Guest post by David Archibald
Successful prediction of levels of solar activity suggests that prediction of other phenomena driven by solar activity might also be successful, and useful. Sea level rise is a concern of some people. President Obama said in June 2008 that his nomination in the Democratic primaries was “the moment when the rise of the oceans began to slow”.
The above graph shows the satellite data from the University of Colorado from late 1992. A change of trend is evident in 2004. Prior to that, sea level was rising at 4.2 mm/annum, and after 2004 at 1.5 mm/annum. 2003 was the recent peak in solar activity in terms of flares, F10.7 flux and proton flux. It is likely that the lower rate of rise post 2004 is due to lower subsequent solar activity.
The CSIRO compiled tide gauge data from 1870. The graph above shows that data with the subsequent satellite data plotted together.
The modern retreat of glaciers began in 1860. Initially sea level rose at 1.0 mm/annum. After 1930, it almost doubled to 1.9 mm/annum. This is a well-defined uptrend, now 80 years long.
Our prediction of a 2° C decline in temperature for the mid-latitudes over Solar Cycles 24 and 25 suggests that sea level will stop rising, and should start falling at some point prior to 2032.
The graph above combines the satellite data with the prior ten years of tide gauge data and shows the bounds of the long term rise at 1.9 mm/annum post 1030. Sea level could remain flat for another ten years before that trend in sea level rise is broken.
tallbloke
“The figures aren’t exact, but are in the ballpark.”
Don’t think so http://www.agu.org/pubs/crossref/2004/2003JC002260.shtml
and I’m sure I have seen lower figures but can’t find them at the moment.
Here 0.6 W/m2 is mentioned
http://pielkeclimatesci.wordpress.com/2009/02/09/update-on-a-comparison-of-upper-ocean-heat-content-changes-with-the-giss-model-predictions/
Leif Svalgaard says:
December 1, 2010 at 2:16 pm
tallbloke says:
December 1, 2010 at 2:07 pm
Leif, apologies, my mind was racing ahead and my words were too elliptical. I’m considering the steric component of sea level rise, not the overall rise including melt.
You were directly using David’s graph as mounting ‘evidence’…
Part of the evidence. One more metric with an inflection point around 1930.
John A says:
December 1, 2010 at 11:54 am
I only get to put this sort of stuff out because I am filling a vacuum. You can displace me by doing better work yourself and offering it up for the common weal.
We know that sea level will stop rising in the next ten years or so. Can we predict that year that will be? That is why I looked at the sea level data. I found that making a worthwhile prediction would be very difficult.
tallbloke says:
December 1, 2010 at 12:40 pm
ftp://ftp.marine.csiro.au/pub/white/church_white_grl_gmsl.zip
That is the link to the tide data at the CSIRO website.
tallbloke says:
December 1, 2010 at 1:25 pm (Edit)
Here’s the graph of tropical cloud cover vs temperature from the ISCCP data:
http://tallbloke.files.wordpress.com/2010/11/isccp-temp.jpg
I believe Roy Spencer had a ballpark figure for the solar forcing increase per 1% drop in cloud cover, but I can’t recall it. Anyone?
######
do you accept all the physics required to produce those cloud cover data points?
lgl says:
December 1, 2010 at 3:04 pm (Edit)
Here 0.6 W/m2 is mentioned
http://pielkeclimatesci.wordpress.com/2009/02/09/update-on-a-comparison-of-upper-ocean-heat-content-changes-with-the-giss-model-predictions/
Thanks lgl. I note Pielke Sr is referring to the overall radiative imbalance rather than the forcing on the ocean. I think that given the reduced cloud cover and increased insolation at the surface, we’d also have seen increased outgoing radiation (due to less cloud cover) and so I’m not sure my 4W/m^2 additional forcing on the ocean surface is incompatible with Pielke Sr’s radiative imbalance figure. Maybe I’ve got some terminology wrong and this is what made you and Steve Fitzpatrick think i was referring to radiative balance?
Leif might like to read that article and note that Pielke Sr doesn’t see a problem in offering a W/m^2 figure ‘corresponding’ to a Joules figure. I’ll email Pielke Sr and ask him how he made the calc.
Tallbloke,
“The profile of the temperature falloff to the thermocline is well known, and pretty much linear below the well mixed layer. ”
It approximates linear perhaps for 100 to 200 meters, but the thermocline is pretty much an exponential decay function, as was first explained by Walter Munk in 1966
http://oceanworld.tamu.edu/resources/ocng_textbook/chapter08/chapter08_05.htm
The (very) slow turbulent eddy down-mixing process opposes a continuous upwelling of very cold water from the abyss (the thermohaline circulation), which leads to the exponential decay shape.
re post by; Leif Svalgaard says: December 1, 2010 at 2:11 pm
Hi Leif,
Thanks for your reply and added info w/ link. I wasn’t arguing the main premise (e.g., solar cycles do/don’t cause inflection points in sea level rate of change), just taking issue with the logic used in that one post of yours. You’ll note that elsewhere I argued that there seem to be a number of inflection points similar to 2004 when looked at on that same time frame – and that for the correlation to be potentially meaningful, one would have to be able to show similar inflection points for similar sun cycle changes (or vice versa), along with a lack of said inflection points where there were not sun cycle changes. If that were done, then the correlation would become interesting and worth digging into – even tho it would still just be a correlation.
Leif, if the Earth can be used to measure the Sun’s magnetism, how can variations in the Earth’s field itself be accounted for?
Also, I was under the impression that while sun spot count could be reasonably correlated to our typical measure of irradiance, it had recently been discovered that the level of irradiance isn’t constant across all frequencies as it had been long thought to be – e.g., that while decreasing in the bands we had been typically measuring, it could actually be increasing in other bands and vice versa? Please forgive me if I’ve utterly mucked this up… but wasn’t there some research along these lines recently, with implications regarding how the amount of energy we thought we’d been receiving from the sun based on our measurements may not have been accurate?
tallbloke says:
December 1, 2010 at 3:53 pm
“You were directly using David’s graph as mounting ‘evidence’…”
Part of the evidence. One more metric with an inflection point around 1930
You were saying that the heat content was decreasing from the 1800s until 1930, but David’s graph showed an increase in absolute level; hardly evidence. And solar activity, BTW, was decreasing from the 1870s until 1900s, where does that fit?
tallbloke says:
December 1, 2010 at 4:09 pm
Leif might like to read that article and note that Pielke Sr doesn’t see a problem in offering a W/m^2 figure ‘corresponding’ to a Joules figure.
He says: “700 meters of the ocean of 2.45 * 10**22 Joules per year, which corresponds to a radiative imbalance of ~1.50 Watts per square meter.”.
Joules per year can be converted to Watts. Joules alone cannot. To get to the ‘per square meter’ you must make assumptions about the radiative balance [have a model for it]. I don’t see you having offered one. But perhaps the issue will get resolved once you tell us how you got the 4 W/m2.
Rational Debate says:
December 1, 2010 at 5:33 pm
just taking issue with the logic used in that one post of yours.
My logic holds under the premise of the topic, namely that a few years of solar change can bring about a significant amount of sea level change.
Leif, if the Earth can be used to measure the Sun’s magnetism, how can variations in the Earth’s field itself be accounted for?
They have very different time scales, so don’t interfere with each other.
the amount of energy we thought we’d been receiving from the sun based on our measurements may not have been accurate?
The energy we measure is measured accurately. What we have discovered is that the distribution over wavelength varies with the wavelength [one band offsettting another band] with the result that the total varies a lot less.
“Our prediction of a 2° C decline in temperature for the mid-latitudes over Solar Cycles 24 and 25 suggests…”
Sounds worse than Maunder, how did you come to that conclusion ?
Ulric Lyons says:
December 1, 2010 at 6:00 pm
Two degrees for the US-Canadian border, one and a half for Norway for example.
See this: http://wattsupwiththat.com/2010/07/13/solar-driven-temperature-decline-predicted-for-norway-by-a-norwegian/
“””””
December 1, 2010 at 12:25 am
I think the correct Latin phrase is “post hoc ergo propter hoc”. And its a fallacy. “””””
Well I have heard:- “post hoc non propter hoc”. But I have never heard yours before.
I have a shirt which proclaims “Nullum gratuitum prandium.” Probably some Indian friend of Pachauri.
“”””” Rational Debate says:
December 1, 2010 at 5:33 pm
just taking issue with the logic used in that one post of yours.
My logic holds under the premise of the topic, namely that a few years of solar change can bring about a significant amount of sea level change.
Leif, if the Earth can be used to measure the Sun’s magnetism, how can variations in the Earth’s field itself be accounted for? “””””
Don’t they measure solar magnetism, as well as that of other stars using the Zeeman effect of hyperfine splitting of spectral lines. There’s one kind that splits one line into two equispaced about the original, and the other gives you back the central line as well. The amount of split (probably frequency or wave number) is proportional to the magnetic field, and I guess which set of lines you get, depends on whether you are looking parallel to the magnetic field or perpendicular to it. I think the two satellite lines are also plane polarised perpendicular to each other.
Well don’t quote me on that; you better look it up somewhere (other than wikipedia).
George E. Smith says:
December 1, 2010 at 7:29 pm
Don’t they measure solar magnetism, as well as that of other stars using the Zeeman effect of hyperfine splitting of spectral lines.
Yes, that is one way. I am co-builder of just such a solar observatory: http://www.leif.org/EOS/1977SoPh-54-353.pdf
But there is another way: the solar wind drags the Sun’s magnetic field out into interplanetary space where it interacts with the Earth’s magnetic field. We understand the nature of this interaction and can use the response of the Earth’s magnetic field to measure this solar magnetic field, e.g. http://www.leif.org/research/IAGA2008LS-final.pdf
It would be nice to hear the input of one of the world’s most preeminent sea-level experts on this issue.
I have emailed and alerted Nils-Axel Morner on this and hopefully he will join in this discussion!
Chris
Norfolk, VA, USA
An Engineer says:
By what mechanism, backed up by empirical data, is this supposed to work?
Rational Debate says:
It is an inriguing correlation, but what possible mechanism could tie solar changes to sudden sea level rate changes?
Try crustal deformation. The same thing that causes earthquakes and volcanoes, that also are shown to vary with the “climate cycles”…
http://chiefio.wordpress.com/2010/12/01/of-quakes-and-climate/
The only “hard bit” is that a lot of folks make some very good arguments for why periodic changes of earths rate of rotation (that also correlates with PDO cycles and solar cycles – paper by Ian Wilson IIRC) can’t be driven by solar motion changes…
But that we have correlation of Length of Day, PDO, Solar cycles of motion, ENSO, volcanoes and earthquakes is pretty much being confirmed. But “correlation is not causality”…
Until someone works out how planets and the sun can have “spin orbit coupling” just like sub atomic particles do, we are left to admire the correlations and repeat the mantra … not causality not causality not causality…
http://en.wikipedia.org/wiki/Spin–orbit_interaction
Why it happens at sub atomic levels yet is not to be considered at macro levels is left as an exercise for the physicists among us…
E.M.Smith says:
December 1, 2010 at 9:02 pm
Until someone works out how planets and the sun can have “spin orbit coupling” subatomic particles have because of the strength of the electromagnetic forces that ‘keep them in orbit’. Neither the Sun nor the planets are electrically charged and their gravitational forces are much too weak [the distances too large] for any spin-orbit coupling. Move the planets 100 times closer to the Sun, the spin-orbit coupling will increase a million-fold and there will be some effects. There exist planetary systems where this is happening. just not ours.
E.M.Smith says:
December 1, 2010 at 9:02 pm
But that we have correlation of Length of Day, PDO, Solar cycles of motion, ENSO, volcanoes and earthquakes is pretty much being confirmed. But “correlation is not causality”…
Until someone works out how planets and the sun can have “spin orbit coupling” just like sub atomic particles do, we are left to admire the correlations and repeat the mantra … not causality not causality not causality…
Yes. But.
The spin orbit coupling that was proposed by he who cannot be named on this site was not between the planets and the Sun directly, but between the irregular motion of the Sun’s own orbit about centre of mass of the solar system, and it’s own spin.
The orbit of the Sun is of course determined by the planets, mostly the gas giants.
I won’t derail this thread with discussion of this, so if people are interested they should look through the archives here and on Geoff Sharp’s and my blogs.
Leif Svalgaard says:
December 1, 2010 at 5:54 pm (Edit)
tallbloke says:
December 1, 2010 at 4:09 pm
Leif might like to read that article and note that Pielke Sr doesn’t see a problem in offering a W/m^2 figure ‘corresponding’ to a Joules figure.
He says: “700 meters of the ocean of 2.45 * 10**22 Joules per year, which corresponds to a radiative imbalance of ~1.50 Watts per square meter.”.
Joules per year can be converted to Watts. Joules alone cannot. To get to the ‘per square meter’ you must make assumptions about the radiative balance [have a model for it]. I don’t see you having offered one. But perhaps the issue will get resolved once you tell us how you got the 4 W/m2.
Thanks Leif, you have jogged my memory cell a bit. W=J/s. I remember working out the number of seconds in a decade (315576000), and dividing the ~8*10^22J increase in ocean heat content by that number. Then dividing that by the number of square metres on earth’s surface (5.1*10^14). But because the amount of energy which went into the ocean was over the whole decade 1993-2003 I’m remembering now that my figure was 8*10^23J not 8*10^22J.
I’ve just crunched it again and I’m getting 4.97W/m^2
I seem to remember you and I agreed 8*10^23 was a bit on the high side after we looked at the uncertainty of the ARGO/XBT data splice so I dropped it to something like 6.5*10^23J when I did the calc. This will probably explain the discrepancy between my 4W/m^2 and the 4.97W/m^2 I’ve just recalculated.
Thanks for your hint and apologies for my post crash mental difficulties causing the confusion.
Steve Fitzpatrick says:
December 1, 2010 at 5:18 pm (Edit)
Tallbloke,
“The profile of the temperature falloff to the thermocline is well known, and pretty much linear below the well mixed layer. ”
It approximates linear perhaps for 100 to 200 meters, but the thermocline is pretty much an exponential decay function, as was first explained by Walter Munk in 1966
http://oceanworld.tamu.edu/resources/ocng_textbook/chapter08/chapter08_05.htm
The (very) slow turbulent eddy down-mixing process opposes a continuous upwelling of very cold water from the abyss (the thermohaline circulation), which leads to the exponential decay shape.
Steve, thanks again. I had been going off a graphic I found along with a description on a different site that was similar to this one.
http://marinebio.org/i/temperature/Thermocline.gif
Using a rough engineers estimate and a thermocline depth of 1000m (not 700 as I said earlier) I came up with my 0.15C the temperature of the top 1000m must have risen to account for the amount of thermal expansion needed to get the steric sea level rise seen in the 1993-2003 decade, equating to ~0.3C increase in temp at the surface. This figure is confirmed by SST observations.
The maths of a more accurate calc are beyond me, so if you’d like to pitch in and help I’d be very grateful.
David Archibald says:
December 1, 2010 at 3:53 pm
ftp://ftp.marine.csiro.au/pub/white/church_white_grl_gmsl.zip
That is the link to the tide data at the CSIRO website.
Thank you David.
Leif Svalgaard says:
December 1, 2010 at 5:40 pm
tallbloke says:
December 1, 2010 at 3:53 pm
Part of the evidence. One more metric with an inflection point around 1930
You were saying that the heat content was decreasing from the 1800s until 1930, but David’s graph showed an increase in absolute level; hardly evidence. And solar activity, BTW, was decreasing from the 1870s until 1900s, where does that fit?
I’ll let you know after I’ve crunched the data David has kindly provided. The absolute level is composed of several factors, and as long as the data is reconciled with relative land height variation, the main two are ice melt and steric changes due to changing insolation *at the surface*.
Solar activity dropped from around 46SSN to 38SSN over the 1870-1930 period. My approximate ocean equilibrium value is around 40SSN but will be affected by cloud cover levels. Given the generally low sunspot number, and Nir Shaviv’s work on using the oceans as a calorimeter, my guess is that it was generally cloudier during that time, especially given the high number of el ninos in the late 1800’s throwing heat and water vapour into the atmosphere, increasing humidity. They will also have depleted the ocean heat content. So although SST’s were up before the ~1910 low point, this was more due to higher atmospheric temps (diminishing the differential between ocean and atmospheric temp) and high SST, than high ocean heat content.
Getting a better grasp of these relationships will help us understand variation in atmospheric temperature and the underlying absorption of solar radiation into the ocean which drives it in a non-linear (on the decadal scale) way.
Steven Mosher says:
December 1, 2010 at 3:59 pm (Edit)
tallbloke says:
December 1, 2010 at 1:25 pm (Edit)
Here’s the graph of tropical cloud cover vs temperature from the ISCCP data:
http://tallbloke.files.wordpress.com/2010/11/isccp-temp.jpg
I believe Roy Spencer had a ballpark figure for the solar forcing increase per 1% drop in cloud cover, but I can’t recall it. Anyone?
######
do you accept all the physics required to produce those cloud cover data points?
Steve, I’m well aware of the problems involved with collating and integrating data from low orbit weather satellites. However, ISCCP data is in reasonably good agreement with completely different methods of calculating Earth albedo (Palle et al), so I think the diligent scientists on the ISCCP project deserve some credit, rather then the generally negative vibes they’ve had from co2 warmists who don’t like the implications of their data.
tallbloke
But ~8*10^22J is for the whole decade, not one year so there’s your *10
http://bobtisdale.blogspot.com/2009/07/ohc-trends-presented-by-levitus-et-al.html
lgl says:
December 2, 2010 at 2:13 am
But ~8*10^22J is for the whole decade, not one year so there’s your *10
tallbloke says:
December 2, 2010 at 12:19 am
Thanks Leif, you have jogged my memory cell a bit. W=J/s. I remember working out the number of seconds in a decade (315576000), and dividing the ~8*10^22J increase in ocean heat content by that number. Then dividing that by the number of square metres on earth’s surface (5.1*10^14). But because the amount of energy which went into the ocean was over the whole decade 1993-2003 I’m remembering now that my figure was 8*10^23J not 8*10^22J.
Levitus has been doing strange things to OHC figures for 10 years. See my comments on Bob’s earlier posts on this issue. e.g. http://bobtisdale.blogspot.com/2009/03/latest-revisions-to-ocean-heat-content.html
Looking at the figures I’m talking of in that thread, my memory is hopelessly confused on this and I’m going to have to go through the calc from the beginning again.