Biased Bob Tisdale is all at sea

Bob Tisdale is one of the pseudo-scientists who frequently writes articles for Anthony Watts' climate conspiracy blog, WUWT. His articles are often overly long, overly tedious and overly wrong. He is also a greenhouse effect denier. He thinks that it's the sun that's causing global warming through El Ninos and blobs. He's wrong.

Bob's what you'd might regard as a plodder. Not a quick-witted chap. He's not a research scientist and has never boasted of any educational qualifications. Where his expertise lies, if you can call it that, is in regurgitating the same mix of pseudo-science tinged now and again with real science - over and over and over again.

Bob's usual subject is ENSO, with occasional forays into sea surface temperature in general. Hunched over his keyboard, he relies on the years of difficult and arduous field work, clear thinking and clever, complicated analysis from scientists, but he despises their efforts in true denier fashion. He doesn't understand most of it so he cranks out his sunlight fueled nonsense and some deniers lap it up.

At one stage Bob even claimed to have made a new discovery. No-one bothered to set him straight - that he was several years and a vast body of science too late and, as well as that, he missed the point by a Pacific Decadal Oscillation. That's what happens when you don't read the literature. (Like his rival, Willis Eschenbach, thinking he was the first person to discover that thunderstorms carry heat aloft.)

Bob's airy battle with sea surface temperatures

Bob's latest beef is with the global sea surface temperature (archived here). He's got a bee in his bonnet about the new version of the Extended Reconstructed Sea Surface Temperature ERSST version 4. Bob has inadvertently or deliberately misrepresented the work - probably a bit of both.

Sea surface temperature is measured by observations taken on ships as well as buoys. (Also from satellites - but not ERSST v4.) The ships have changed over time as has the way temperature is taken - a lot. So to get a set of data that is consistent over decades, scientists have to make all sorts of correcting adjustments, using fancy algorithms.

As described in one of the main ERSSTv4 papers, Huang15, in ERSSTv4 there were two main bias adjustments. One to adjust bias in ship SSTs over time - using HadNMAT2; and one to adjust for bias between the ship and buoy SSTs. Bob Tisdale only focused on the former, which constitute only 10% of the measurements in the period post-1980.

Bob neither knows nor cares about all that. He's got only one thing on his mind. He wants to prove that scientists don't know nuffin.  And to try to best Tamino (as if!). He wrote:

NOAA used the night marine air temperature dataset (HADNMAT2) from the UK Met office as a reference for their new ERSST.v4 dataset. But the short-term warming rate of the new NOAA ERSST.v4 data during the global warming slowdown is much higher than the HADNMAT2 data. See Figure 1, which was first presented in my open letter to Tom Karl. That graph serves as the basis for my statements (1) that the recent update to the NOAA sea surface temperature data cannot be justified by the dataset that was used a reference for those adjustments, and, in turn, (2) that the NOAA adjustments are overcooked.

He's wrong. HadNMAT2 wasn't used as a reference in the way that Bob implies. It was only used for bias adjustments to ship sea surface temperatures. As stated in Liu15 (about ERSST v4):

... ship SST measurements are adjusted through reference to the most recent night time marine air temperatures (NMAT) from the Hadley Centre.

Air temperature helps determine ship bias

When the NOAA researchers set out to get an more accurate reconstruction of ship sea surface temperatures, they looked at the night time marine air temperature (not sea surface temperature) as reconstructed in a special data set, the Hadley Centre night time marine air temperature data set or HadNMAT2. That allowed them to correct for bias in ship data, which was the first step they took after quality control.

Observations from buoys are more accurate and more consistent than those from ships. So the scientists also looked to where ships and buoys were co-located, which enabled them to make further corrections to the data. Bob's article wasn't about this aspect. He was only concerned with the first lot of bias adjustments I mentioned. The ones to correct a bias in ship sea surface temperature by using HadNMAT.

It's complicated

Scientists have to make all sorts of calculations to find out how the sea surface temperature changes. You just have to read the papers describing ERSST v4 and how it compares with HadSST to get an appreciation of the detailed work that has gone into working it out. It's complicated. (Some of them are listed below.)

You'd not know how much work went into ERSST v4 if you only read WUWT. If you relied on quacks like Bob Tisdale, you'd probably think that getting sea surface temperatures is a piece of cake. You'd think that all you had to do was go to the Hadley Centre in the UK and ask them to give you the Hadley Centre Nighttime Marine Air Temperatures (NMAT) and that would be that. It wouldn't matter what the daytime marine air temperatures were. Nor would it matter what the temperature of the sea surface was. Nor that these night time observations can be quite sparse. All you'd have to do is go to KNMI Climate Explorer, grab some data (collected over decades, and developed through years of painstaking work by scientists) and plot the temperature of the air about ten metres or so above the sea surface and you'd have your answer.

Correcting for bias in ERSST v4

Thing is, you'd be as wrong as Bob Tisdale. Bob thinks that every observation in ERSST v4 was adjusted to match HadNMAT2. Or he thinks that they should have been. He's wrong of course. There are papers describing how NMAT was used. It was used to work out the bias in ship observations, not all the observations combined.

The passage below is from a paper by Huang et al (2015) describing how ERSST v4 was constructed. It describes how ship data and buoy data were adjusted separately:

The SSTs from ships or buoys were accepted (rejected) under a QC criterion that observed SSTs differ from the first-guess SST [sea surface temperature] from ERSST.v3b by less (more) than 4 times standard deviation (STD) of SST (Smith and Reynolds 2003).

The ship and buoy SSTs that have passed QC were then converted into SSTAs [sea surface temperature anomalies] by subtracting the SST climatology (1971–2000) at their in situ locations in monthly resolution. The ship SSTA was adjusted based on the NMAT comparators; buoy SSTA was adjusted by a mean difference of 0.12°C between ship and buoy observations (section 5). The ship and buoy SSTAs were merged and bin-averaged into monthly ‘‘superobservations’’ on a 2° x 2° grid. 

As for why HadNMAT was used, the same paper explains:

The latest release of Hadley NMAT version 2 (HadNMAT2) from 1856 to 2010 (Kent et al. 2013) provided better quality-controlled NMAT, which includes adjustments for increased ship deck height, removal of artifacts, and increased spatial coverage due to added records. These NMAT data are better suited to identifying SST biases in ERSST, and therefore the bias adjustments in ERSST version 4 (ERSST.v4) have been estimated throughout the period of record instead of exclusively to account for pre-1941 biases as in v3b.

Bob is comparing the wrong things - his data is largely from buoys not ships

Thing is, too, that Bob was writing about the so-called "hiatus". Now in recent times, since about 1980, ship measurements are only about 10% of the data. From Huang15:

Since 1980 the global marine observations have gone from a mix of roughly 10% buoys and 90% ship-based measurements to 90% buoys and 10% ship measurements (Kennedy et al. 2011).

[Quote added by Sou 27 July 2015, since RomanM thinks I'm "out to lunch"] 

So that means that the HadMNAT2 adjustments only applied to 10% of the observations. The other 90% of data comes from buoys. So when Bob puts up his chart showing that NMAT observations have a different slope from the slope of ERSST v4, he's not demonstrating that ERSST v4 is wrong. In the period that Bob was most interested in, only about 10% of the observations are affected by the bias adjustment with HadNMAT.

There's more. I was advised by one of the researchers that where both ships and buoys exist, the algorithms give buoys more than six times the weighting given to ship data. That means that over the last twenty thirty five years of ERSST v4, the results are dominated by data from buoys.

That's still only part of the story

The researchers considered other comparisons, including UKMONMAT.  In their Part 2 paper by Liu et al (2015), the authors explain that if instead of HadNMAT, they'd used UKMONMAT, then there would have been a stronger warming in a number of regions - east of Greenland, over the Greenland Sea and the Barents Sea, in the Southern Ocean, and several marginal seas around Japan and the Bering Strait. They illustrated the difference in Figure 4 C below:

FIG. 4. Maps of differences of the SST warming trend (1910–2012) between individual SPP runs and the ERSST.v4 operational run. (a)–(n) Results from the following SPP runs: (c)UKMONMATbias adjustment. Source: Liu15

What that illustrates is that Bob is wrong if he thinks that the bias adjustment algorithm (using HadNMAT2) has an equal effect across the board. Obviously it doesn't, or there wouldn't have been these regional differences appearing when a different set of marine air temperatures was used for bias adjustment. When the HadNMAT2 bias correction is applied there is a ripple effect that can have different impacts in different regions.

Bob and his silly graphs claiming 0.04C/decade is "much different"!

Bob plonks up this quickie chart to "prove" that he knew better than all the scientists who've been studying sea surface temperature for decades:

First of all let's clear things up. Bob's ERSSTv4 line has almost no influence from the HadNMAT bias adjusting algorithm. That's because most of the data is from buoys, not ships. And as I wrote above, where ships and buoys both exist, the buoys have six times more weighting. So he's not showing what he thinks he's showing. If he could extract it, he should be comparing the less than 10% of the data that comes from ships with his HadNMAT plot, not the full ERSSTv4.

Put that to one side, what about his claim that the two charts are not a good match. Well I disagree. Not only are the values very closely matched (the lines overlap for most of the period he shows), but the difference in the trend is minor. Bob calls HadNMAT2 a "much lower rate" - yeah, all of 0.04C/decade difference!

As well as that his rates would be strongly affected by the end points.  To illustrate, here is the same chart with the period extended to 1990 to 2015 - click to enlarge it:

Data source: KNMI Climate Explorer

If you are wondering if it makes a difference that the data for HadNMAT2 only goes to December 2010 in KNMI, well not really. In both cases the difference is marginal, only 0.001°C/decade:

Data source: KNMI Climate Explorer

And below is the chart from 1950 to 2015. I've added in Reynolds OI v2, Bob's favourite SST, as a bonus. HadNMAT has almost the same slope as ERSST v4 going back to 1950. Reynolds is slightly less - a difference of only 0.001C/decade!

In his chart above, Bob claimed (again) that HadNMAT2 was used as "the reference" for ERSST v4. It wasn't. It was used to correct bias in ship sea surface temperatures only, around 10% of the measurements over the period that Bob's plotted. The other 90% of measurements came from the more accurate buoys.

Then he blithely proclaims that:

Bottom line for the NOAA adjustments: I can support with data my statements about the NOAA ERSST.v4 sea surface temperature-based global temperature products being unjustifiably overcooked.

That's it. That's the sum total of Bob's rebuttal. A simplistic chart and a few words saying they are wrong. No analysis. Clearly no understanding of how HadNMAT was used or why. Nor any understanding of ERSST v4 itself.

Seeing the period in question has almost no influence from HadNMAT, Bob's chart is not showing what he thinks it does. What the HadNMAT is showing is that the night time marine air temperatures as measured on ships are not inconsistent with sea surface temperatures as measured mostly by buoys in the period that Bob looked at.

What is ironic is that some of the work that underpins ERSST v4 and HadNMAT was by one of the authors of Bob's favourite sea surface temperature data sets (Reynolds OI v2), Richard W Reynolds. He and Thomas M. Smith co-wrote a number of papers on sea surface temperatures. The team that developed ERSST v4 said the methodology follows that described in a 2003 paper of Smith and Reynolds.

If you read any of the papers listed below, you'll quickly realise the ridiculous in Bob's weak attempt to "rebut" the scientists. Bob plonks up a five minute chart that he doesn't understand, about work that he could never hope to understand, and proclaims that the scientists are fraudulent or, in his words, their "data are unjustifiably overcooked". In stark contrast, the work by scientists is based years of work and very detailed analysis of observations taken over 150 years or more. It's far from a simple matter of plonking up a couple of charts. The scientists' work is rigorous, very complex and a galaxy away from Bob's ridiculously simplistic attempt to refute it.

From the WUWT comments

This is a long article so I'll just post one choice comment:

wickedwenchfan doesn't care about bucket materials, insulated vs not insulated, deck heights, engine room intake readings, or any of that trivial stuff that concerns scientists, all they want is the raw data.

July 20, 2015 at 6:36 am
Let’s face it, data should be data. Full stop. End of story. No excuses. No adjustments.
Once you adjust data, even once, for whatever reason, we are no longer doing science. We are doing opinion and personal bias.
Put everything back as it was. Show on graphs where instruments changed, or where airports were built, or where orbits declined, etc, etc, but let the data speak for itself and leave everyone the opportunity to interpret it as is.

References and further reading

Huang, Boyin, Viva F. Banzon, Eric Freeman, Jay Lawrimore, Wei Liu, Thomas C. Peterson, Thomas M. Smith, Peter W. Thorne, Scott D. Woodruff, and Huai-Min Zhang. "Extended Reconstructed Sea Surface Temperature Version 4 (ERSST. v4). Part I: Upgrades and Intercomparisons." Journal of Climate 28, no. 3 (2015): 911-930. doi: http://dx.doi.org/10.1175/JCLI-D-14-00006.1 (pdf here)

Liu, Wei, Boyin Huang, Peter W. Thorne, Viva F. Banzon, Huai-Min Zhang, Eric Freeman, Jay Lawrimore, Thomas C. Peterson, Thomas M. Smith, and Scott D. Woodruff. "Extended Reconstructed Sea Surface Temperature Version 4 (ERSST. v4): Part II. Parametric and Structural Uncertainty Estimations." Journal of Climate 28, no. 3 (2015): 931-951. doi: http://dx.doi.org/10.1175/JCLI-D-14-00007.1 (pdf here)

Smith, Thomas M., and Richard W. Reynolds. "Bias corrections for historical sea surface temperatures based on marine air temperatures." Journal of Climate 15, no. 1 (2002): 73-87. doi: http://dx.doi.org/10.1175/1520-0442(2002)015<0073:BCFHSS>2.0.CO;2 (open access)

Smith, Thomas M., and Richard W. Reynolds. "Extended reconstruction of global sea surface temperatures based on COADS data (1854-1997)." Journal of Climate 16, no. 10 (2003): 1495-1510. doi: http://dx.doi.org/10.1175/1520-0442-16.10.1495 (open access)

Parker, D. E., C. K. Folland, and M. Jackson. "Marine surface temperature: observed variations and data requirements." Climatic Change 31, no. 2-4 (1995): 559-600. (Web version here)

Kent, Elizabeth C., Nick A. Rayner, David I. Berry, Michael Saunby, Bengamin I. Moat, John J. Kennedy, and David E. Parker. "Global analysis of night marine air temperature and its uncertainty since 1880: The HadNMAT2 data set." Journal of Geophysical Research: Atmospheres 118, no. 3 (2013): 1281-1298. doi:10.1002/jgrd.50152, 2013 (open access)

Kennedy, J. J., N. A. Rayner, R. O. Smith, D. E. Parker, and M. Saunby, 2011: "Reassessing biases and other uncertainties in sea surface temperature observations measured in situ since 1850: 2. Biases and homogenization." J. Geophys. Res., 116, D14104, doi:10.1029/2010JD015220. (open access)

Tamino has an article, too - Fundamental Differences between Bob Tisdale and Reality

This HotWhopper article describes the NOAA paper that prompted the frenzy of denial:
NOAA: No pause in the global surface temperature, with more references and further reading, and more on the denier protests