Guest Post by Bob Tisdale
Monthly sea surface temperature data from NOAA (ERSST.v3b) are available online well before the global land air+sea surface temperature suppliers (NASA GISS and NOAA NCDC) publish their monthly updates. In addition to annual (January to December) data, GISS also presents what is known as the Meteorological Annual Mean data, using the months of December to November. So, in advance of the release of the GISS and NCDC combined products, let’s see how the global sea surface temperatures in 2014 compare to the previous record high back in 1998, on a Meteorological Annual Mean basis, using the dataset employed by both suppliers (NOAA’s ERSST.v3b). We’ll start the data presentations in 1997 to make the comparison easier, and I’ve also provided color-coded horizontal lines to show the 1998 (blue) and 2014 (red) values. Last, I’ve presented the data in absolute terms, not as anomalies, but that makes no difference.
QUICK NOTE FOR THOSE WHO ARE FIRST DISCOVERING THAT GLOBAL SEA SURFACE TEMPERATURES ARE AT RECORD-HIGH LEVELS: The record high global surface temperatures in 2014 are primarily a response to a prolonged weather event in the North Pacific. See the post here. Putting that in more familiar terms: You often hear meteorologists talking about a blocking high during a heat wave. So it might be easier to think of the record high sea surface temperatures globally being caused by a prolonged blocking high in the extratropical North Pacific, south of Alaska. [End note.]
As shown in Figure 1, global sea surface temperatures (based on NOAA’s ERSST.v3b dataset) from pole to pole were 0.02 deg C higher in (Dec to Nov) 2014 than they were in (Dec to Nov) 1998. That’s read two one-hundredths of a deg C. Two. All of the hubbub this year boils down to two one-hundredths of a deg C.
Figure 1
But GISS masks (excludes) sea surface temperature data in the polar oceans, anywhere sea ice has existed. And NCDC, as far as I know, excludes polar sea surface temperature data seasonally when there is sea ice in a grid. So let’s take a look at global ocean data excluding the polar oceans, Figure 2. That is, we’re looking at the data for the latitudes of 60S-60N. And as you can see, on a Meteorological Annual Mean basis, 2014 surpassed 1998 by only 0.02 deg C for those latitudes as well with the NOAA ERSST.v3b data.
Figure 2
WHAT ABOUT NOAA’S SATELLITE-ENHANCED SEA SURFACE TEMPERATURE DATASET, REYNOLDS OI.V2?
In 2008, NOAA introduced its ERSST.v3 data, which included satellite-data starting around 1982. NOAA quickly discovered the satellite-enhancements rearranged the record high years for their combined land+ocean data, making 1998 once again higher than one of the years in the early-to-mid 2000s. If memory serves, the differences were measured in a few hundredths of a deg C. But NOAA didn’t like that, so they removed the satellite-enhanced data from their long-term sea surface temperature reconstruction, making it a poorer-quality dataset, and renamed the dataset ERSST.v3b.
In 2013, GISS switched to NOAA’s ERSST.v3b sea surface temperature dataset from a combination of the UKMO HADISST data before the satellite era and Reynolds OI.v2 satellite-enhanced data starting in 1982. That of course increased the trend of the GISS data a small amount since 1900 and rearranged the highest temperature rankings in the 2000s a tiny amount in their GISS Land-Ocean Temperature Index (LOTI) data.
NOAA still supplies a satellite-enhanced sea surface temperature dataset. It runs from November 1981 to present. It’s known as Reynolds OI.v2, and it’s the dataset I use in my monthly sea surface temperature anomaly updates. (See the November 2014 update here.) It’s the basis for many of the daily sea surface temperature anomaly maps you see online.
[sarc on] Good thing both data suppliers went to all the trouble to change sea surface temperature datasets. On a “Meteorological Annual Mean” basis, the satellite-enhanced sea surface temperature data in 2014 was 0.01 deg C higher in 2014 than in 1998. They gained one one-hundredth of a deg C by excluding the better-quality satellite-enhanced data. Whew!! [sarc off]
See Figures 3 and 4 for the global “Meteorological Annual Mean” sea surface temperatures from 1997 to 2014, with and without the polar oceans, based on the Reynolds OI.v2 data.
Figure 3
# # #
Figure 4
In a few days, GISS and NCDC will publish their November 2014 values. It will be interesting to see where they fall…and also see how alarmists try to make it appear as though the end of the world is near.
And Ted Williams batted .406 in 1941 only because he had 6 hits in his final two games.
It doesn’t really matter what “caused” the record highs. It’s noteworthy, actually, that we saw record high temps in the absence of a large El Nino event, which we had in 1998.
For the proper perspective, you really need to put that in Hiroshimas.
Actually it has been warming for a long, long, long time. It is called an interglacial warming period. And it’s a good thing or we would all still be hunkered down in caves.
Is it just me or has HadCRUT4 become trendier in the history the last couple of months? Seems like the whole dataset tilted a bit.
You may wish to check out:
http://wattsupwiththat.com/2014/11/05/hadcrut4-adjustments-discovering-missing-data-or-reinterpreting-existing-data-now-includes-september-data/
and
http://wattsupwiththat.com/2014/10/05/is-wti-dead-and-hadcrut-adjusts-up-again-now-includes-august-data-except-for-hadcrut4-2-and-hadsst3/
Wish you had said “(Hot) Whopping…” in the title.
Hadsst3 just came out for November. The yearly average after 11 months is 0.481, which is 0.065 above the 12 month 1998 average of 0.416.
The previous high month of July 1998 with an anomaly of 0.526 was beaten in every month from June to October.
wbrozek, thanks, I’ll go take a look.
Is the ERSST.v3 dataset still avalaible somewhere ?
Simon, I don’t believe it is.
Bob Tisdale,
When the hidden heat comes busting out from ocean depths, any day now, not even your sarcasm will save the pause.
The pause will soon be replaced with a 5 degree surge.
Please graph that.
Seriously, Steve? How did the warming get there, what’s the process?
Repeal the the 2nd Law of Thermodynamics, first.
That’s after you’ve repealed gravity, so that warm water can lie under cold.
Steve Oregon, thanks. That made me laugh, considering I only presented global data in this post. Will that be a global 5 deg C surge? If so, how will that occur, when the subsurface temperatures of the oceans to depths of 2000 meters have only increased at a rate of 0.03 deg C over the last decade?
Hi Bob,
I was wondering if you had any thoughts on the strange coexistence (pointed out by Steve Goddard) on the unisys sst anomaly map of a +6c bright red between Greenland and Iceland with more extensive than normal ice pack, which indeed almost reaches Iceland. It is a bit odd isn’t it, especially coming after last months rumpus about massive -6 anomalies in mid Atlantic? I seem to remember you said last month you didn’t use unisys, is this why?
Steve
Hi Steve. First, I no longer pay close attention to daily sea surface temperature maps. Too much weather noise. But what Steve Goddard picked up looks odd, doesn’t it?
http://stevengoddard.wordpress.com/2014/12/10/something-is-very-rotten-in-denmark/
Here’s a different view from the same source of data:
http://polar.ncep.noaa.gov/sst/rtg_high_res/
I also don’t know enough about that specific sea surface temperature dataset that Unisys is using to comment either way.
I liked the old Unisys color coding of their maps, especially when viewing their animations. Then they ran into a problem with the new dataset and then changed the color coding. I haven’t been back.
Hi Bob,
Thanks for that. I’m interested you say ‘too much weather noise’ in the daily sst maps. I suppose I’d imagined the sst to be more or less free of noise, given that sea temperatures change more slowly than land. But if you’re trying to pick up the really long trends and anomalies then any short term changes are noise, especially if they produce the range of odd data as recently.
And If it weren’t for the contradiction of the ice pack, how would we know?
Steve
At first glance the dataset name “ERSST” reminded me of the German word ‘ersatz’ which means ‘fake’.
It is a bit fake, in the sense that it stands for “Extended Reconstructed Sea Surface Temperature”, and so uses “statistical methods” to reconstruct missing data.
From the NOAA website, where you can download the ERSST data in various formats:
That last sentence is surprising because it suggests that the satellite data may have been thrown out because it didn’t produce the rankings they wanted. Freudian slip? Who decides which rankings are correct?
Having said all that, I see nothing wrong in using data like this for research purposes. I view it as an attempt to harmonize historical data and render it in a seamless and uniform way. Of course it will contain errors, and they have provided an error estimate for each value (which we should plot and compare to the time series). So I think it could be useful for climate studies and such, but should not be interpreted as the “Gospel of Meteorological Truth Down to a Hundredth of a Degree”!
But that’s exactly what the MSM are doing with this info.
Here’s the link to the ERSST website:
http://www.ncdc.noaa.gov/data-access/marineocean-data/extended-reconstructed-sea-surface-temperature-ersst-v3b
… a paper which explains the layered approach that was used to create the datasets. And also a rationale (“cold bias”) for throwing out the satellite data, which still seems a bit ‘biased’ to me (because all data sources are ‘biased’ in one way or another).
http://www.ncdc.noaa.gov/sites/default/files/attachments/Improvements-NOAAs-Historical-Merged-land-Ocean-Temp-Analysis-1880-2006_0.pdf
All of these temperature measurements are, for the most part, made with resistance temperature devices(RTD) which use a calibrated platinum resistance element. These typically range from +-.3 degC error down to 0.0001degC for for a few thousand dollars. In any case, the averages, statistics, and the ERSST temperature reconstructions do not affect the error range of the instrument. The arithmetic doesn’t change the instrument error, which still applies to the final result.
In filling data with statistical methods does not supply any new data or information. It is simply done to make it easier to average a picture of the data in some way and increases the uncertainty in the results by an unknown amount.
“ERSST v3b does not include satellite data, which were found to cause a cold bias significant enough to change the rankings of months.” This statement shows their plain, old bias. It could just as easily mean that the ERSST statistics are causing a warm bias. It shows that the errors in the ERSST numbers are larger than they indicate, since they do not have the courtesy to show how big the changes would be.
There are a number of physical processes which can yield “thermometers” with extra-ordinary Temperature resolution.
Hewlett Packard used to make a quartz crystal thermometer instrument that was based on a totally unique quart crystal oriented “cut” that had a highly linear Temperature coefficient of frequency.
Many cuts, yield parabolic or cubic TCFs, but HP researchers, theoretically calculated this doubly rotated orientation that was very linear over a considerable range. I have no recollection of what the oscillation mode was, or the operating range. The thing is that such devices can make extremely small Temperature changes apparent.
Well whoopee !!
The problem in Temperature measurement, and / or control, is to know just what it is that, is at THAT Temperature.
Same thing goes for PRTs, which are anything but highly linear.
So we can’t expect extreme “accuracy” to formal Temperature scales, but we can expect good resolutions, and good repeatability.
But that is only for the Temperature of the thermometric element itself.
Who, or whatever else may be at that Temperature is where all the misery lies.
Thanks, Bob.
I think 0.02 deg C looks like no change at all, and the global warming is still stopped.
The most significant thing is not what is being observed, but what is NOT being observed. The climate models predict a CATASTROPHIC rise of up to 6 degrees per century.
What really bugs me is that it was very cold last winter and this winter in Indiana. To hear warmest year on record is really aggravating.
You mean coz it was cold at your house, it must have been cold everywhere? Mmmm, I think the planet is bigger than your back yard.
Hoo Boy!
If this rapid ocean heating and ‘acidification’ continue unabated, in 50,000 years or so it will be like simmering the clams and fishes in wine sauce!
Oh, the Humanity….
Bob,
Thanks for the outstanding post re sea surface temperature.
As an engineer it is obvious that hundredths of a degree is statistically not relevant. But that won’t stop the team from misleading the public, probably by not including the minuscule amount of the increase. Of course this is only effective with a complicit media that fails to challenge misinformation.
One question I have (I apologize if you have weighed in on this before). It is a simple concept that as ice forms, energy is released into the surroundings, either warming it or reducing the amount of cooling. How much of a factor is this in “warming” the surrounding warmer.
Enthalpy of fusion for (pure) water is 334 kJ/kg. You should be able to Google it for salt water.
Basically UAH and RSS datasets are set surface dataset, since over 70% of the surface is ocean. As time passes, HADSST3 and UAH/RSS are more and more of. I wonder why?
Could someone explain?
More False Precision Syndrome.
I just love it when these morons at NASAN NOAA etc. quote temperatures to degrees of precision literally orders of magnitude in excess of the discriminatory ability of their instruments.
Do they think everyone – even the non-scientifically trained layman – is that stupid?
Averages are not measurements. The denominator in an average is a “perfect number” and it’s precision is perfect.
Yes it is. But it isn’t the precision of the denominator which is at issue here. Nice try though.
Mr Davidmhoffer
..
This post should help you understand precision in averages.
..
http://wattsupwiththat.com/2014/12/12/meteorological-annual-mean-dec-nov-global-sea-surface-temperatures-set-a-record-high-in-2014-by-a-whopping/#comment-1812936
actually they do , but not just in climate ‘science’ they pull the trick of claiming an ability to measurement to levels of accuracy that is beyond the ability of the instruments they use to measure with . There is an idea that if you throw enough maths at it you can overcome the limitations of such physical realities and just add in error bars to cover your rear end.
Steve Goddard’s page showing sea surface temperatures around Greenland at +6C above average while anomalous ice is present demonstrates that the books are being cooked by the fanatics.
“Record temperatures” were on the cards at the start of year, regardless of what happened in the real world.
http://theconversation.com/are-we-heading-for-a-worrying-super-el-nino-26090
– – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Mark Harrigan
PhD Physicist
If it [a super-worrying El Nino] happens I confidently predict that those at the forefront of climate change denial currently falsely claiming “there has been no warming since 1998” will jump on the El Nino as a one off event and therefore “not to be counted” as an indicator that the planet continues to demonstrate all the signs of increased heat retention.*
– – – – – – – – – – – – – – – – – – – – – – – – – – – – –
* Great Lakes record ice cover, record duration, earliest recorded formation,
* Arctic ice extent 1.7million square kilometres more than 2 years ago
* Antarctic Sea Ice extent, highest on record
* Fall snow cover in Northern Hemisphere most extensive on record
The territory is cold while the map is warm. But the map is not the territory.
Reblogged this on Centinel2012 and commented:
And just think this is the best they they can com up with, to small to mean anything ; but what really significant is this value is about .5 degrees C lower than the average of the IPCC GCM computer models. Which means they are not even in the ball park, I’ve done some complex computer models and if one of my models performed this badly I’d be a bit embarrassed!
Here is a plot of the ERRST reconstruction going back to 1880, showing the NOAA calculated uncertainties:
http://i59.tinypic.com/15zihwl.jpg
The uncertainty of measurements prior to 1950 is understandable. But why does the uncertainty suddenly jump an order of magnitude or so, starting in 2000? That’s about plus-or-minus 0.1C! (And we’re quibbling about 0.01C nudges?) Yikes, looks like another attempt to draw a hockey stick.
Also, the above uncertainty plot is restricted to data between 60S and 60N latitudes. NOAA acknowledges even larger uncertainty in the Arctic and Antarctic ERRST reconstructions. (All of Greenland lies above 60N).
https://climatedataguide.ucar.edu/climate-data/sst-data-noaa-extended-reconstruction-ssts-version-3-ersstv3-3b
Hmm, I erred, it’s “ERSST”, not “ERRST”. (Errly in the morning here).
[No problem, thank you for noting the ERRER, it ist noted. 8<) .mod]
Such a tiny amount , despite so much effort going into finding a ‘increase ‘ shows one of two things , either their rubbish at it , quite possible given the amazingly poor scientific practice seen in this area or that despite their ‘best efforts ‘ the data cannot be tortured enough to get anything more. Either way although their have succeeded in getting the headlines they wanted this time, in the long term their claims are getting farer away from reality everyday.
Has it increased or hasn’t it?
Come on – this is not good enough. We need some dramatic headlines – like – Mann spotted some boiling sea water on the sands of San Jose.
Seems obvious “the pause” reduced glacial runoff which in turn decreased cooling, contributing to the supposed rise, however statistically insignificant or w/i the margin of error…
I wish we could lasso a number of these ARGO buoys, recalibrate them, and determine how much their sensors may have drifted, and how much larger is their measurement uncertainty after years at sea. When one begins to trumpet loudly changes of 0.01C we are in for a sanity check.