

crosspatch

Bernie:
There is an article today over at Climate Audit today that illustrates your point about longer term trends. It is a study in Finland of old trees that were found in ponds. They show periods of much warmer temperatures during the Holocene than we are looking at today with pine forests extending to where there are none today because it is too cold now for them to grow. The graphic there shows where the tree lines were doing earlier in this interglacial.

Between c. 8000 and 4000 cal. yr BP pines were growing at 350/400 m higher altitudes than at present and the shift in mean July temperatures compared with 1961/1990 climate normals was ~2.5/2.6 deg C. … During the ‘Mediaeval Warm Period’ the distribution area of pine was 7200 km2 more extensive than at present, and pines were growing at 40/80 m higher altitudes. For this period, the mean July temperature reconstruction
shows ~0.55 deg C shift compared with the present.

So today’s temperatures aren’t anywhere near “unprecedented” nor are they particularly unusual. If you consider that 6000 years ago temperatures were some 2.5 degrees C warmer than now in Finland, we have cooled considerably since then. In fact, just about every long term (meaning more than 2000 years) temperature chart I have seen shows us in a long gradual cooling trend with each cool period cooler than the previous and each warm period cooler than the past (e.g. the late 1990’s being slightly cooler than the early 1930’s). This also corresponds to finding of wood uncovered from receding alpine glaciers in Europe exposing wood that dates back some 5000 ybp. Many of those glaciated valleys were ice free 5000 years ago and had been ice free long enough to become forested where there is nothing but ice today.
Gore’s story is fine for people who have no perspective beyond their own lifetime or that of their parents (who tell stories of much harsher winters). But in a longer term perspective that must be researched to be found (which few average people seem to have the inclination to do), we find that we are actually colder now than it was when the Mesopotamians were brewing their first batches of beer.

Joel Shore

Anthony Watts says:

The point being made here is that the last 10 years hasn’t met with some model expectations.

Adam Sullivan says:

While one can set different periods and get different trends from the dataset available, it is unmistakable that the climate models of 7 to 10 years ago which were used to scare the crap out of thinking people worldwide are now proven inaccurate.

Okay guys, so can you show me that the data seen over the last 10 years for the U.S. is out-of-the-range of what climate models project? You seem to be making a comparison between the observational data and the climate models without actually showing us anything from the climate models. To see what the climate models predict on a global scale (where noise is much less dominant than on a regional scale), you might want to start here: http://www.realclimate.org/index.php/archives/2008/05/what-the-ipcc-models-really-say/langswitch_lang/fr
CodeTech says:

Remember, AGW alarmists tell us that CO2 should be resulting in increased temperature, increasing trends, and once the heat is captured it has no place else to go. Therefore, ANY cooling (not just a decade’s worth) has disproved the AGW theory. It’s time to let it go, guys. Man up and get a new pet theory.

No, you don’t disprove a theory by making up a strawman version of the theory which has no relation to the actual theory and then finding data in contradiction to this. Perhaps you should go and study what the AGW predictions really are before you talk about them.
Keith says:

I’m just waiting for the cries that this is only a regional trend, since this is from the U. S. National Climatic Data Center, not a true Global record rendering. Even more, that the cooling trend is only due to the last two years’ aberration , and ten years do not a climatic trend make.
Gavin? Tamino? I’ve stated the party line for you, and your cue has been said, so please stroll in from the wings now.

Yeah, why should we worry about details like statistical significance? It is much more fun just to find (cherrypick) data that fits our preconceptions regardless of such boring issues and then use it to validate our preconceptions.

Maybe a better discussion would be to haul out the hoaxes first 1988 IPCC model predictions and see how reality has followed the computer models. Afterall, we do have 20 years of tracking the hoax. Could be an interesting exercise. It would be OK to allow 5 year updates, as long as the older projections are included, so the IPCC predictions would better match reality — VBG.
Like with the NASA sunspot predictions, at some point even you are better at darts than the official predictors.

Dell Hunt, Jackson, Michigan

Also interesting, if you select December only in table format, Dec 2008 was the 35 coldest (80th warmest) December on record.
So if you thought it was colder than normal, you are right.
My heating bill agrees it was pretty cold in December 08.

L Nettles

“ak (11:38:04) :
@L Nettles, i would choose time span that didn’t start with the peak of a solar cycle and end at it’s trough. wouldn’t you agree?”
I personally think linear trends on cyclic data are what got us here. I expect a reversion to the mean.
I notice that you only criticized the selection and did not choose a starting point or defend your selection. If I had to choose a starting point it would be in the Medieval Warm Period, but we don’t have that data, any trend using only post 1850 data is useless.

terry46

Off topic but I just saw where Hathaway has changed the forecast for sun cycle 24 yet again.I just love this. When thing don’t go as projected just change your forecast .It’s the same with recording the temps .If it doesn’t fit the template then change the way you get your numbers .How many more years of egg on the face are the global warming crowd going to have to take it before they wake up?

King of Cool

Australia shows a very similar trend to the USA:
http://www.bom.gov.au/announcements/media_releases/climate/change/20090105.shtml
But the BOM still continues to write in terms that everything is on the up and up (and it was good old La Nina that temporarily stopped the rise in 2008).
Will be a very interesting to read their comments this time next year if 2009 continues to go down. No snowy owls sighted yet.

Ed Scott

Current Arctic Sea Ice Extent (2009/01/01)
http://www.socc.ca/seaice/seaice_current_e.cfm
The image / animation depicts recent Northern Hemisphere sea ice extent and concentration as estimated from the Special Sensor Microwave Imager (SSM/I) on board a U.S. Defence Meteorological Satellite. Sea ice concentrations are estimated from the 19.3 and 37 GHz channels of the sensor using the Canadian sea ice algorithm.

crosspatch

Tarpon:
Here is a graphic that compares observed temperatures (magenta) against the various IPCC predicted scenarios. You can probably find more by digging around in Icecap’s “cold storage” area.
As you can see, the not a single one of the IPCC forecast scenarios have been anywhere close to accurate. Even their “best case” scenario is 1 degree C high and the trend is incorrect in sign (they show a positive trend, observed temperatures show a negative trend).