Dr. Roy Spencer on the sea level spat between Gavin and Willis

Dr Roy Spencer writes:

There is a continuing debate over sea level rise, especially how much will occur in the future. The most annoying part of the news media reporting on the issue is that they imply sea level rise is all the fault of humans.

This is why the acceleration of sea level rise is what is usually debated, because sea level has been rising naturally, for at least 100 years before humans could be blamed. So, the two questions really are (1) Has sea level rise accelerated?, and (2) how much of the acceleration is due to humans?

Yesterday’s spat between Gavin Schmidt and Willis Eschenbach dealt with the question of whether sea level rise has accelerated or not. Gavin says it has. Willis says not, or at least not by a statistically significant amount.

I’m going to look at the data in a very simple and straightforward manner. I’ll use what I believe is the same data they did (Church & White, from CSIRO, updated through 2013 here), and plot a trend line for the data before 1950 (before humans could reasonably be blamed), and one for the data after 1950:

If we assume that the trend prior to 1950 was natural (we really did not emit much CO2 into the atmosphere before then) and that the following increase in the trend since 1950 was 100% due to humans, we get a human influence of only about 0.3 inches per decade, or 1 inch every 30 years.

Even though it looks like there is some evidence of even stronger acceleration more recently, sea level has varied naturally on multi-decadal time scales, and it is dangerous to extrapolate any short-term trends far into the future.

Climate models aren’t of much help in determining the human contribution because we have no idea how much of recent warming and the glacial melt was natural versus human-caused.

Models still can’t explain why glaciers started melting in the mid-1800s, just like they can’t explain why it warmed up so much from the mid-1800s to the mid-1900s.

The bottom line is that, even if (1) we assume the Church & White tide gauge data are correct, and (2) 100% of the recent acceleration is due to humans, it leads to only 0.3 inches per decade that is our fault, a total of 2 inches since 1950.

As Judith Curry mentioned in her continuing series of posts on sea level rise, we should heed the words of the famous oceanographer, Carl Wunsch, who said,

“At best, the determination and attribution of global-mean sea-level change lies at the very edge of knowledge and technology. Both systematic and random errors are of concern, the former particularly, because of the changes in technology and sampling methods over the many decades, the latter from the very great spatial and temporal variability. It remains possible that the database is insufficient to compute mean sea-level trends with the accuracy necessary to discuss the impact of global warming, as disappointing as this conclusion may be.”

Read more at Dr. Roy Spencer’s website


newest oldest most voted
Notify of
R. Shearer

Sea levels have been rising for the better part of this inter-glacial period and often much more than is happening now.

Doesn’t appear to be happening now, at least since 2000: http://www.psmsl.org/data/obtaining/stations/1159.php

M Courtney

Willis says not, or at least not by a statistically significant amount.

In fairness to Willis (and I do hate to be fair to Willis) he definitely said that the acceleration of sea level rise was not statistically significant.

He did not say that it was known for reasonably certain that sea level rise had not accelerated.
At least, not on this site.


I would have guessed that we emitted a lot of CO2 during WWII.


Not significantly more.
Civilian production was diverted for military use.

Michael Kelly

Because of the accelerated production of the war effort,we actually omitted much more CO3.


I’m not familiar with CO3.
Regardless, while military production was way up, civilian production pretty much ground to a halt.


I see the time stamps have reverted back to PST.


CO2 was up compared to the Depression, but still not like the postwar period, with so many cars and mass production.

Burning cities in Europe and Asia of course added something.



CO3(2-) or CO3-^2

A salt of carbonic acid.

Michael J. Dunn

I suppose this is a research project. Aside from increased war production, there was tremendous war DESTRUCTION: detonation of explosives, combustion of gunpowder and aviation and Diesel fuels, burning of cities and refineries and combat vehicles and vegetated land. All of which led to increased emission of carbon dioxide. Need I mention flame throwers?


All of which isn’t a pimple on the posterior of the economic miracles of the postwar period.

Bloke down the pub

The more fossil fuels we burn, the stronger the economy becomes and the more resilient we are able to build our defences against the sea.


Yes, that attitude defeated the prophecy of population control advocates and their solutions, including many that migrated to the contemporary catastrophic anthropogenic global warming prediction. It saved lives, advanced welfare, and generally improved our and Posterity’s quality of life .


“Models still can’t explain why glaciers started melting in the mid-1800s, just like they can’t explain why it warmed up so much from the mid-1800s to the mid-1900s.”

The models are all mostly calibrated using aerosols and various parameters on the late 20th century temperature records. So anything anyone claims to give us in terms or attribution or “human fingerprint” about that period is pseudoscience junk. But that hasn’t stopped Glaciologists from layering another model on top of that junk to produce estimates of anthropogenic glacial melt contributions to SLR.

For example:

A new model for global glacier change and sea-level rise
Quoting the abstract:

“The anticipated retreat of glaciers around the globe will pose far-reaching challenges to the management of fresh water resources and significantly contribute to sea-level rise within the coming decades. Here, we present a new model for calculating the twenty-first century mass changes of all glaciers on Earth outside the ice sheets. The Global Glacier Evolution Model (GloGEM) includes mass loss due to frontal ablation at marine-terminating glacier fronts and accounts for glacier advance/retreat and surface elevation changes. Simulations are driven with monthly near-surface air temperature and precipitation from 14 Global Circulation Models forced by RCP2.6, RCP4.5, and RCP8.5 emission scenarios. Depending on the scenario, the model yields a global glacier volume loss of 25–48% between 2010 and 2100. ”

Seriously, it’s dubious glacier models layered on top of failed climate models. And with more hand tuning of individual glacier melt parameters on top of tuning. This is what passes for climate science today.
It’s how they keep the alarmist rhetoric going.

But they can’t model the melting of the glaciers before 1950, especially the high melt years between 1910 and 1945, because if they did, as Roy points out on global temperatures, just minor fraction of melt after 1950 could then be anthropogenic.

In a Nature article by the climatist Stefan Ramsdorf of the Pottsdam Institute, (aka, continetal Europe’s foremost climate carnival barkers), Rahmsdorf lamented that,

Modeling Sea Level Rise
By: Stefan Rahmstorf (Potsdam Institute for Climate Impact Research)
“But when the different modeled components are added up, their sum falls well short of the observed sea level rise for the past decades (IPCC 2007; Rahmstorf et al. 2007). At this stage it must be concluded that physical modeling of sea level rise does not yet provide reliable results, which is the motivation to turn to semi-empirical methods.”

So what have they done? They turned to “Semi-empirical” models” which they could tune to get whatever output they needed to accelerate SLR well above AR4’s 3.2 mm/year.

(the new WP format here at WUWT won’t automatically fetch images… sigh)

Figure legend: Projection of sea-level rise from 1990 to 2100.
Projection of sea-level rise from 1990 to 2100, based on IPCC temperature projections for three different emission scenarios (B1, A2 and A1FI). The sea-level range projected in the IPCC (IPCC 2007) for these scenarios is shown for comparison in the bars on the bottom right. Also shown is the observations-based annual global sea-level data (Church and White 2006) (red) including artificial reservoir correction (Chao et al. 2008). The key difference between projections is that the IPCC projections would be realised if the rate of rise remains roughly constant at the 3.2 mm/year that is observed since satellite measurements of sea level began in 1993 (Church and White 2011), while the semi-empirical projections predict a future acceleration of sea level rise.

How convenient.

Alan Tomalty

I certainly don’t trust satellite measurements of sea level for a 100 different reasons

Measuring sea level with satellites is high comedy.

Measuring sea level with tide gages is low comedy, since
land movements are included in the numbers.

For a better “measurement”: When liberals
start selling their expensive oceanfront property
because of rising sea level, then we MAY HAVE a problem.

Over 399 feet of the past 400 foot rise of sea level,
in the past 20,000 years was NOT caused by manmade CO2.


Love your comparison of natural v human-caused sea level rise in the current interglacial period, Richard. Certainly puts our contribution into perspective.


Even if sea level were rising @ 20 mm per year, it STILL would not deserve even a small portion of the kerfuffle that this silly IPCC scare-fest (scam-fesr?) pumps out each week.

But of course, its actual rise is so absurdly tiny and slow that it’s easy to not even notice a change of antly kind during a 75 year long human life.

But we must perpetually get freaked, and run around like Chicken Little, and write endless journal papers about nothing much at all, as though something had actually changed above noise in the bearly moving watermark. 🐔

It’s a real worry!


I agree, the silly sea-level scare-mongering is absurd. Anyone w/half a brain should see the chicken-little aspect. I guess the scare-mongers have less than half a brain.

matthew dalby

One thing that is rarely mentioned in discussions on increasing sea level is the amount of water that is being taken from deep aquifers. This is often called fossil water because it has been underground for thousands of years. Many aquifers are being drained (largely for agriculture) at a rate many hundreds of times faster than they can naturally refill. Once the water reaches the surface it ultimately ends up in the oceans. Can’t remember the source, but it has been claimed that adding all this extra water to the surface could be causing sea levels to rise by 1 mm a year, or 1 cm a decade. 1 inch is roughly 2.5 cm, so 1 cm is roughly 0.4 inches. Therefore if humans are responsible for an extra 0.3 inches of sea level rise per decade most of this could be due to the draining of deep aquifers (which I’m pretty sure didn’t happen before 1950) rather than supposed man made climate change.

Mickey Reno

By the same token, no one ever talks about human built reservoirs to stop water running off that otherwise would have flowed into the sea, nor do they consider the climate effects of large, artificially irrigated land, with higher humidity due to surface water that would not have otherwise been present.

One possible contributor to late 20th century warming that does not care about CO2 could very well be the added water vapor GHE over land, due to human irrigation and reservoir evaporation.

Crispin in Waterloo but really in Ulaanbaatar

Matthew D

You are on the right track: Water abstraction is well known, modestly document in terms of volume (especially these days) and that water is not being replenished very quickly. It absolutely is ending up in the oceans because the atmosphere is drying slightly. There is nowhere else to put it.

Total volume of groundwater taken per decade, gives rise X and the total rise Y is sort of known. Y-X = Z = ocean level rise due to all other effects. If Z is less than or equal to zero, there are no ‘other effects’.

As a higher CO2 concentration means plants need less water, we can either increase the land area under cultivation, or dial back and sea level will slowly drop.


I like the chart from some 12,000 years ago, when there was a lot of ice and sea levels were low.
Because there was so much ice, warming was slow. Then, going into some 8,000 years ago it got a lot warmer and sea levels rose steeply. Makes sense.
Glacial ice no longer extends to Chicago, and with much less ice to melt the rise in sea levels has slowed.
And as the terminally anxious go on about, alpine glaciers have retreated hugely.
So, the only serious ice is depressing Greenland and is unlikely to suddenly melt. And then there is Antarctica where the ice has been building for millions of years.
On periodic interglacials, all that is needed is each year’s summer melt exceeds the previous year’s.
On similarly periodic ice ages, all that is needed is that each year’s accumulation of snow and ice to be greater than the year before. Over a couple of decades a trend can be determined.
On the Northern Hemisphere chart of Snow Cover for the trailing year the plot has been mainly on the high side of the mean curve. In the early winter and through the late spring.
Of interest is that the plot for the seasonal low last August-September, the cover was above the the standard deviation band. The melt season was weak and the snow would be reflecting heat to wherever.
It will be fascinating to see how this goes over the months to September.
Will the plot stay above?
Maybe a two-year trend?

Bob Hoye
( An ancient financial analyst who completed a degree in geophysics in 1962. 1957 was declared “International Geophysics Year” and it was motivating.)

jDr. S. Jeevananda Reddy

Few points:

After 50s, extraction of oil, gas, water — subsidence; destruction of coastline for urbanization and to meet other human needs/greed;

Before 50s and after 50s network — different, play major role

So how can we compare prior and after 50s?

Dr. S. Jeevananda Reddy

Nice analysis. Still a bit problematic, since C&W changed their tide gauge sample in each successive paper. That sample matters because of vertical land motion at each gauge.
For a fairly rock solid ‘no acceleration’ conclusion, see my previous guest post here “Sea level rise, acceleration, and closure’. Covers all the issues.

Agreed. On Twitter, or when sloppy and in a hurry, I say “no acceleration.” That’s shorthand for, “There’s been no significant, sustained acceleration in the rate of sea-level rise, over the last nine or more decades, detectable in the measurement data from any of the longest, highest-quality, coastal sea-level records.” Which is right.

That is true at every site with a very long, high-quality measurement record. If you do a quadratic regression over the MSL data, depending on the exact date interval you analyze, you may find either a slight acceleration or deceleration, but unless you choose a starting date prior to the late 1920s, you’ll find no practically-significant difference from perfect linearity. In fact, for the great majority of cases, the acceleration or deceleration doesn’t even manage statistical significance.

What do I mean by “practically-significant,” you might wonder? I mean that, if the acceleration or deceleration continued for a century, it wouldn’t affect sea-level by more than a few inches. That means it’s likely dwarfed by common coastal processes like vertical land motion, sedimentation, and erosion, so it is of no practical significance.

For instance, here’s one of the very best Pacific tide gauges. It is at a nearly ideal location (mid-ocean, which minimizes ENSO effects), on a very tectonically stable island, with very little vertical land motion, and a very trustworthy, 100% continuous, >113-year measurement record (1905/1 through 2018/3):


As you can see, there have been many five-year to ten-year “sloshes-up” and “sloshes-down,” but there’s been no sustained acceleration, and no apparent effect from rising CO2 levels.

The linear trend is +1.482 ±0.212 mm/year (which is perfectly typical).

Quadratic regression calculates an acceleration of -0.00539 ±0.01450 mm/yr².

The minus sign means deceleration, but it is nowhere near statistically significant.


{cont’d…} (I’m posting in parts because I got “500 Internal Server Error” when I tried to post the whole comment}

Hmmmm… I see that the graph didn’t show up as an image. Maybe an IMG tag will do it?

Nope. The IMG tag just disappears.

And I got another “500 Internal Server Error” when I tried to use & lt ; to make a less-than symbol.


To calculate the effect of a century of sustained acceleration on sea-level, you divide the acceleration by two, and multiply it by the number of years squared, 100² = 10,000. In this case, -0.00539/2 × 10,000 = -27 mm (about one inch).

That illustrates a rule-of-thumb that’s worth memorizing: if you see claimed sea-level acceleration or deceleration numbers on the order of 0.01 mm/yr² or less, you can stop calculating and immediately pronounce it practically insignificant, regardless of whether it is statistically significant.

However, the calculation above actually understates the effect of projecting the quadratic curve out another 100 years, compared to a linear projection, because the starting rate of SLR is wrong. On the quadratic curve, the point of “average” (linear) trend is the midpoint, not the endpoint. So to see the difference at 100 years out, between the linear and quadratic projections, we should calculate from that mid-date, rather than the current date. In this case, that adds 56.6 years, so we should multiply half the acceleration by 156.6² = 24,524.

-0.00539/2 × 24,524 = -66 mm = -2.6 inches (still of no practical significance).


Church & White have been down this “acceleration” road before. Twelve years ago they published the most famous sea-level paper of all, A 20th Century Acceleration in Global Sea-Level Rise, known everywhere as “Church & White (2006).”

It was the first study anywhere which claimed to have detected an acceleration in sea-level rise over the 20th century. Midway through the paper they finally tell us what that 20th century acceleration was:

    “For the 20th century alone, the acceleration is smaller at 0.008 ± 0.008 mm/yr² (95%).”

(The paper failed to mention that all of the “20th century acceleration” which their quadratic regression detected had actually occurred prior to the 1930s, but never mind that.)

So, applying the rule-of-thumb above, the first thing you should notice is that 0.008 mm/yr² of acceleration, even if correct, is practically insignificant. It is so tiny that it just plain doesn’t matter.

In 2009 they posted on their web site a new set of averaged sea-level data, from a different set of tide gauges. But they published no paper about it, and I wondered why not. So I duplicated their 2006 paper’s analysis, using their new data, and not only did it, too, show slight deceleration after 1925, all the 20th century acceleration had gone away, too. Even for the full 20th century their data showed a slight (statistically insignificant) deceleration.

My guess is that the reason they wrote no paper about it was that the title would have had to have been something like this:

Church and White (2009), Never mind: no 20th century acceleration in global sea-level rise, after all.

# # #

Marine charts are corrected for GPS. They are drawn to the nearest 1 foot. Yet no correction for sea level rise. Even though most charts are more than 200 years old.

Funny how when lives are on the line sea level rise disappears.

Warren Blair

Church & White eh.
Then you must also study the work of another local boy from down-under (no longer with us unfortunately).

Steve Case

Church and White seem to find rates of sea level rise much higher than what you can glean out of the tide gauge record. Considering that they used less than 50% of them because they tossed out any tide gauge in an estuary and that seems to be a lot of them. For example they tossed out Churchill Manitoba
on Hudson Bay. It’s the only tide gauge for nearly 1,000 miles in any direction, and they tossed it. Sea level is falling there. If you follow the link you will find a map showing that it’s about 1.5 miles from the open bay, and the Churchill River is about 1.5 miles wide at that point.

Even if you use the tide gauges they selected you don’t come up with the rates of sea level rise they do with any sort of straight forward analysis.

Here’s a link to my post
from a few days ago covering this issue.

Smart Rock

Steve – Churchill is a really bad example to use. Not only is it on an estuary, but most of the flow of the Churchill River was diverted into the Nelson River to feed hydro-electric developments in the 1970s or 1980s. The town of Churchill had to move the intake for their water supply way upstream to keep getting fresh water. Hard to quantify the effect of that on water level in the estuary, but hard not to imagine there would be an effect.

Then again, land in the Hudson Bay region is rising due to isostatic rebound, probably faster than anywhere else on earth. Just look at all those raised beaches.

Plus, Hudson Bay has lower salinity than the open ocean. It has a huge catchment area feeding fresh water, it is iced over for 6 months of the year (less evaporation), it is shallow (average ±100 m) and has a limited connection to the Arctic ocean. It’s not hard to imagine the average salinity changing from wet years to dry years, and that will affect the water level. The less saline it is, the higher the water level needed to maintain hydrostatic equilibrium.

Even if they had GPS correction, I would still distrust any measurement of SLR from Churchill.

Sorry to be negative. I spent a good deal of time in Churchill waiting for the fog to lift so we could fly out to exploration projects in what is now Nunavut, so I know it fairly well. Mostly I remember the fog.

Steve Case

Thanks for the reply. If there’s a signal of overall sea level rise from the decreasing land ice from around the globe, that should be reflected in all the tide gauges. A simple quadratic plot of the Churchill data from the PSMSL
shows that any overall rate change is not positive. Here’s a link to Dave Burtan’s site to illustrate the point:



Anything related to Climate or the Environment that comes from Australia’s CSIRO or Australian Universities is nothing Anti Science and political Agitprop. It would be like using data from Lysenko while studying Biology or Agriculture.

Actually, our emerging understanding of the multi-generational effects of epigenetic expression in response to environmental cues is bringing us some respect for Lysenko, whatever other blunders he may have made.


Where were sea levels before the multi-mile (miles thick folks) thick ice melted on the land in North America over thousands of years to form the Great Lakes? Just sayin…

Really, think about it…..

comment image

Kristi Silber

“If we assume that the trend prior to 1950 was natural (we really did not emit much CO2 into the atmosphere before then)”

I don’t think that’s a safe assumption. The change in CO2 isn’t just due to fossil fuel use, it’s also due to change in land use. A lot of forests were burned off, both contributing CO2 to the atmosphere and decreasing the sink.

I also don’t understand the argument that seas were already rising naturally.

I suppose the explanation for the graphs is that they are based on bad data?


“I also don’t understand the argument that seas were already rising naturally.”
Of course you don’t, What sort of people wanting to understand would spread material from a self-acknowledged agit-prop agency “2 Degrees Institute”, instead of looking skeptically at it and looking for the tricks? A few of which are quite obvious:
* start date was obviously cherry-picked to fit the narrative…
* …including a zoom to make 1 mm look significant, when previous sea-level change were 1000 to 100,000 higher
* while we now have some data to monitor micro changes at <1 year time scale, such data doesn't exist for past time. How practical, they don't even have to erase the past, just sharpening the present image is enough to create an image of change out of a situation just the same as before.

But that would had been if you wanted to understand, of course…


I don’t think that’s a safe assumption. The change in CO2 isn’t just due to fossil fuel use, it’s also due to change in land use. A lot of forests were burned off, both contributing CO2 to the atmosphere and decreasing the sink.

As I live near a sink that wasn’t there in 1900, but was there in 1950, I tend to get picky at this point.

Whatever you say about 1900-1950, the emissions were much smaller and atmospheric portion was near to what you people want it to be. Now we have 410 ppm and a pretty good 2 ppm/yr increase. So there should be an acceleration. If there isn’t, CO2 has practically no effect.

There is a bloody good motivation to find acceleration, the fact it is hiding (and not in the ocean) is a travesty.

Hoyt Clagwell

Kristi, I don’t know what to make of your sea level graph. The bottom time increments show four places marked as 1K years ago, and two places marked 0K years ago. I have no idea what time frame that chart is supposed to be showing. Can you explain it please?


“I also don’t understand the argument that seas were already rising naturally.”

The oceans were warming coming out of the LIA. This caused the oceans to release some of the CO2 that had been stored there.


A lot of forests were burned off, on the other hand a lot of forests were allowed to grow back.
In the US it is estimated that by the middle of the last century there were more forests in the US than there were at the time of the revolution.
This is due to the opening of the midwest which allowed the creation of large efficient farms. Which in turn caused the many smaller and farms in New England and other areas to close and return to forest.

Jeff Alberts

And how many forests were burned by American Indians?

Depuis des millions d’années il y a des milliards de km³ d’eaux douces (venus des pluies, des fleuves & des rivières) qui se sont déversés dans les mers & océans… SANS QU’ELLES OU ILS NE MONTENT !!! Çà alors ! Tout simplement parce que l’eau s’infiltre continuellement dans les planchers océaniques et maritimes vers le magma où cette soupe toxique (les poissons chient dans la mer !) y est chauffée/bouillie et remonte donc (comme dans une cafetière électrique) vers les sources (chaudes ou froides suivant l’altitude) et vers les nappes phréatique qu’elle remplit.

For millions of years there are billions of km³ of fresh water (from rains, rivers & rivers) that have poured into the seas & oceans … WITHOUT WHERE THEY DO NOT UP !! ! That’s it! Quite simply because water continuously seeps into the ocean and sea floors to the magma where this poisonous soup (the fish shit in the sea!) Is heated / boiled and goes up (as in a coffee maker) to the sources (hot or cold depending on the altitude) and towards the water tables it fills.


Aquifers are filled from above, not below.

“If we assume that the trend prior to 1950 was natural (we really did not emit much CO2 into the atmosphere before then) and that the following increase in the trend since 1950 was 100% due to humans, we get a human influence of only about 0.3 inches per decade, or 1 inch every 30 years.”
That actually isn’t a correct calculation of accelerating rate. It looks like a change of 2″ attributed to 60 years of acceleration. If the acceleration is assumed constant, the correct formula is
2″ = 1/2*a*t²
Using decade units, a=4/36=0.11″/dec². That would make the current rate attributed to CO₂, 0.66″/dec. But more to the point, it implies the rise over the next 60 years, with the same acceleration, would be 6″ (4*2-2).

But assuming constant acceleration isn’t right, because the CO₂ forcing hasn’t been constant, but rising. If you assume that acceleration is itself increasing linearly, the rise in the next 60 years would be 14″.

Now I can just hear the chorus – you can’t extrapolate. But there is no points in calculating rates and acceleration, unless you are thinking about what might happen if that rate and acceleration kept up.

” at least not by a statistically significant amount”
This just means that you can’t assume as a predictive law that the rate in other circumstances would be what you observe to date. It might be higher or lower. But what you observe becomes the best estimate.

The important point is that accelerated sea level rise was a predicted result of CO₂ emission. We look at the results, and see that happening. So the prediction is looking good. Having a prediction work out may not prove a theory, but it is a step on the way. The theory can’t to better than get it right.


if one were to extrapolate the real 1910-1945 surface T rise into a modeled (presumed) SLR acceleration, then NewOrleans, South Florida, and NYC would already be under 30cm of seawater.
But they ain’t.

@Nick Stokes

Are there failed predictions related to elevating CO2 emissions? It seems fairly easy to say sea levels are going to continue rising and subsequently blame it on CO2 while conveniently ignoring or downplaying natural variability. After all, sea levels have been rising naturally since the last glacial maximum. It’s like rolling a ball down a hill and predicting it’s going to continue rolling down that hill. That kind of prediction is what’s known as low hanging fruit. Can we predict when sea level rise will accelerate or decelerate as it does naturally? It seems obvious that natural variability hasn’t suddenly stopped. So, how do you reliably distinguish between natural variability and anthropogenic forcing in sea level rise??

There’s no sarcasm or snark intended in what I’ve said or in my questions. It seems to be self-evident that these are important questions to ask and have answered.

“So, how do you reliably distinguish between natural variability and anthropogenic forcing in sea level rise?”
Maybe you can’t, reliably. That’s the point about lack of statistical significance. But it doesn’t mean insignificant. If a change is consistent with an AGW cause, then that is support for AGW. If it could also have happened naturally, that means that it wasn’t a very strong test, but it isn’t a negative. If it is a weak test, the merit goes the other way too. A negative result might be because natural variation countered an AGW effect. Still a negative.

Generally, if you can’t “reliably distinguish” between natural and anthro as causes, the remedy is to await more data, to get a stronger test.


“If a change is consistent with an AGW cause, then that is support for AGW”.

I am not sure I understand this statement. Doesn’t the statement “is consistent with” mean “doesn’t disprove”? If it means “is support for” why not just say that instead of “is consistent with”? I would appreciate any clarification on this you can supply.

“Is consistent with” is a math statement you can test for. “Is support for” is a matter of judgement. How much support it provides depends on what other possibilities it is consistent with.


“‘Is consistent with’ is a math statement you can test for. “Is support for” is a matter of judgment. How much support it provides depends on what other possibilities it is consistent with.”

Thanks for responding to my question. If I understand your answer correctly, when you make the statement ““If a change is consistent with an AGW cause, then that is support for AGW” you are not making a mathematical statement, but simply expressing an opinion or “judgment” of its support of AGW.

“when you make the statement”
Well, it starts with “If…”. But according to what is said here, there was acceleration, and that is consistent with AGW, and provides at least some support. How much support depends on whether the observation of acceleration might have arisen from natural variation. As time goes on, natural variation is a more unlikely explanation, and support for AGW increases (if acceleration is sustained).


Thanks again for your patience in answering my questions. One of the reason I like this site is that very knowledgeable people such as yourself are willing to post comments and answer questions.

“…there was acceleration, and that is consistent with AGW…” & “…the observation of acceleration might have arisen from natural variation…”

In other words acceleration in sea level rise is consistent with both AGW and Natural Variation.

“As time goes on, natural variation is a more unlikely explanation, and support for AGW increases (if acceleration is sustained).”

I have not seen any studies that indicates that Natural Variation is an unlikely explanation for a sustained acceleration in sea level rise. Can you point me to a study that says that this is true?

” Can you point me to a study that says that this is true?”
I do not follow the sea level story very closely. I didn’t say that natural was unlikely, just becoming less likely as time goes on. The quantification of this is statistical significance, and as quoted, Wilis says that acceleration is not currently significant. I presume that is based on a quantitative study.


No climate predictions are ever falsified. They just get adjusted after the fact.

But there is no points in calculating rates and acceleration, unless you are thinking about what might happen if that rate and acceleration kept up.

Good point, but just juggling with mathematical variables is a poor basis for extrapolating.

You need to justify the drivers by a theory or model based on physical realities.
In this case I am not aware of any physical property that can be directly connected to drive the acceleration of sea level.

What you can justify is a connection between climate forcing and additional sea level rise per decade, not acceleration per se.

Or, do I miss something here?



Calling it a theory doesn’t make it a Theory.

The premise that CO2 is in some significant way related to (causes) temperature increases and other calamities has missed its predictive mark.

The data associated with SLR was available and evident prior to the “CO2 will accelerate SLR” claim, so there is no confirmed prediction. If you want confirmation of a prediction you can’t start with data that is already doing what you want it to do.



I have a theory that my friend is psychic.

And just this morning she predicted that the NBA finals will take at least 6 games to determine a champion.

Having this prediction work out may not prove my theory, but it is a step on the way. The theory can’t to better than get it right.

Nick wrote, “But assuming constant acceleration isn’t right, because the CO₂ forcing hasn’t been constant, but rising.”

Not much. The CO2 level increase has accelerated over the last few decades, but because of the logarithmically diminishing “greenhouse” effect, the CO2 forcing increase has been only just slightly more than linear.

There was a substantial acceleration in CO2 forcing in the 1960s, but look how linear it’s been for the last few decades:


The acceleration has been diminishing for half a century, and now is only barely above zero. So the CO2 forcing trend is now very close to linear. I would bet dollars to donuts (hmmm… that expression doesn’t work so well anymore, does it?), that is, I’m very confident that the CO2 forcing trend will be less than linear over the next sixty years.

Alexander Carpenter

I can’t believe that no commenter has pointed out that any linearity of the “natural” sea-level change rate is nothing more than an assumption, the kind of ungrounded and habitual assumption that is native to reductionist linearizing science, and one of its greatest weaknesses. Perhaps I missed in any such observation here in all the noise and posturing.

Since we really do not know all the forces and effects involved, it is gratuitous to simply accept that assumption and make another, that humans are responsible (“CO2 forcing”) for a change in the rate (instead of some other phenomenon or influence that we don’t really understand). Or maybe the rate is naturally variable, perhaps even cyclic.

I agree with Willis about the statistical significance, if only because we have insufficient detail in past changes whose proxy data is averaged out, while the current data is not. And then I have to ask, where are the error bars? And, will longer observation-periods tell us something different? Do we really need to rush to judgement?

Yet another tempest in a teapot.


Visual evidence? I see many pictures of shoreline properties and even island properties that were constructed in the early 1900’s and they are STILL as they were when they were first constructed. Pictures can’t lie. There’s that famous ‘tree’ on an island too. We need to show more of these images for those who are inclined to take their cAGW gospel from the MSM who just love to both dumb-it-down and hype-it-up (DIDHUP) in the same article.

Alan Tomalty

This whole sea level rise threat is such a farce anyway. Even if all 200000 glaciers in the world melted the sea level rise would only be 400mm. That is less than 16 inches. Antarctica will never melt; even if global average temperatures go up 4C; because almost all of the continent is way below freezing even in the Antarctic summer. Since the Arctic ice sits on the ocean; sea level will not rise even if the whole Arctic melted. What is left to melt? Greenland. Okay Consider this. So even if the CAGW alarmists are right and the temp rises to 4C over today, how much of Greenland would melt? Well in the vast interior of Greenland only the very top of the ice melts in the summertime. Even the IPCC has stated that a 3C global rise in temperature over the next 80 years would only result in a 1 metre rise in sea level. I dispute that because the vast majority of Greenland’s ice sheet never even comes close to reaching 0C. It would take a much larger increase of temperature to melt it. Everybody just seems to take the alarmist view on this without looking at the actual size of ice that would have to melt. Recently an engineer calculated that it would take 105000 years to melt the Antarctica even if you had all the energy of the world running blowtorches melting the ice. Greenland is not nearly as big but to try and melt even a thousandth of Greenland even if you had access to industrial melters on every last inch of the Greenland interior, would be a futile task. Greenland has 2,850,000 km3 of ice.
All of it would have to melt to raise sea level by 7 metres. This is just not going to happen especially even with a 4C average global temperature rise which is at the high catastrophic range of IPCC predictions. You just cannot melt that large a block of ice with air temperatures 4C higher . This is because you are dealing with averages here. The summit which is 2 miles above sea level in the interior has an air temperature range of -26C in winter to 0C in summer. Summer in this part of Greenland is only 2 months long. Temperatures in the other 10 months of the year are below
-10 C. So 2 months of summer is just not enough time to melt an appreciable amount of ice. Increasing the average global temperature by 4C will not make the interior go above freezing because of the elevation.
Additionally, the weight of the Greenland ice has depressed the interior of the continent and disrupted any drainage that existed prior to being covered in ice. If the ice should be completely melted, a significant fraction of the water won’t make it to the oceans until isostatic rebound removes the ‘bowl.’ The bottom line is that theoretical calculations converting the ice volume of Greenland to an increase in ocean level overstates the immediate effect.

I would also like to draw your attention to this graph


It shows the alltime record summer temp for Summit station in inland Greenland. Notice that it barely got above 0C. Since summers are only 2 months long here how in the hell is Greenland supposed to melt any appreciable amount even if global temps went up 4C. The summit is 2 miles high and the mean thickness of the ice in all of Greenland is 2135 metres or 7000 ft. Since this total of ice is 2850000 km3 , how would this melt in 2 months? It wouldn’t. fall and winter would come and the ice would refreeze. Spring would come again and as you see on the graph there wouldn’t be any melting in the spring even if global world temperatures soared above an increase of 4C. Sure Greenland has been losing ice mass over last 20 years but this has happened thousands of times in the past. There was less ice at the end of the 1930s in Greenland than there is today.

To further cement this hypothesis of Greenland ice sheet not melting from of top, there have been studies that the melting is happening from underneath because of a volcanic ridge extending from Iceland right to the Arctic. Even the alarmist scientists are admitting that the top of Greenlands interior ice sheet is not melting and that the upper surface every year gets fresh snow/ice and the reason that there is a net loss of ice is the amount of icebergs calving off on the shore line. These icebergs have calved off for millions of years and the volcanic activity has come and gone for millions of years.The alarmists will argue that the calving of the icebergs on the coast of Greenland is increasing with global warming.
However, that demonstrates a lack of understanding of just how calving works. Calving is a breakup of ice shelfs at the coast caused by pressure from the ice sheets as the ice is forced to the sea. Calving is just as likely to happen when it is cold or warm. Calving has been going on ever since Greenland formed ice sheets.

The global alarmist position is a farce on every level.


water has a volumetric thermal expansion coefficient of ~210/10^6 per K, or 0.21 mm/(m.K) if constrained to expand only in a single direction. That is, to rise the level by 2.1 mm you need to rise by a full 1°C the temperature of a water column 10 m, or any similar combination (0.01K x 1000 m water column, etc.)
water has a thermal capacity of 4.18 J/(gK), that is, the aforementioned temperature rise would require 41.8
10^6 J. This in turn would require a ~1.3 W steady yearlong power input.
1 W year average variation is not particularly noticeable, and well within usual natural variation of sunpower.
And that’s thermal expansion alone
2 mm/year can be done by Nature without even noticing.
And human don’t even care, of course. We don’t need a century to adapt to a 20 cm rise, or even a full meter. It can be done without even noticing, the way Chigago did
If need be.

David Dibbell

On his own blog, which I read regularly, Dr. Spencer’s headline for this post is “Sea Level Rise: Human Portion is Small.” Evidently he means that even if there is a contribution, it is small at most. But it also remains true that the human portion of any acceleration could be zero or negative. Future history remains to be written.


Factors affecting sea and shore level changes as explained by Dr. Nils-Axel Mörner:
Coastal dynamics: Erosion, silting, sediment transport, continental runoff, air pressure changes, long term trends, storms, hurricanes and tsunamis.
Changes in land level: compaction, geoid deformation, seismotectonics, hydro, sediment and glacial-isostasy,
Sea level changes: stearic effects, temperature, salinity, basin volume changes (long term tectonic and glacial rebound), geoid deformation, and glacial eustasy.
Isostasy is the general state of equilibrium of the Earth’s crust, with rise and fall of land relative to sea level. Eustasy is the uniform change of sea level throughout the world.
This puts a more realistic slant on all the CO2 horror stories, doesn’t It?


I don’t know about you but determining Absolute Sea Level in 1850, using tidal guages which record Relative Sea Level only, makes me cringe.

Just how large are the error bars on that graph anyway?


230,000 people are estimated to have died following the 2004 Boxing Day Tsunami. Amazingly far fewer, approx 16,000 in the Japanese Tsunami of 2011. While the Japanese one was of notable size, neither were especially unusual geologically.
With sea-level rises like that, it is quite beyond me why people spend precious taxpayer dollars worrying about 3 or 6 mm per year when we have decades to decide clearly which it is.
The money spent on climate change would be far better spent put towards building resilient societies that can withstand foreseeable disasters. Instead we are plunging headlong into a world where many entirely natural perturbations (earthquake, volcano, meteorite, pandemic, solar flare, tsunami – did I forget any?) could send us into ‘Mad Max’ territory.

Alan Tomalty

Is this the same Gavin Schmidt who earlier this year coauthored a paper on exploring evidence in earth’s past of an alien race who were wiped out on earth (before homo sapiens developed ) because of their advanced use of fossil fuels causing CO2 to then cause runaway global warming? Gavin Schmidt leader of men and aliens NOT.


The late, great John Daly on Tasmanian sea level:



No on 1 and 0 on 2.

comment image


Have to agree.

Humans haven’t warmed the air enough globally to have any impact on sea level rise since 1950, although we could possibly have had a minor effect from irrigation and similar activities.


Remember to tie all claims back to the CO2 molecule and its underlying physics:
1) The chart dog-legs. CO2 shows a log decay. It would not cause a dog-leg or acceleration.
2) Just look at the Battery Park tidal gauge, there is no trend or acceleration.
3) Thermal expansion can cause an increase in sea level. CO2 doesn’t warm the oceans, visible radiation does.
4) Sea Ice and the Greenland glaciers are melting from BELOW. CO2 doesn’t cause warm oceans or geothermal heat.
5) The dog-leg doesn’t match the dog-legs on the Hockeystick chart. Either way, CO2 can’t cause a dog-leg.

Steve O

The lower the understanding of the basic drivers of major trends in the data, the lower the justification for expensive actions. How much evidence is appropriate for what level of spending is a political question, not a scientific one. We absolutely do not have enough justification for the extreme actions that are being proposed.

And scientists who overstate evidence are putting their thumb on the scale, and are thus betraying the public trust.


The volume of sea ice in the Arctic is currently the largest in 5 years.
comment image


Disappointing coming from a professional scientist.

There is no such thing as a “NATURAL RISE” in sea level: Changing SL is CAUSED by:
1) An increase in temperature (thermal expansion.
2) An increase in volume – caused by changes in the amount of ice (and water) on land.
3) Very slow changes in the volume of ocean basins.

The first two are caused by rising temperature, but the third is not. We know that the 3) is not responsible for the alleged “natural rise” in SL (0.5 inches/decade) between 1850 (when tide gauge records became useful) and 1950 (when the IPCC asserts rising GHGs became the predominant driver of temperature change). How do we know this? Simple. 0.5 inches/decade is 50 inches/millennium or 1.25 m/millennium. Over the last 4 millennia, we are positive that there hasn’t been 5 m of SLR. The rate of rise is indistinguishable from ZERO.

Some scientists even believe that SL may have FALLEN slightly since the end of the Holocene Climate Optimum in parallel with the fall in temperature.

Frank wrote, “There is no such thing as a ‘NATURAL RISE’ in sea level…”

Of course there is. “Natural” sea-level rise is SLR which is not caused by man. Dr. Spencer was perfectly clear that that’s what he meant.

Sea-level has been rising and falling for as long as there have been oceans — and it’s perfectly natural.

Frank continued, “Changing SL is CAUSED by…
2) An increase in volume – caused by changes in the amount of ice (and water) on land…
… caused by rising temperature…”

Not entirely. Many things other than “rising temperature” can change the amount of ice/water sequestered on land.

The most important factor affecting ice sheet mass balance is snowfall, but climate alarmists rarely mention it. They talk plenty about melting ice and glacier calving, and mostly ignore the factor which is much more important and much greater in magnitude than either of those, because snowfall accumulation ADDS to ice sheet mass, and subtracts from sea-level, which means it is not useful to the alarmist narrative.

A warming climate causes processes which both increase and decrease sea-level rise. Based on the sea-level measurement record, it seems that when there’s no Laurentide ice sheet to melt those processes roughly balance each other.

The magnitude and importance of snowfall on ice sheet mass balance (and thus sea-level) is illustrated by the story of Glacier Girl.

She’s a WWII Lockheed P-38 Lightning which was extracted in pieces from beneath 268 feet of accumulated ice and snow (mostly ice), fifty years after she made an emergency landing on the Greenland Ice Sheet.

Do the arithmetic and you’ll calculate an astonishing number: more than 5 feet of ice per year, which is equivalent to more than seventy feet of annual snowfall!

That ice and snow represents evaporated water, mostly from the Arctic and North Atlantic Oceans, which then fell as ocean-effect snow on the Greenland Ice Sheet:

The story of Glacier Girl is fascinating. You can read more about it here:
and here:

So, the key question is: what happens to snowfall in a warming climate?

The answer is that it increases, for two reasons.

First, it increases simply because warmer air holds more moisture. Every meteorologist knows that the biggest snowfalls occur when the temperature is not too far below freezing.

Second, a warmer climate should reduce sea-ice extent, increasing evaporation from the Arctic, North Atlantic, and Southern Oceans, and thereby increasing Lake/Ocean-Effect Snowfall (LOES) downwind. (Ice-covered water does not produce LOES.)

When additional snow falls on ice sheets and glaciers, it adds to ice mass accumulation and subtracts from sea-level.

So, which effects are greater in a warming world? Those which decrease ice mass and increase sea-level, or those which increase ice mass and decreases sea-level?

The best evidence to answer that question is history, because we already have some experience with the effect of temperature changes on sea-level. What it tells us is that those opposing processes must be very similar in magnitude, because the approximately 1°C of warming which the Earth has experienced since “pre-industrial” (Little Ice Age) conditions was associated with only a very, very slight acceleration in sea-level rise, all of it more than 85 years ago. Sea-level is rising no faster now than it was nine decades ago, when CO2 was under 310 ppmv, and CH4 was 1.0 ppmv. A 30% increase in CO2 and an 80% increase in CH4 have caused no acceleration at all in the rate of sea-level rise, and there’s no reason to suppose that similar future increases will cause dissimilar results.

Frank continued, “…Simple. 0.5 inches/decade is 50 inches/millennium or 1.25 m/millennium. Over the last 4 millennia, we are positive that there hasn’t been 5 m of SLR. The rate of rise is indistinguishable from ZERO.”

Your arithmetic would only be correct if you assumed that the rate of sea-level change has not varied over the last 4000 years. Why would you assume that?

Climate alarmists frequently do arithmetic like that, with sea-level or temperature, and declare that recent increases in either or both are unprecedented in thousands of years, because of evidence that temperatures or sea-levels were many degrees or meters different thousands of years ago. It is a common fallacy.

The evidence is very strong that there’s nothing at all unusual about the modest warming which the Earth has experienced over the last century, or the minute amount of sea-level rise.

That fallacy is a product of statistical confusion. Paleoclimate information, inferred from indirect evidence like marine sediments, is naturally “smoothed,” by processes which blend the evidence from consecutive decades, centuries, and millennia. As any engineer knows, when you smooth a graph, sharp fluctuations disappear. But many people apparently don’t know that. They see a paleoclimate graph and say, “look, it took x-thousand years to change by a few degrees or a few feet, that’s much slower than the 20th century!” But, of course, they have no way of knowing how many times it went up or down much more rapidly than that during a single decade during that x-thousand years.

What’s more, other than the recent higher CO2 levels, from fossil fuel use, the current Modern Climate Optimum is very similar to the Medieval Warm Period and the Roman Climate Optimum, and probably cooler than it was during the Holocene Climate Optimum and most of the Eemian interglacial.

Frank continued, “Some scientists even believe that SL may have FALLEN slightly since the end of the Holocene Climate Optimum in parallel with the fall in temperature.”

Not necessarily “slightly.” if you google search for “Holocene highstand” you’ll find a number of studies which concluded that, circa 3000 BC, in many places, and at least in most of the tropics, sea-levels were considerably higher than present.

That might be because, according Zwally (2015), the Antarctic ice sheet (especially the EAS) has been growing during the Holocene. Here’s an excerpt from the abstract (with my translations added in [brackets]):

“EA [East Antarctic] dynamic thickening [of the ice sheet] of 147 Gt a–1 [Gt/year] is a continuing response to increased accumulation (>50%) [of snow & ice] since the early Holocene.”