Leif Svalgaard writes in comments:
We plan to submit tomorrow to JGR the following…
http://www.leif.org/research/IDV09.pdf (preprint)
…showing the run of the heliospheric magnetic field since 1835 [not a typo]. I plan to discuss the whole peer-review process here on WUWT, complete with nasty comments by the reviewers and our responses. This will be an illustration of the peer-review process as it unfolds. Should be interesting.
I’ll say. I’ve taken some of the most interesting graphics and put them up for WUWT readers, along with the abstract.
IDV09 and Heliospheric Magnetic field 1835-2009
Leif Svalgaard1 and Edward W. Cliver2
Stanford University, HEPL, Cedar Hall, Via Ortega, Stanford, CA 94305-4085
Space Vehicles Directorate, Air Force Research Laboratory, Hanscom AFB, MA 01731-3010
Abstract.
We use recently acquired archival data to substantiate and extend the IDV index of long-term geomagnetic activity, particularly for years from 1872-1902 for which the initial version of the index (IDV05) was based on observations from very few stations. The new IDV series (IDV09) now includes the interval from 1835-2009, vs. 1872-2004 for IDV05. The HMF strength derived from IDV09 agrees closely with that based on IDV05 over the period of overlap. Comparison of the IDV09-based HMF strength with other recent reconstructions of solar wind B yields a strong consensus between the series based on geomagnetic data, but significant lack of support for a series based on the 10Be cosmic ray radionuclide.
The reconstructed data in the graphic below, from the paper, is quite interesting. Currently, we appear to be at the lowest point in the record.
Click for larger images.
Here’s the comparison with the Be10 isotope record:
L Bowser 06:11:22
Saying it should be Leaf was just my little joke.
Leif Svalgaard 07:14:00
I didn’t know Lief meant “dear” in Dutch; thanks for pointing that out, darling.
Leif Svalgaard (07:35:52) :
tallbloke (06:20:26) :
I am really, really pleased to see Leif and Edward’s reconstruction agrees with the ~65-70% increase in B from 1900 to 1990 found by Mike Lockwood et al 2009
What has happened is that Lockwood has changed his values so that they now reflect the ‘facts’ [our reconstruction] because he has seen the light and abandoned his old method for the ones we have pioneered. Another point is the ‘rise’ from 1900 to 1990. Cycle averages are as follows:
14 5.2
15 6.0
16 5.9
17 6.6
18 7.1
19 7.1 <== 37% higher than cycle 14
20 6.2
21 7.0
22 7.0
23 6.3
24 5.0 <= this is a guess [a prediction if you like]
As the graphs shows there is a 10[0]-yr 'wave' with a maximum [37% higher] near the middle [say cycle 19] and low value at both ends [1901 and 2009].
"This fits well with my own model"
Leif’s law: data is good if it fits well with my own model 🙂
tallbloke (06:59:07) :
OT UAH anomaly is out for sept. 0.42C
That’s what no spots do 🙂
Always teasing. 🙂
The currently high SST’s and consequently the high LT temps fit with my theory regarding the oceans going into heat release mode when sunspots are low.
The data is good anyway, but as I pointed out, my model is ‘scalable’ and can accomodate alterations to data and new discoveries about climate factors, at the expense of certainty, which I see as a chimaera anyway.
You have a 35% rise from cycle 14 to 22, Lockwood et al 67%. This discrepancy is due to the difference between your reconstruction and theirs around cycle 14. The sunspot numbers seem to agree better with their version, but I’ll await your sunspot numbers revision and revisit this point. Any comment about the size of the cycle 14 discrepancy in the light of the fact that Lockwood has adopted your method?
“Leif Svalgaard (07:39:52) :
Bill Illis (06:52:52) :
Your newest data continues that some small correlation. Any thoughts about that?
Since we are now back [as far as the Sun is concerned] to conditions of 108 years ago, one might ask if our climate is too. I don’t think it is.”
My two cents: however, wasn’t the climate warming up from a colder period (LIA) and now we’re cooling (slightly) from a warmer currrent period? Perhaps we are now at a crossroads, so to speak and in a few years’ time we will see more correlation with climate from 100+ years ago.
OT for a minor grammar lesson: “Its” is the possessive pronoun and “It’s” the conjunction for “it is.” When you’re not sure which one use, substitute “it is” in your sentence. If the sentence does not make sense, use “its.” If the sentence does make sense, use “it’s.” Lesson over.
Can someone please explain what the potential implications of this paper are for the AGW hypothesis. I fear that progress being made in alternative theories of AGW are being ignored by the mainstream media due to their complexity and a lack of an layman level explanation. There may also be a bias but we should not always assume that! If such explanations already exist then can you please post a link. Thanks.
Ed Scott (08:26:37) :
In mid-August the University of East Anglia’s Climate Research Unit (CRU) disclosed that it had destroyed the raw data for its global surface temperature data set because of an alleged lack of storage space.
It is ironic, professor H. H. Lamb, most notable 20th century climatologist (creator of the famous temperature graph), left the UK Meteorological Office to found the Climatic Research Unit at the University of East Anglia !
Thank you Leif.
A very interesting extension to the data series.
I find your long range graph of B(IDV) on page 9 really fascinating.
Especially where cycle 20 is completely below your green trend line.
Your interpretation is interesting:
Cycle 23 looks remarkably like cycle 13, including the very deep solar minimum following both cycles, likely presaging a weak cycle 24 as predicted from the solar polar fields [Svalgaard et al., 2005]. It is clear that we are returning to conditions prevailing a century ago.
However, my perspective, pattern matching and interpretation is slightly different… which reflects the saying that beauty is in the eye of the beholder… so please have patience….
For example:
If I was looking for a “match” for cycle 13 then I would choose cycle 22 because of the strong double peak and long decline.
If I was looking for a “match” for cycle 23 then I would choose cycle 19 because of the strong double peak and long rounded end to the decline. If this means anything then it might indicate we are in for a re-run of cycle 20 but at a lower level…
My differences in perspective also come from looking at your green trend line.
The recent peak, circa 1980, is well below the previous peak (sometime before 1835) and the subsequent downward gradient is relatively steep.
Cycle 13 was at the bottom of a trough in the trend line.
Cycle 23 is just after a peak in the trend line…
So, overall, I think you will have to look further back in the records to find a match for cycle 23 because it looks like the trend line is heading further down into uncharted territory.
Leif Svalgaard (07:35:52) :
tallbloke (06:59:07) :
OT UAH anomaly is out for sept. 0.42C
That’s what no spots do 🙂
Well, the ocean heat content was a lot lower in 1900 and had been fairly flat since the mid 1800’s before dipping from the 1890’s according to my model. I think we will see what no or low numbers of spots do in terms of temperature in another 7-10 years.
bill (02:23:14) :
if you are refering to the cyan line then isn’t this a 200year cycle?
The cyan line has a minimum at 1900 and about now. A cycle is defined from one phase to the same phase hits again, e.g. from min to min, so from 1900 to now is about 100 years, not 200.
I think we should wait for you to extend the data back to 1800 and see what happens over the next couple of years before we make any firm judgement about that. The rapid drop in solar activity (judging by sunspot numbers) around 1803 and the current extended minimum might show a more definite ‘cycle length’.
Leif Svalgaard (07:14:00)
The cyan line has a minimum at 1900 and about now. A cycle is defined from one phase to the same phase hits again, e.g. from min to min, so from 1900 to now is about 100 years, not 200.
No problem with that!
However there is really only one actual max and one actual min on the plot so it is difficult to get min-min or max-max
My 200 year was assuming min to max was 0.5 cycle 1900-1980 (only true if symmetry rules) ie 1 cycle = 160 years = 200 years if you squint a bit!
tallbloke (08:47:19) :
You have a 35% rise from cycle 14 to 22, Lockwood et al 67%. This discrepancy is due to the difference between your reconstruction and theirs around cycle 14.
The sunspot numbers seem to agree better with their version
No, if you actually did a real comparison you’ll see that they don’t. Here is such a real comparison: http://www.leif.org/research/B-IDV-Rz-LRF-Obs.png The oval shows cycle 14. We can determine B from IDV (blue curve), from the sunspot number Rz (green curve) and Lockwood’s method (purple curve). It is clear that Lockwood’s data [purple] deviates most from the Rz-derived B. BTW, you can see the differences between the blue and green curves in the early part of the plot, stemming from me using the official, too low, sunspot number. Anyway, your comment is typical for how you see what you want to see without actually looking carefully and doing a minimum of analysis.
Any comment about the size of the cycle 14 discrepancy in the light of the fact that Lockwood has adopted your method?
Well, he has adopted his own variant of my method. If he was using my method 100% he should find the the last decimal place exactly what I have. If you actually READ his paper you can see he has his own indices. The important point is that the ‘principles’ of the analysis are the same: using two different indices with different dependencies on solar wind properties. That insight of mine is what he has acknowledged as valid and fully embraced.
Bill Illis (06:52:52) :
My two cents: however, wasn’t the climate warming up from a colder period (LIA) and now we’re cooling (slightly) from a warmer currrent period? Perhaps we are now at a crossroads, so to speak and in a few years’ time we will see more correlation with climate from 100+ years ago.
It sounds like you want it both ways: solar activity responsible for LIA and also for shorter term variations on top of that. And what we’ll see in some time cannot be used as an argument for what we are seeing now.
“Its” is sometimes just fast typing omitting [‘] by mistake.
Tim Cullen MalagaView (09:03:45) :
I find your long range graph of B(IDV) on page 9 really fascinating. Especially where cycle 20 is completely below your green trend line.
Cycle 20 is indeed an interesting case.
So, overall, I think you will have to look further back in the records to find a match for cycle 23 because it looks like the trend line is heading further down into uncharted territory.
http://www.leif.org/research/SC23%20is%20like%20SC13.png
shows that SC23 was much like SC13. SC24 is predicted to be like SC14, so I don’t think there is any evidence that we are headed for uncharted territory.
tallbloke (06:59:07) :
I think we should wait for you to extend the data back to 1800 and see what happens over the next couple of years before we make any firm judgement about that. The rapid drop in solar activity (judging by sunspot numbers) around 1803 and the current extended minimum might show a more definite ‘cycle length’.
Yet you make firm judgments the other way. As I show in this comment, SC23 is much like SC23, and SC24 is getting to be a year old already, so all evidence is just a repeat of 100+ years ago. [except the climate is not doing that].
lines 199-200: “A 4th-order polynomial fit suggests a ~100 year Gleissberg cycle.”
I won’t ask why you fit a polynomial, but I will ask why you didn’t fit a sine curve since that has some theoretical basis. Would it make any sense to fit a curve to the minimums instead to the whole series?
The possibility, which this data appears to be suggesting, is that the sun’s magnetic field doesn’t change much on centennial timescales. This is, to say the least, troubling, because it makes it very difficult to understand how past changes in climate on those time scales could have occurred-what might cause them? Apparently not the sun, so it has to be something mysterious.
Say Leif, is it you experience that new scientific discoveries seem to raise as many if not more questions than they answer? The Universe is puzzling indeed.
tallbloke (06:59:07) :
As I show in this comment, SC23 was much like SC13,
We mus take heart that Phil Jones’ “VALUE-ADDED” version remains.
This deserves a nomination for quote of the week/ month.
Please excuse the interruption. Congratulations, Leif, on your submission.
Looks very interesting, and I look forward to seeign the whole shebang plus Leif’s inputs on the PR process.
Just a casual look at the first graph, and I see a passing resemblance to my mental picture of the sunspot counts over the same time interval. Notably the steady rise of the peaks up to the all time max reached during the IGY of 1957/58, and the following historically high counts; although the peak following the IGY in Leif’s magnetic field data is surprisingly subdued but then revovers to something like what the susnpot peaks look like.
So the real interesting thing is whether the present downward trend is just a short term anomaly or whether we can expect the general downward behavior to persist for some decades yet. Well I’m not in a predictive mood; but it does beg the question as to whether we may not like what we may be about to receive.
Thanks for sharing with us at this point Leif. time to shift the mental focus from the ice to the fire.
Hopefully you can also educate us Leif on what you solar Physicists think this all means.
George
Leif Svalgaard (09:56:35) :
tallbloke (06:59:07) :
As I show in this comment, SC23 was much like SC13,
And on this basis you predict cycle 24 will be similar to cycle 14. Fair enough, we’ll see how it turns out.
However I can see plenty of other cycles which show similarities to each other, but the following ones don’t. The sun’s activity isn’t regular, although some degree of similarity can be seen between some runs of cycles. It’s as if there were several simultaneous cycles occurring, which sometimes amplify and sometimes cancel each others effects. This suggests to me there are various resonances of differeng amplitudes and frequencies occurring which affect solar activity.
Time and effort will tell us more.
complete with nasty comments by the reviewers
Can the WUWT reviewers join in? ;~P
Moliterno (03:54:04) :
I’m actually kinda curious about the typical pronunciation of the name in English, Swedish, and other languages.
Danish and Norwegian are pretty close. In Norway Leif is a common name and is pronounced approximately like english speakers would pronounce ‘safe’ (just substitute the s with an l) .
Leif Svalgaard (09:54:14) :
http://go2.wordpress.com/?id=725X1342&site=wattsupwiththat.wordpress.com&url=http%3A%2F%2Fwww.leif.org%2Fresearch%2FSC23%2520is%2520like%2520SC13.png
shows that SC23 was much like SC13. SC24 is predicted to be like SC14, so I don’t think there is any evidence that we are headed for uncharted territory.
When we are crystal ball gazing regarding SC24 then the playing field is a lot more level…. so please forgive my presumption in replying.
Although there are clear similarities between SC13 and SC23 (as shown in your linked diagram above) there are also differences. This is especially marked in the final transition years of SC23 which form a curved line compared to the straight lines of SC13. This transitional difference indicates (to me at least) that SC24 will be different to SC14… and curved transitions are not common in your graph… so my guess (for that is all we can do) is that SC24 will be similar to SC20 but at a lower level… but this speculation is the side show…
The main point is clearly demonstrated by the green trend line where it is clear that we are not returning to conditions prevailing a century ago because the trend line has already dipped below the level seen a century ago.
Given the smoothness of the green trend line it is apparent that something quite remarkable will have to occur if we are to return to the conditions prevailing a century ago.
Additionally, if this downward trend continues then we will be literally in uncharted territory i.e. below the graphed levels for the period 1835 through 2009.
Leif Svalgaard (09:54:14) :
tallbloke (08:47:19) :
You have a 35% rise from cycle 14 to 22, Lockwood et al 67%. This discrepancy is due to the difference between your reconstruction and theirs around cycle 14.
The sunspot numbers seem to agree better with their version
No, if you actually did a real comparison you’ll see that they don’t. Here is such a real comparison: http://www.leif.org/research/B-IDV-Rz-LRF-Obs.png The oval shows cycle 14. We can determine B from IDV (blue curve), from the sunspot number Rz (green curve) and Lockwood’s method (purple curve).
According to this plot of sunspot numbers, cycle 14 had around 1/3 of the amplitude of cycle 19.
http://www.woodfortrees.org/plot/sidc-ssn/from:1860/to:1960/mean:12
But your Rz sunspot curve shows cycle 14 being around 3/4 the amplitude of cycle 19.
I think this is where the discrepancy arises between your interpretation and mine. You are perhaps using your ‘revised’ sunspot numbers, which of course, agree better with your revised magnetic data. However, not all solar physicists agree with your sunspot number revisions. It looks like we mauy be in danger of getting into semantic arguments about the context of the word ‘real’. 😉
Leif Svalgaard (09:54:14):
It sounds like you want it both ways: solar activity responsible for LIA and also for shorter term variations on top of that. And what we’ll see in some time cannot be used as an argument for what we are seeing now.
Coldest part of the Little Ice Age is usually linked with the Maunder minimum 1650 – 1720 and the temperature charts associated with American scientist J. Eddy.
In Europe, the temperature chart produced by climatologist H. Lamb shows that coldest period was nearly over as the solar Maunder minimum is starting. The coolest period was apparently 1600-1650, when there is a reasonable record of the sunspot activity. I am not sure if H. Lamb was aware of the Maunder minimum when he produced his charts, it was first published in 1965 and has been updated several times since. The IPCC used the graph in their early publications.
J. Eddy is credited with discovering the Maunder minimum, his chart was used in The IPCC Assessment, Cambridge University Press, Cambridge, 1990.
It could be speculated that both are right in their own way, since the charts should be considered as regional rather than global.
http://www.vukcevic.talktalk.net/LIA.gif
What does the current science have to say ?
Is there some significance known to the sinusoidal average with a period at 1/16th of the sunspot period ?
bill (09:46:57) :
My 200 year was assuming min to max was 0.5 cycle 1900-1980 (only true if symmetry rules) ie 1 cycle = 160 years = 200 years if you squint a bit!
The Gleissberg cycle is traditionally ~90 years. The 100-yr ‘cycle’ was noticed by my little [five at the time] grandson Peter when he saw this plot: http://sidc.oma.be/html/wolfaml.html the cycles are low at the left- and right-hand side…
Toto (09:55:33) :
I won’t ask why you fit a polynomial, but I will ask why you didn’t fit a sine curve since that has some theoretical basis.
The whole idea of fitting is not really physical because we don’t know what the curve form should be. To be honest I use the 4th-degree pol for easy of drawing the ‘wave’ and it has no further significance. In the old days, I would just have drawn it by free-hand. I mention the 4th pol because if I didn’t, the referee [and many of you] would ask: “how did you draw the wave?”
Andrew (09:56:12) :
The possibility, which this data appears to be suggesting, is that the sun’s magnetic field doesn’t change much on centennial timescales. This is, to say the least, troubling, because it makes it very difficult to understand how past changes in climate on those time scales could have occurred-what might cause them? Apparently not the sun
I think that has been a leitmotif through all my postings here at WUWT and elsewhere.
Say Leif, is it you experience that new scientific discoveries seem to raise as many if not more questions than they answer? The Universe is puzzling indeed.
Yes, that is always the way it is: the more we know, the more good questions we can ask.
George E. Smith (10:32:36) :
Just a casual look at the first graph, and I see a passing resemblance to my mental picture of the sunspot counts over the same time interval.
That is, indeed, the case. See section 7 of http://www.leif.org/research/The%20IDV%20index%20-%20its%20derivation%20and%20use.pdf
tallbloke (10:43:43) :
And on this basis you predict cycle 24 will be similar to cycle 14. Fair enough, we’ll see how it turns out.
No, not at all. This http://www.leif.org/research/AGU%20Fall%202008%20SH51A-1593.pdf and this http://www.leif.org/research/Cycle%2024%20Smallest%20100%20years.pdf is the basis.
Summarized here: http://www.leif.org/research/Predicting%20the%20Solar%20Cycle.pdf
Carsten Arnholm, Norway (10:53:33) :
Danish and Norwegian are pretty close. In Norway Leif is a common name and is pronounced approximately like english speakers would pronounce ’safe’ (just substitute the s with an l) .
And in Danish, Leif is pronounced ‘lyf’. When I worked in Japan, they called me ‘raifu’ which the closest they could get to ‘leif’.
Tim Cullen MalagaView (10:59:37) :
The main point is clearly demonstrated by the green trend line[…] Additionally, if this downward trend continues then we will be literally in uncharted territory i.e. below the graphed levels for the period 1835 through 2009.
As I explained, the trend line is in sense bogus and should not be extrapolated too much. It was only to show my interpretation of the Gleissberg ‘wave’. Others would say there are two waves, some would say none, just random, etc. Don’t extrapolate a polynomial fit. We had a discussion about that on WUWT a while back, when the issue was if a cooling has begun.
tallbloke (11:09:23) :
But your Rz sunspot curve shows cycle 14 being around 3/4 the amplitude of cycle 19.
No, again, look at it carefully. My plot does not show Rz, but B calculated from Rz. The formula is B = 4.3 + 0.3 * SQRT(Rz).
However, not all solar physicists agree with your sunspot number revisions.
That is moot, as the original, official number was used.
Vukcevic (11:12:10) :
What does the current science have to say ?
There is little detailed correlation between solar activity and temperature. e.g. http://www.leif.org/research/Loehle-Temps-and-TSI.png The LIA lasted several centuries and [as you point out] the coldest temperatures were at the high solar activity when Galileo [and others] first observed the spots. If you look back through the record the past 2000 years [shown in my link above] it is hard to get the notion that there is any relationship at all.
Neo (11:36:42) :
Is there some significance known to the sinusoidal average with a period at 1/16th of the sunspot period ?
Not sure what you are asking. On its face, I would say no.
Whenever I write Leif’s name and I’m wondering is it ‘e’ before ‘i’ or the normal ‘i’ before ‘e’ except after ‘c’, I remember it by thinking ‘Leif is abnormal.’ 😉
tallbloke (09:04:13) : Well, the ocean heat content was a lot lower in 1900 and had been fairly flat since the mid 1800’s before dipping from the 1890’s according to my model. I think we will see what no or low numbers of spots do in terms of temperature in another 7-10 years.
Excellent! No matter what caused the little ice age, it takes a long time to heat the oceans.
Andrew (09:56:12): The possibility, which this data appears to be suggesting, is that the sun’s magnetic field doesn’t change much on centennial timescales. This is, to say the least, troubling, because it makes it very difficult to understand how past changes in climate on those time scales could have occurred-what might cause them? Apparently not the sun, so it has to be something mysterious.
If you think that a reduced magnetic field instantly will heat our planet, it would be equivalent to state that a cold cabin can be instantly heated from -10 C to +20 C – it’s not going to happen! The thermal mass and the thermal time constant of the oceans is huge, heating takes centuries! The statistical variation in the magnetic field from 1850 to 2009 is quite dramatic:
min 4.1800 nT
max 9.6000 nT
mean 6.3446 nT
var 1.3959 nT
But we have to very patient to see the ocean temperature change if it’s being influenced by the magnetic field (I did not say it is!), since everything that changes the energy in the oceans has to go through the first law of thermodynamics,
m•cp•dT/dt = Qin – Qout
People seem to think that the equation for the temperature in the oceans is not a differential equation, but some simple algebraic equation:
T(t) = A + B + C + D
However, that’s not the case, the correct equation is a first order differential with a huge time constant. Is this difficult to understand? The differential equation means that it takes a longer time to heat larger objects with large density and large heat capacity than smaller objects with small density and small heat capacity. Does it take a long time to heat your cabin? Imagine the oceans! And it also means that one has to add up the contributions to the energy and temperature over time, since the temperature of a large body can never be changed instantly. Please read Wikipedia:
Contrary to other thermodynamic quantities such as entropy and heat, whose microscopic definitions are valid even far away from thermodynamic equilibrium, temperature being an average energy per particle can only be defined at thermodynamic equilibrium, or at least local thermodynamic equilibrium (see below).
As a system receives heat, its temperature rises; similarly, a loss of heat from the system tends to decrease its temperature (at the—uncommon—exception of negative temperature; see below).
http://en.wikipedia.org/wiki/Temperature
So I give up ! I just pored over my table of the isotopes, trying to find some animal that decays to 4Be10.
3Li6 and 3Li7 aren’t going up there for anybody, and what would they be doing in the atmosphere ?
So maybe 4Be9, the normal isotope can capture a neutron; but then what the blazes is it doing in the atmosphere either. So maybe 5B10 can capture a neutron, and then do an ordinary beta decay, but the table shows no such decay for 5B10; which also has no business being in the atmosphere.
So getting wilder I looked for an alpha emission from 6C14; but no such luck and not much of that either in the atmosphere.
So what am I missing; how do I get from Nitrogen or Oxygen to 4Be10 ?
Enquiring minds want to know !
George