People send me stuff. Yesterday I got a note suggesting I have a look at what NOAA/NCDC’s “climate at a glance” was showing for trends in the 21st century so far.
I decided to take a look.
Have a look at NOAA’s Time Series calculator
http://www.ncdc.noaa.gov/cag/time-series/global
It is now displaying a cooling trend commencing in 2001 – 2013. Ensure you are on the Global tab; Annual; 2001;2013;Land and Ocean. Then in the Options Tab click; Display Trend; per century; 2001;2013. Then click plot. These result give you a -0.05 per/century over 13 years.
-0.05 is hardly significant (even though they claim +0.05 of 1 degree over a two month period of May and June this year proves global warming)
I verified that,
…and did my own.
This plot mostly matches what he says, though I prefer doing decadal scale trends on decadal scale data plots:
Fig 1. Source: http://www.ncdc.noaa.gov/cag/time-series/global/globe/land_ocean/ytd/12/2001-2013?trend=true&trend_base=10&firsttrendyear=2001&lasttrendyear=2013
(IMPORTANT NOTE: NCDC’s link generator on their web page creates a pre-broken link, so if you use the source links I provide from NCDC, be sure to manually set it to Annual from the default Year-to-Date and press plot again, otherwise you’ll end up with an incorrect plot.)
The trend is -0.01C/decade, essentially flat, no statistically significant trend. And if you want to make that a nice tidy package for the 21st century new millenium, the 2000-2013 trend is nearly equally statistically insignificant, and would be flat except for the fact that the year 2000 was a bit cool. It’s the typical problem of trend line sensitivity to endpoints on short data sets.
Fig 2. Source: http://www.ncdc.noaa.gov/cag/time-series/global/globe/land_ocean/ytd/12/2000-2013?trend=true&trend_base=10&firsttrendyear=2000&lasttrendyear=2013
But the lack of a trend on the Land Ocean Temperature Index plots isn’t what I find most interesting or significant – the difference between land and ocean is more interesting.
First the oceans in the 21st century:
Fig 3. Source: http://www.ncdc.noaa.gov/cag/time-series/global/globe/ocean/ytd/12/2000-2013?trend=true&trend_base=10&firsttrendyear=2000&lasttrendyear=2013
With only an insignificant +0.01C/decade trend, it seems Trenberth’s missing heat is still missing, and the oceans have stubbornly refused to play out the role that CO2 crunching models have prescribed. I suppose I just can’t get all that excited even though there’s a lot of squawking about the month of June being smashingly record-warm in the oceans:
The record was driven largely by warmer than normal ocean surfaces. Last month saw the highest temperatures on the water for any June on record, and the highest departure from the average for any single month ever. Average global land surface temperatures for June 2014 were also the seventh hottest June ever recorded.
Well, gosh, 2014 isn’t over yet, and we’ve been told time and again that a single month of anomalously low temperature means nothing in the scheme of climate things, and so it must go for a single month of high temperatures.
But, here is what I find most interesting, note the difference in trend from Figure 2 which is land+ocean index (LOTI) and Figure 3 which is just ocean (OTI) below. Have a look at the same period for land (LTI), which has a rate +0.13C/decade or 13 times higher than the ocean index in the same period:
Fig 4. Source: http://www.ncdc.noaa.gov/cag/time-series/global/globe/land/ytd/12/2000-2013?trend=true&trend_base=10&firsttrendyear=2000&lasttrendyear=2013
This difference between land and ocean trends is quite large, and some divergence would be expected, since the oceans affect the atmosphere above them far more than land as a stabilizing heat sink.
But, it seems in the USA, the Land Temperature Index isn’t cooperating with expectations or even warming at all. It seems the USA has been cooling in the 21st century at a rate of -0.09F/decade (-0.05C/decade):
Fig5. Source: http://www.ncdc.noaa.gov/cag/time-series/us/110/00/tavg/ytd/12/2000-2013?base_prd=true&firstbaseyear=1901&lastbaseyear=2000&trend=true&trend_base=100&firsttrendyear=2000&lasttrendyear=2013
It seems that that oceans aren’t warming, the contiguous USA isn’t warming, but the land surface of the rest of the world has been so far in the 21st century.
Meanwhile, MLO annual data shows carbon dioxide has risen from 369.52 ppm in the year 2000, to 396.48 in 2013, an increase of ~ 7.3%, but we don’t see a corresponding increase in global temperature for the same period perhaps because climate is a non-linear system and/or because we are close to saturation of the logarithmic effect of CO2 induced warming in our atmosphere. Global temperature has been mostly flat. Where’s those posited warming climate feedbacks when we need them?
Now, to alleviate the inevitable screams of not showing the “full picture” of temperature from the overly excitable that comment here under a variety of nom-de-plumes, I offer the entire LOTI plot from NCDC:
Fig 6. Source: http://www.ncdc.noaa.gov/cag/time-series/global/globe/land_ocean/ytd/12/1880-2013?trend=true&trend_base=100&firsttrendyear=1880&lasttrendyear=2013
To my eye, I see a natural sine wave, which I’ve traced below on the same graph in solid grey, with extrapolated segments in dashed grey:
Fig 7. Source: http://www.ncdc.noaa.gov/cag/time-series/global/globe/land_ocean/ytd/12/1880-2013?trend=true&trend_base=100&firsttrendyear=1880&lasttrendyear=2013 plus hand drawn sine wave from the author.
It seems to me that our current “pause” might simply be that we are at the top of that sine wave I see, and that we might actually see some cooling ahead, assuming it isn’t all adjusted away by the next “improvement” from NCDC.
I’ll leave you all to the squabble which will surely follow.
The problem with fitting a sine curve to the data is that it could be extended backwards as well as forward. Unfortunately doing so would eliminate most of the LIA. No doubt you would therefore say that extrapolating backwards is invalid, but if extending back is wrong why is projecting forwards right?
jimmi_the_dalek:
You ask in your post at July 27, 2014 at 11:28 am which says in total
Please don’t pretend to know what others would say, especially when your red-herring asserts that none would do what some do; for example, Akasofu’s fit can be seen e.g. here.
Richard
JJ says:
July 27, 2014 at 8:39 am
JJ, this was an excellent rebuttal to Rex and others who believe the current calendar convention (which is based on a calculated birth year of Jesus, and thus is correctly designated AD) began with year 0.
I just add 1 (2?) point:
Rex says:
July 26, 2014 at 4:49 pm
“You count your millenia as starting with 1AD … I count mine as
starting a year earlier. ”
Then, Rex, you have chosen to use your own private convention (noted by JJ). To be consistent, you should also call the 1st day of each month “0”, the 2nd day should be called “1”, Etc.
The point of a convention is to reduce confusion in communication. Does yours do that? Even if your convention was actually more internally coherent, it would still need to be universally adopted to be useful.
SR
according to GISS the trend for 2000-2013 was 0.08 C and for 2001-2013 is was 0.01 C. so whenever you decide to start the 21st century, the trend is basically the same
JJ says: July 27, 2014 at 8:39 am
JJ Thanks for a very interesting post.
richardscourtney says: July 27, 2014 at 6:53 am
Richard
To put some context to my comment, my father was a mathematics teacher. His view was that mathematics is the pre-eminent science and that mathematical advances were often made well in advance of any conceivable application for the discovery. In his opinion the discovery of the number zero was pivotal to the development of science. One example he gave of mathematical advancement was the challenge to the assumption that the internal angles of a triangle always sum to 180 degrees. If we suppose that the internal angles of a triangle can sum to more than 180, then it turns out that we have developed spherical geometry where, for example, a triangle composed of three great circles can have internal angles that sum to 270 degrees (3 right angles).
As a non-mathematician my concern was that mathematicians appear to invent new numbers, whereas he was very clear that mathematicians discover new numbers. We had many interesting discussions about this point. At school my biggest maths concern was the identification of the number e, the base of the natural logarithm. It was my inability to understand how the value of e was derived that caused me the greatest difficulty in trusting the validity of calculus as a mathematical process.
Years later I read of a definition of e that is both rigorous and intellectually satisfying. The natural number e is defined as the point on the X-axis where the space beneath the curve Y=1/X and the X axis, between the lines X=1 and X=e, has an area of 1. It has always been my view that an expert can be identified as the person who is capable of explaining a concept. If they can’t explain it to you then they are not an expert.
For more comments on the history of zero see here and here.
The sine curve is amusing but a 200 year cycle is a bit mischievous.
I prefer Akasofu’s sine curve fit, too. As per Richardscourtney (above)
If you remove the linear trend out of the little Ice Age and sort out the real cyclic pattern …
But then you have to allow for a huge crank up of of current temperatures and a huge drop in the old data Garbage in Garbage out Who knows what the real trend is?
But yes – There is a PAUSE.
“I’ll leave you all to the squabble which will surely follow.”
There’s nothing sure about what the ups and downs will be year-to-year, but seeing as reasonable skeptics eschew Dragon-slayer ‘physics’, we can expect temps to go up over the longer term as long as GHGs accumulate in the atmosphere. How much is what is in question. There’s nothing sure about the sine wave continuing, either. One supposed cycle does not a pattern make.
I’ve been quietly spectating the climate change saga for the past couple weeks. Before then I must admit I was largely unaware and blissfully so. The correlation between the news and scientific claims can be alarming: specifically the connection to the fires in the Northwest Territory, now reported to be the largest in 10,000 years.
An upward trend until recently seems irrefutable, but here the world sits on some plateau. Whether the temperature will continue up or down, I’m finding harder to rationalize, at least through Internet searches. There are times people seem to be pulling from completely different data sets, though each has it’s own credibility.
Today, I stumbled on the solar variability line of reasoning fronted by many skeptics. Solar seems to be the strongest case against a significant human signal on temperature, and it has received credence from Russian astrophysicist and NASA alike. http://sppiblog.org/news/russian-scientists-we-could-face-cooling-period-for-200-250-years
In a way, I very much want the sun’s influence to be the biggest factor. But I’m not overly optimistic that it’s the case, or the ramifications would be any less difficult to deal with considering the burgeoning population.
…It’s and interesting, and sometimes worrisome, time to be alive. I’m sure I’m not the first person born this generation or any of those previous to say so.
We’re used to thinking of blocks of years like decades and centuries as starting with zero. We’re used to thinking of “the Fifties” as starting in 1950, and “the noughties” as starting in 2000. Further, when we speak of a person as being in his twenties, we again think of the period as starting with a zero.
For consistency with that ingrained habit, it would reduce confusion if we adopted the convention that “the 20th century” started in 1900 (on a zero) and that the 2nd century began at the start of the year 100. (Leaving the 1st “century” with only 99 years. So what if it’s a shorty?)
Sine wave? What mechanism do you propose with that period? Do you think you can fit a sine wave to a half-cycle of data? LOL!
Actually the scale is shown at 1 degree Fahrenheight
[Marked in degrees of Fearandhype ? .mod]
You know, I’m just kind of DENSE (Pb)…You see, when I run these calcs for 1996 to 2014, I get.02 degrees C increase per decade for the Ocean temps. When I run it for the land temps I get 0.10 degrees C per decade. AND when I run the COMBINED, I get 0.04 degrees C per decade. However, if I use the 1/7 that the LAND represents of the Earth’s surface, and multiply the 0.02 C/decade by 6/7, and likewise the 0.10 Degrees C by 1/7 (for the ocean metric), and ADD the two I get 0.31 degrees C for the combined value. Whereas the web page gives me 0.40…or about a 30% error in that rather straight forward linear combination.
As I said, I’m dense (Pb)…and maybe I don’t understand this “high level” math. Perhaps instead of my simple linear thinking, I should be looking for the Poles in the complex plot of this data, circle them with a closed boundary (the Bromwhich Contour) and do the complex integral to figure out the residuals and AMAZINGLY I’ll get that 0.40 C…or then again, maybe not.
excellent analysis. Please have a look at my MACD analysis which finds a ~60yr oscillation:
http://wattsupwiththat.com/2013/10/01/if-climate-data-were-a-stock-now-would-be-the-time-to-sell/
1879 to 1911: -0.0076°C/yr, R2=0.18 (stat. sign.)
1911 to 1945: +0.0141°C/yr, R2=0.52 (stat. sign.)
1945 to 1976: -0.0020°C/yr, R2=0.02 (stat. not sign.)
1976 to 2007: +0.0193°C/yr, R2=0.64 (stat. sign.)
This might be a smaller oscillation within a larger (your grey line in Fig 7. suggests a 100-120 yr oscillation).
The ~60yr oscillation has been described extensively elsewhere as (Scafetta 2014, The complex planetary synchronization structure of the solar system) . Moreover, Scafetta (2014) also describe a ~115yr oscillation. I find this very interesting and well-fitting with the data.