Guest Post by Willis Eschenbach [See update at the end]
Since publishing my last two posts here and here on the Church and White (“C&W”) 2011 sea level dataset, some folks have queried why I didn’t use the Church and White 2013 dataset instead. The answer is simple. It’s because of the hockeystick.
What hockeystick, you might ask? Why, the C&W hockeystick … the figure below shows the difference between the C&W 2013 and the C&W 2011 data.
Figure 1. Red line shows the difference between the sea levels of the two C&W datasets. Larger values show where the 2013 dataset has higher sea levels than the 2011 dataset.
Now, I can understand the slight, ~ ± 1 mm differences between the two datasets in most of the record. They’ve recalculated with a slightly different algorithm, or maybe used a somewhat different mix of tide gauges and weights, with resulting small differences.
And I can understand the offset in most of the record, where the 2013 data is generally about 2 mm less than the 2011 data. It’s an anomaly so the zero point doesn’t matter.
What I don’t understand is the blade of the hockeystick from about 1992, the start of the satellite sea level era, to 2009 when the C&W 2011 record ended.
It looks to me like they’ve munged the data in order to make the data take up a smooth transition from the smaller acceleration of the tidal gauge record to the larger acceleration of the satellite record. That way, they could claim that the tide gauges and the satellite data agree … and if that is the case, I can only say bad scientists, no cookies …
Anyhow, that’s the only explanation I can think for this hockeystick … suggestions gladly accepted.
[UPDATE] I thought I might point out the biggest problem we have with sea level estimates, which is the lack of data. Here’s a graph showing how many tide gauges we have by the number of years of data that they have:
So for example, there are only 35 tide gauges which have 120 years or more of data … and that only gets us back to about the year 1900.
Not only that, but because the tide gauges are on land, the world’s oceans are only sampled around the edges and on a few islands …
Here’s another look, this time at how many gauges we have going at any given time:
Note that at the maximum, only half of the tide gauges in the historical record were in operation.
I bring this up to point out that we simply cannot place much faith in the tidal gauge data to provide any “global” average sea level rise.
Here, amazingly, we’ve had rain again. It hardly ever rains this late in the spring in Northern California, and the redwood forest around our house is redolent with the earthy odors of new growth and old decay … I blame global warming. After all, global warming has been blamed for a host of bad things from low birth rates to volcanic eruptions, so why not balance the scales by blaming it for a lovely gentle late spring rain?
Best of this wonderful wet world to you and yours,
w.
MY USUAL REQUEST: When you comment, I ask, beg, implore, request, and importune you to QUOTE THE EXACT WORDS THAT YOU ARE DISCUSSING so we can all understand who and what you are talking about.
DATA: The C&W data is here, h/t to Nick Stokes for finding it.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
I suspect we have tide gauges everywhere it actually matters what the sea level is. While it’s technically interesting to measure sea level globally, what does it actually matter what the level is in the middle of the ocean? The fish don’t care and I suspect even they can’t notice if it goes up or down by a few meters or even a few tens of meters.
In terms of consequences of sea level rise that justify spending any money, the only thing that matters is level relative to significant human structures and navigable waterways. Which I strongly suspect is exactly where the tide gauges are. Funny about that.
Everything that everybody said, ………. OR
8 mm
Getting WordPress to post that is a better trick then their sea level graft
Did anyone happen to see this:
“We are currently attempting to more thoroughly evaluate the methodology and to improve the reconstruction so that it better represents the variability (Legresy et al., Workshop on Global and Regional Sea Level variability and change, Mallorca, Spain, June 2015).”
It’s right under the link to the 2013 data
http://www.cmar.csiro.au/sealevel/sl_data_cmar.html
Seems like they would have been done with it by now, but who knows? Forgot to update the site? I don’t presume to know. Maybe if there are doubts, one should ask one of the researchers, or use a different dataset. Of course, that means passing up a chance to imply wrongdoing. This thread is chock full of guesses and assertions about the ways the scientists have committed fraud. If there’s a perfectly good explanation, who cares? As long as it looks like corruption, that’s all that counts – it’s worth posting. Find enough examples of things you don’t understand, and you can build a case to make it seem like the whole dam* industry is full of low-life scumbags. It doesn’t matter if the examples later turn out to have valid explanations – the goal is to instill the idea of misconduct, corruption and falsity, not to reveal the truth.
These days the ends justify the means, apparently. Saving conservatives from having to compromise on energy policy is worth destroying the public trust in science. Saving liberals from having to question their values and admit their arrogance is worth alienating half of Americans and dividing the nation.
It’s politics. The Swamp is not confined to D.C., it’s in us all. It’s power, entitlement, hypocrisy and tribalism. It’s ideology triumphing over pragmatism, bias over reason. It’s the human characteristics that must be acknowledged and tamed in order to live productively together in today’s world. We did not evolve in the environment we’ve created, and we need to put effort and thought into making it work socially as well as logistically. If it means anything to call ourselves Americans, it necessarily means recognizing that we are ALL part of this group called Americans, and as wonderfully diverse as we are, we all have something to contribute. It’s up to us whether we want to contribute destructively or constructively.
Oh, dear! Rambling again. Sorry.
(P.S. I know there are many who post here who aren’t American. I don’t mean to exclude you, and I’m not in favor of American exceptionalism, but I love my country.)
I find it mildly amusing that warmists like satellite sea level measurements because they show acceleration when grafted onto the tide gauge data. But they really don’t like satellite measurements of mid-troposphere temperature because they support the post-1998 “hiatus”.
It’s reminiscent of the way we skeptics like satellite temperatures but find good reasons to not trust satellite sea level measurements.
The difference, of course, is that we are right and they are wrong. How could it be otherwise?
Somehow the /sarc tag fell off.
Smart Rock: You are vastly overgeneralizing by treating all satellite measurements alike. Satellites are mere platforms. It is crucial to distinguish which instruments they are carrying, what those instruments are trying to measure, what the capabilities of the satellite platforms are relevant to the needs of the instruments and the intended measurements. And unlike you, I actually have years of experience designing and managing all the above–at NASA.
Thanks, Tom. I’m always amazed at the experience of the various people that read this blog.
The problem that I have is that the satellites are at an altitude of ~ 500 km, and they are trying to measure the ocean surface to an accuracy on the order of ± a half of a millimetre or so, in order to give answers to the nearest mm …
This means that it is attempting to measure the distance to a constantly moving surface to one part per billion (1E+9). This type of accuracy, one part per billion, is difficult to maintain over time in a laboratory, much less in space … not impossible, you understand, but very difficult. Instrument drift, orbital drift, exact position measurement, thermal expansion, ocean wave heights, there are lots of things that have to be known to extreme accuracy for it all to work. That sucker is doing one complete orbit of the earth every 56 minutes, which means if the timing is off by one second there’s an 11 km error in the location. And of course the orbit is not perfectly circular, so you have to constantly know the instantaneous altitude of the satellite at all times to something like a tenth of a mm or so …
Not impossible to do, as I said … just very, very difficult.
As a result, I trust the satellite data as to precision much more than I trust it as to accuracy … which means that e.g. the 10-day snapshots of the entire globe do show the relative heights in different parts of the planet, but the longer-term data that makes up the decadal and longer trends are much less trustworthy.
Best regards, any comments gladly accepted,
w.
Willis: Your arguments from ignorance and incredulity fail against knowledge. I suggest you take an inferential statistics class, paying special attention to the topics of populations versus samples. Read and really try to understand Nick Stokes’s comments here and on his blog. Read the actual academic papers before you criticize. Read spacecraft, space mission, and engineering and scientific project official proposals and reports. When you do not understand something in any of those, read the references on that thing, and if you still don’t understand, read a relevant textbook. Yes, doing all that will take you full days, weeks, months, or years for some topics. That’s how the people who actually do that work you criticize, actually plan and do their work. Space is hard. Science is hard.
Tom Dayton, you need to be more specific for your criticism to be valid. Make your argument otherwise your comment is meaningless.
Tom Dayton May 27, 2018 at 9:27 am Edit
Tom, you do realize that your entire response to me is an argument from ignorance and incredulity, right?
What I said, in essence, was that measuring something like the ocean surface levels reliably and repeatably from 500 km away to one part per billion over decades is a very difficult technical problem. I listed a host of specific problems that need to be overcome. The difficulty has been proven and highlighted by the large number of revisions in the satellite sea level data over time. Heck, Church and White in the link Nick Stokes just gave are claiming that there are still significant errors in the results. And that doesn’t count the unknown unknowns.
I’m not arguing from ignorance. I’m arguing from knowledge as well as the actual experience of people trying to do the measurements and having a very difficult time doing it.
As to “paying special attention to the topics of populations versus samples”, I suspect that I understand them as well as most do, perhaps better. I know I understand them better than many professional scientists, because I’ve seen their laughable errors.
But since all you’ve done is handwave rather than making specific
objections, so far we don’t know … in that regard you might enjoy my post entitled Decimals of Precision.
M.W. Plia May 27, 2018 at 1:58 pm
What he said …
w.
Willis,
Now, I can understand the slight, ~ ± 1 mm differences between the two datasets in most of the record.
I can’t, because those differences don’t exist when comparing the real data. Between 1880 and 1989 the difference is exactly -1.6. Now, should I immediately accuse you of intentional dishonesty and fraud as you have? Or should I first propose a more likely reasonable explanation – that you’re still comparing with a digitisation rather than the real numbers, which is introducing errors but no real intentional bias. There is no change to the algorithm or in the mix of tide gauges, which is obvious because there’s actually no change at all over that time.
The reason why you get increasing differences (both positive and negative) closer to the present has already been explained to you: Additional tide gauge data made available after the 2011 paper was published. Church and White provide a list of tide gauges used in the paper in the link you already have. If you look down that list you’ll find that many of the locations do not have data up to 2009. In many cases this is not because no data had recorded at those locations, but because the recorded data had not yet been reported/logged to the PSMSL repository used by Church and White. And of course, the closer to the present the more likely it is not to have been reported yet, which is why no changes prior to 1990.
This can also be plainly seen by looking at how the uncertainty estimate increases for recent years in the original data, primarily due to coverage deficiencies. By the 2015 update, much of the deficiency had been filled in by those reports coming in, so that the 2009 annual uncertainty estimate substantially tightened. Whereas we can see that the 2013 annual estimate (the final year in the update reconstruction) has about the same uncertainty as 2009 in the original, because it too is affected by reporting lag.
Thanks, Paul. You’re right about the small ± 1 mm errors, they are indeed from the digitizing. Makes me happy to see that my digitizing is consistently that good.
Regarding the rest, it may just be that the few new tide gauges that they added are enough to change the trend of the other 750 gauges in the dataset by a very large amount in the direction of greater acceleration … but given that C&W have nailed their colors to the mast of the good ship “ACCELERATION!”, it may not be …
I’m also not seeing how adding even 10 or 20 new datasets could change the sea level by 8 mm in a decade.
You say they explain it “in the paper in the link [I] already have” … if you’d be kind enough to repost that link, I’ll look at it, because I haven’t been able to find the paper.
Your answer to the question of the small 1mm errors is appreciated,
w.
So, if you don’t understand it the way Tom understands it, then spend the rest of your life studying it the way Tom did, down to every last minute detail, EXACTLY the way he assimilated his perception, … which, oh by the way, will distract you for the rest of your life, from raising legitimate concerns.
Not saying Tom’s method is the wrong method. I’m saying that there might be a better way to point out that somebody has not considered relevant facts than to tell them to do another whole lifetime of studying to erase any questioning — a lifetime that a person does not have, of course.
Tom, might I suggest you take a couple of years of ballet, then some chef classes, then read a good text about how to mix alcoholic beverages, a class or two in how to set a table, how to use your eating utensils in a dignified, noble manner, how many times to chew your food for maximum digestion, in order to understand your body’s most elegant placement, and your best choices of food/drink and best way of eating/drinking them, … all BEFORE telling me that you don’t like the way I cook my broccoli and you don’t like the choice of wine I serve with it. (^_^)
The response just came across as a bit snobbish.
Regardless of the possibility of one’s great mastery of actually achieving (or believing that one can achieve) accuracy within a tiny fraction of a mm — LET ME REPEAT, <b<fraction of a mm
— the important fact seems to be that we are talking about MILLIMETERS. And we are talking about a FRACTION of a millimeter assessed GLOBALLY.
Admittedly, I speak from “ignorance and Incredulity” as to whether such accuracy is possible, but I’m pretty sure I have a feel for what a millimeter is and how the Earth has changed by far more than fractions of these or even multiples of the wholes of these very small units of measure.
Again:
8 mm</b<
… considering the ENTIRE surface of Earth, weighed against MILLIONS of years of sea level change, BEFORE humans ever set foot (precursor fin, tentacle, or proto-foot) on the planet.
Oops, coding flub.
8 mm
… and ANOTHER coding flub before that one, .. but, oh well, what the hell, I give up for now, you get the message.
Willis: “they are trying to measure the ocean surface to an accuracy on the order of ± a half of a millimetre or so, in order to give answers to the nearest mm”
According to NASA, “The primary instrument on Jason-3 is a radar altimeter. The altimeter will measure sea-level variations over the global ocean with very high accuracy (as 1.3 inches or 3.3 centimeters, with a goal of achieving 1 inch or 2.5 centimeters)”
According to Wikipedia, “From orbit 1,330 kilometers above Earth, TOPEX/Poseidon provided measurements of the surface height of 95 percent of the ice-free ocean to an accuracy of 3.3 centimeters.”
According to the IPCC, “The current accuracy of TOPEX/POSEIDON data allows global average sea level to be estimated to a precision of several millimetres every 10 days, with the absolute accuracy limited by systematic errors.”
So the trick here is to measure many times and take an average. But “Satellite altimeters measure the ocean’s surface height, passing over each point on the ocean’s surface every week or two.”, so can the IPCC be right?
Interesting aside: “In addition to mapping sea-surface height, the other primary mission for the altimeters is to provide maps of wave height and direction (Section 8.3.2 in Chapter 8). Mean wave-height estimates accurate to ±1 m or 25% of the actual wave height are possible from RADAR altimeter backscatter. This measurement is made possible by looking at the waveform returned to the satellite from the altimeter reflection. The slope with which it returns is a function of the significant wave height on the surface of the ocean. Many experiments have been carried out to verify this assertion.”
https://www.sciencedirect.com/science/book/9780750645522