Guest post by David Archibald
This is a little bit amusing. In February, I had a post on the solar – sea level relationship which quantified the sea level fall to come to the end of Solar Cycle 25:
http://wattsupwiththat.com/2012/02/03/quantifying-sea-level-fall/
The site “Skeptical Science” has to date carried two pieces in response to that February post: http://www.skepticalscience.com/Why_David_Archibald_is_wrong_about_solar_sea_level.html
and http://www.skepticalscience.com/Why_David_Archibald_is_wrong_about_solar_sea_level_1B.html
My February post was 624 words and 6 figures. The Skeptical Science responses to date total 3,446 words and 17 figures. The relationship I found between solar activity and sea level is 0.045 mm per unit of annual sunspot number. The threshold between rising and falling seal level is a sunspot amplitude of 40. Below 40, sea level falls. Above that, it rises.
So let’s apply that relationship to the know sunspot record back to the beginning of the Maunder Minimum and see what it tells us. This is the result:
Figure 1: Back-tested Sea Level from 1645
The figure shows sea level falling through the Maunder Minimum due to the lack of sunspots and then fluctuating in a band about 60 mm wide before increasing rapidly from 1934. It then shows sea level peaking in 2003 before declining 40 mm to 2040.
That is pretty much in agreement with the data from the last 150 years, as per this figure combining coastal tide gauge records to 2001 and the satellite record thereafter:
Figure 2: Sea Level Rise 1850 with a Projection 2040
The glaciers started retreating in 1859, with sea level responding with a rise of 1 mm per annum up to 1930. There was an inflection point in 1930 with the rate of sea level rise almost doubling to 1.9 mm per annum. Sea level also stopped rising from 2003. So the back-tested model and the sea level record are in agreement for at least the last 150 years.
Jevrejeva et al (http://www.psmsl.org/products/reconstructions/2008GL033611.pdf) reconstructed sea level back to 1700:
Figure 3: Global Mean Sea Level Reconstruction since 1700
This longer term reconstruction shows the rise of sea level once the glaciers started retreating. It also shows the acceleration of sea level rise from the early 1930s. As Solanki noted in 2004, the Sun was more active in the second half of the 20th Century than at any time in the previous 8,000 years: http://earthobservatory.nasa.gov/Newsroom/view.php?id=25538 A sea level response to that would be expected.
In summary, the sea level trend fluctuations driven by the internal variability of the ocean-atmosphere coupled system were overprinted by higher solar activity from 1933 to 2003. The period of best fit within that, from 1948 to 1987, has allowed the solar component of sea level rise to be elucidated.
stevefitzpatrick, the only sea level rise coming from satellite data is the adjustments. No one with a brain trusts satellite data anymore.
Ooops!
“Then measured by the water and got there -2 m / -6.5 ft at low tide and later 4 m / 13.1 ft at the river.” should be “Then measured by the water and got there -2 m / -6.5 ft at low tide and later 4 m / 13.1 ft at high tide.”
Scientists are curious to determine whether WUWT’s moderators actively delete references to contrasting points-of-view.
[REPLY: Why should that question even occur to a knowledgeable scientist? -REP]
REPLY: And, why should I care what a person [Hansel Seuss] who has a fake name, a fake IP, and a fake email have to say? If you have something important to challenge me on, have the courage to put your name on it. Otherwise your opinion and fakery is worthless to me – Anthony
[Reply: As a sporadic moderator, I can tell you that I was given NO direction to delete any contrasting point of view. ( In fact, contrasting points of view increase readership and are to be encouraged). I was given guidance to snip certain offensive language, dampen insulting behaviour and attacks, and not judge based on my evaluation of the technical merit of the argument, only on the “tone” and language used in the delivery. So, no, I do not “actively delete” much of anything other than foul language and attack language. There are some (short list) of posters who are banned from time to time due to repeated infraction of the guide lines; but that’s about it. -ModE ]
stevefitzpatrick says:
April 22, 2012 at 9:31 am
Thanks for the explanation of how El Nino raises the global temperatures.
Energy in, energy out: how much it bunches up in the middle is how warm it gets. Where it shifts around is how one area has a hot spike while another has a cold spike. Unfortunately, Hansen, Gore et al have made people thing that one hot spike a hot world makes.
Interesting, I think the second graph from the paper is more informative.
http://i42.tinypic.com/2lsbjtu.png
rate of change plot shows clearly the cyclic nature plus the longer term rise. The authors fit a linear slope to the rate of change (ie they assume constant acceleration). They seem to discount longer term cyclic variation without saying why.
Rate of change in sea level reflects a power term, ie “forcings”.
Compare to the middle panel of this plot showing the rate of change of temperature:
http://curryja.files.wordpress.com/2012/03/icoads_monthly_adj0_40-triple1.png
http://judithcurry.com/2012/03/15/on-the-adjustments-to-the-hadsst3-data-set-2/
Similar peaks but much less long term rise.
The beginning of their graph seems a lot flatter than the linear rise that could be drawn and the latest bump seems spent despite its only having got to about the average of the 60y cycles.
My reading of that would be either diminishing amplitude or a longer term rise that is flattening out. The earlier flat period would argue in that sense too, though it’s far from clear.
The other thing I’m dubious of is this is all underpinned by the famous GAIA [sic] adjustment for mantle rebound. That is grossly speculative since based on computer models and some estimate for the viscosity of rock ! Error bars please.
Now that adjustment is supposed to account for land rising and ocean basins sinking and hence it gets added to sea level. It’s interesting that the temperature record is similar but much flatter.
This is the first study I’ve seen that allows such a comparison and suggests a touch too much GAIA getting into the “corrections” , yet again.
Still, it’s a very good study and they seem to be doing serious work and do say in two places that their projection is on the assumption that whatever caused the rise continues in the same way.
If anyone knows anything about this subject it must be Dr. Nils-Axel Morner.In 2010 he hoped that by 2012 people would see the truth and that the field data would show that the IPCC is lying about almost everything. We know the true story, the story that we live in a corrupt world. In a world where only corrupt people succeed. These people will never admit that they are/were wrong and give there gained power back. They control the propaganda.I’m afraid it will end in a bloody fight because it has nothing to do with climate and everything with blind power.
Sea Level Basics
http://itsrainmakingtime.com/2010/nilsaxelmorner/
commieBob,
“The glaciers will have stopped melting fifty years from now.”
I’ll be long dead, so it’s not a very interesting bet. BTW, I wasn’t offering a bet, I was making a prediction. ” Glaciers not melting” means temperatures falling by ~0.8C on average compared to today, and much more than that at high latitudes. That is only possible if there is essentially NO sensitivity to GHG forcing. We are not going to have to wait ’till the glaciers stop melting, we need only see low sunspots and a substantial fall in average temperatures over a couple of decades (say, a fall near 0.1C per decade). If you are young enough to be around in 20 or so years, I hope you will remember my prediction: it ain’t gonna happen. 🙂
Stevefitzpatrick says: The measured mass of Greenland ice and Antarctic peninsula ice continue to decline. The measured rate of sea level rise has slowed since 2004, but it is still increasing
Nobody has “measured” the ice mass. All techniques depend on estimates of mantle rebound. (GRACE satellites can not tell the difference between ice mass and land mass so their gravity measurements get… adjusted. )
Measured sea level also gets adjusted for land rising which is based on mantle rebound measurements.
Mean sea level also gets adjusted for ocean basins deepening. Now, no one has even measured that, again is speculation based on computer models of mantle rebound. The official MSL is now hovering mystically above the ocean surface and is getting higher each year. The so-called GAIA adjustment. 😉
Ocean heat content , based on temperatures measured to withing 0.0005 degrees accuracy, right. Any honest error estimate makes that data totally unhelpful in drawing any conclusions, though that will improve with time.come back in a couple of decades.
sunshinehours1,
“the only sea level rise coming from satellite data is the adjustments.”
References and data showing that is the case? The satellites don’t seem terribly out of line with tide gauges (although the tidal gauge data set has more noise than the satellite data), so some references showing that both tide gauges and the satellites are wrong would seem to be called for if you are making that claim.
“No one with a brain trusts satellite data anymore.”
Accusing people of having no brain is neither constructive nor convincing. Measurement data, combined with a reasoned analysis, is more convincing than insults every time.
It’s good to see this type of analysis being further developed to indices like sea level. I derived a similar average sunspot number of around 40 for the ocean equilibrium value from an empirical study a couple of years ago, and performed the same cumulative count integration as done by David Archibald here.
I considered the result in terms of it being a proxy for ocean heat content rather than sea level, but the sea level more or less follows from the OHC anyway, considering water expands as it warms.
http://tallbloke.wordpress.com/2010/07/21/nailing-the-solar-activity-global-temperature-divergence-lie/
David, thanks for your contribution. I must confess I hadn’t read your previous posts. But I must admit, I busted out laughing when I got to this claim in your previous work:
Gosh … so you are saying that if we remove the parts of the record where your claimed relationship doesn’t hold, that your claimed relationship holds quite well. There’s a surprise …
I was again surprised when I came to this graph of yours on a previous post:
Regarding that graph, you say
Well … no. I fear that the correlation is actually quite poor, but our human desire to see patterns over-rides that lack of correlation. This is the problem with eyeballing and “wiggle matching”, we see patterns where none exist.
In fact, the R^2 of your two graphed lines is a meager, pathetic 0.1, and is not statistically significant (p is slightly greater than 0.05).
You could improve your arguments greatly by doing a few things. The first is to actually calculate and report the R^2 of your claimed relationships (including the adjustment for autocorrelation), rather than just eyeballing it and claiming a “striking” correlation where only a very weak correlation exists.
The second is to cite your sources. For example, in Figure 2 you say there are “coastal tide gauge records to 2001 and the satellite record” … which coastal tide gauge records? Which satellite record? With or without the inverted barometer correction?
The third is to explain your choices. For example, the satellite record started in 1992, but you are using “”coastal tide gauge records to 2001 and the satellite record”. Why did you use tide gauge records up to 2001? What did you do during the overlap? Why have you chosen to splice two very disparate record sets (satellite and tide gauges)?
The fourth is to show your work and include your units.
For example, you say:
I don’t understand that. It appears to say that
where SLA is the sea level anomaly in millimetres and SSN is the sunspot number. The problem is that the equation reduces to
But if that’s the case, the sea level would never rise, it would just go up and down with the sunspot number. So there’s a problem with either the units or with your description.
I’m not trying to discourage you, David, I’m trying to encourage you to look at and investigate this scientifically.
My best to you,
w.
ferd berple says:
April 22, 2012 at 10:18 am
One if the great mistakes in modern science is the quest to explain “why”. History shows that “why” is at best temporary in science. “Why” is better left to religion and philosophy.
Why does a dog chase a rabbit? Why does the rabbit run? Does the rabbit run because the dog is chasing it, or does the dog chase the rabbit because the rabbit is running?
In the end the “why” is really not very important. What is important is this: Will the dog catch the rabbit? The ability to predict that answer is the basis of science. It is what gives science value.
—
The first question I would like to ask you is: If science made a mistake by asking “why”, what damage did it cause?
Would science have progressed faster, or further, or in an altogether more exalted and wholesome fashion, had scientists not asked this question? If that is your position, you need to provide more compelling reasons than a rabbit chase.
If, on the other hand, you admit that science did rather well all the same, then your objections to the question that, according to you, should be better left to philosophy or religion, are themselves philosophical, and again would need to be substantiated in order to be taken seriously.
If you have had a chance to interact with any three year old kids, you will have observed that the human mind will just not stop asking “why”. Can we acknowledge that asking “why” evolved because it served us humans well, just like their light-footedness served dogs and rabbits well? If asking “why” has served us well as part of common sense, then, why, precisely, is it such a grave mistake in science?
What constitutes a satisfactory answer to a question “why”? A demonstration of how to predict the observed phenomenon from existing, accepted knowledge. Asking “why” therefore promotes parsimony – if we can work out something from existing knowledge, we don’t have to burden our memory with amorphous, solitary bits of information. It encourages us to test and evaluate the predictive power of our knowledge – can we come up with a satisfactory, consistent explanation, or can’t we, or would we even have predicted a different outcome? If the latter, what part of our knowledge needs changing? Even if it may seem idle now, one day our life may depend on that improved predictive power.
There may be a good natural reason for breakdown in the late 1940’s, there is a similar discontinuity between the NAO and the AMO. I have provided some initial ideas here:
http://www.vukcevic.talktalk.net/theAMO-NAO.htm
fully documented explanation to follow in part II.
Willis, you are correct that David’s description is inadequate. He has. I believe, followed the procedure I used in 2009.
That is:
Work out the ocean equilibrium equivalent in SSN. This is around SSN=40, and also happens to be the average SSN over the length of the sunspot record from 1749 (SIDC).
Then make a cumulative total departing from this value, which in effect performs the integration.
i.e. if the first monthly value is 40, and the second is 46 the index goes up by 6. If the next monthly SSN value is 30, the running total drops by 10 to 36.
The assumption underlying the procedure is that the cumulative total mimics the retention of solar energy input by the ocean, as an approximation of OHC (or sea level).
Since we don’t have long term accurate records of any of these quantities, getting too hung up on R^2 values probably isn’t worth the candle, but as a rough wiggle match where inflection points more or less coincide, it has some qualitative value in helping inform judgements as to how likely it is that solar variation is the dominant process, via whichever amplification caused by cloud cover, ozone levels, latitudinal shifting of climate zones etc you tend to give weight to.
Willis Eschenbach says:
April 22, 2012 at 12:15 pm
Willis, given the fact that the climate is chaotic, I’m not sure it is reasonable to expect a correlation between a given solar parameter and the rate of change of sea level rise. Other actors will still be at work – like your beloved clouds for instance.
P. Solar,
The applied glacial rebound adjustment for sea level is ~+0.3 mm per year, much less than the measured rise. I agree that it would be better if published values for sea level rise always included a clear explanation that the rate of rise relative to the average shoreline will always be 0.3 mm/year lower due to the rebound adjustment.
But rebound is real. Rebound has be directly measured in lots of places, and is especially easy to measure where thick glaciers were present ~18,000 years ago. The upward movement in these regions was directly measured more than 200 years ago; no need for computer modeling. Direct observation of glacial rebound took place in the 18th century (http://en.wikipedia.org/wiki/Post-glacial_rebound#Discovery). Glacial rebound was reasonably well understood and broadly accepted long before “global warming” ever became a political issue (http://www.cr.nps.gov/history/online_books/geology/publications/pp/754a/). Modeling takes into account the reality that the physical volume of the Earth is not subject to significant change; if land is measurably rising in some places, it must be falling elsewhere to maintain constant volume, and also takes into account the measured rebound rates in formerly glaciated regions. The rate of rebound near Hudson Bay in Canada is still ~12.5 mm per year…. easily measured by looking at how Hudson Bay’s shoreline changes over relatively short periods (https://www.dmr.nd.gov/ndgs/ndnotes/Rebound/Glacial%20Rebound.htm). Glacial rebound is nothing new, and in no way the result of global warming research.
With respect to rebound interfering with mass based measurements (like GRACE satellite measurements): Based on multiple GPS altitude sensors mounted on land (not ice), different parts of coastal Greenland show quite different rates of uplift; parts of western Greenland are actually falling, while on average Greenland is rising modestly (~1.5 mm per year, more or less). Applying GPS based altitude measurements to correct for rebound sounds reasonable to me. GRACE based mass loss in Greenland does have considerable uncertainty, but the measured mass change appears consistently much larger than the rebound adjustment.
Argo measures changes which are considerably larger than 0.0005C, at least in the upper 700 meters. But that aside, the individual profile measurements are (of course) much more noisy than required to measure very small temperature changes. However, many thousands of measurements are more accurate on average, and can show very small changes in average temperature profiles. Argo data shows a consistent seasonal pattern of heat content (the seasonal pattern is often removed from the data so that the longer term trend is easier to see). Were Argo actually so terribly inaccurate, a consistent seasonal pattern would seem impossible to find in the data. The Argo measurements were preceded by decades of temperature profile measurements, most taken long before global warming was ever an issue. The main motivation for taking many thousands of temperature profiles was to characterize how military sonar waves would be influenced by the temperature (and so density) profile. There is no reason to believe that temperature profile data from the 1950’s, 60’s, and 70’s was somehow compromised by a desired outcome.
stevefitzpatrick says: There are at least four components which impacted sea level in the last 150 years: glacial melt, thermal expansion (ocean heat accumulation), reservoir volume increase, and ground water pumping.
Don’t forget hydroisostasy, the ongoing subsidence of the ocean floor due to the added water weight post the continental icesheets melting.
I used a CUMSUM analysis of the detrended satellite measurements, I find that the rate of increases falls in mid-2007 and crashes in fall-2010.
http://i179.photobucket.com/albums/w318/DocMartyn/sealevel.jpg
Could you tell me where you got your txt file for the long tide gauge record from?
Given that David has used the same cumulative technique i developed in 2009, I’m wondering if he has yet connected the dots and realized that the deep drops in temperature he has been forecasting on the basis of a deep solar minimum need to be revisited in the light of realisation that solar energy is retained and released from the ocean on long time scales. As such, it provides a buffer against the Sun being quiet for a few cycles.
Winter air temperatures are likely to be low, but I doubt annual average land surface temperatures will be as deeply affected as he previously forecast. Mind you, quite a lot of the Can/Am border is a long way from the ocean…
I’ve just taken a close look at the rate of change from the paper linked in the article.
http://i39.tinypic.com/jqss1z.png
I’ve measured the peaks and troughs on the graph and marked the midpoint of each rise and fall and joined all the peaks and all the troughs.
The midpoints and the peaks are definitely turn down. It should be noted that the rate of change on the graph is still positive Acceleration is
One interesting point is how straight the base line is, there are three lines there but they are indistinguishable. That base line slope is about 0.9mm/year/century and is steeper than the slope of the midpoints and the peaks, so I think their value is a bit high. It also seems that linear acceleration is already dramatically reduced, so the assumption that the same thing will go on for the next 100 years seems unlikely.
Now mantle rebound will be a fixed rate of change on this timescale and that is a very straight line so I’m going to try to find out how much of that slope is due to hypothetical rebound values.
rgbatduke says:
April 22, 2012 at 10:34 am
over a longer timescale the sunspot records span multiple humans, methods, tools and are difficult to compare to contemporary records on a quantitative basis, are they not?
Solar activity also causes ultraviolet emission which in turn creates our ionosphere. The daily rotation of the Earth means that the ionosphere is heating during the day and not during the night. This sets up an electric current whose magnetic effect we can measure on the ground. This effect was discovered way back in 1722 and can be used to calibrate the sunspot number independently of observers, instruments, etc., see e.g. http://www.leif.org/research/Rudolf%20Wolf%20Was%20Right.pdf
I’m curious, though, how you extend the conclusion that the 20th century was exceptional in solar activity back over the entire Holocene.
I was trying to say it was NOT exceptional.
Is this junk science? If so, how do we know — what is wrong with the reconstructions?
The recent sunspot record is used to calibrate the radionuclide proxies so any problem with the recent data feeds into the reconstructions. This is such an important problem that I’m convening a workshop in a couple of weeks to study this: http://www.leif.org/research/Svalgaard_ISSI_Proposal_Base.pdf
The cosmic ray experts also disagree as to what the proxies show, e.g. http://www.leif.org/EOS/muscheler07qsr.pdf
Mike Dubrasich,
I am aware of how glacial rebound (and ocean floor fall) influence sea level. It is accounted for by a ~0.3mm/yr adjustment. Still, “at least four components” would seem consistent with the existence of Earths accommodation of changing mass distribution.
The cosmic ray experts also disagree as to what the proxies show, e.g. http://www.leif.org/EOS/muscheler07qsr.pdf
rgbatduke says:
There was a lovely paragraph in one of the articles where it was pointed out that in many winters, a heavy snowfall can blanket a geologically signficant fraction of e.g. North America, greatly increasing its albedo. Yet, instead of reflecting enough heat to trigger a return to ice age conditions, negative feedbacks almost instantly kick in, the snow melts, and things return to “normal”.
In a system dominated by rapid negative feedbacks, it is a puzzle as to how the climate changes at all. Yet we know the MWP and LIA occurred.
FWIIW, the conclusion I came to was that climate change on a decade to millenium scale occurs through the water based feedbacks themselves. Specifically, factors that affect the phase changes of water, aerosols, particulates and GCRs. They operate by changing surface and planetary albedo (clouds) and the speed of the hydrological cycle.
@Leif
How does the rabbit know it does not like to be eaten? I am sure it has no prior experience of being eaten. I would postulate that rabbit simply recognizes the immanent threat from another animal that is not a rabbit.
The dog? It just chases anything.