Some people cite scientists saying there is a “CO2 control knob” for Earth. No doubt there is, but due to the logarithmic effect of CO2, I think of it like a fine tuning knob, not the main station tuner. That said, a new data picture is emerging of an even bigger knob and lever; a nice bright yellow one.
A few months back, I found a website from NOAA that provides an algorithm and downloadable program for spotting regime shifts in time series data. It was designed by Sergei Rodionov of the NOAA Bering Climate and Ecosystem Center for the purpose of detecting shifts in the Pacific Decadal Oscillation.
Regime shifts are defined as rapid reorganizations of ecosystems from one relatively stable state to another. In the marine environment, regimes may last for several decades and shifts often appear to be associated with changes in the climate system. In the North Pacific, climate regimes are typically described using the concept of Pacific Decadal Oscillation. Regime shifts were also found in many other variables as demonstrated in the Data section of this website (select a variable and then click “Recent trends”).
But data is data, and the program doesn’t care if it is ecosystem data, temperature data, population data, or solar data. It just looks for and identifies abrupt changes that stabilize at a new level. For example, a useful application of the program is to look for shifts in weather data, such as that caused by the PDO. Here we can clearly see the great Pacific Climate Shift of 1976/77:
Another useful application is to use it to identify station moves that result in a temperature shift. It might also be applied to proxy data, such as ice core Oxygen 18 isotope data.
But the program was developed around the PDO. What drives the PDO? Many say the sun, though there are other factors too. It follows to reason then the we might be able to look for solar regime shifts in PDO driven temperature data.
Alan of AppInSys found the same application and has done just that, and the results are quite interesting. The correlation is well aligned, and it demonstrates the solar to PDO connection quite well. I’ll let him tell his story of discovery below. – Anthony
=================================
Climate Regime Shifts
The notion that climate variations often occur in the form of ‘‘regimes’’ began to become appreciated in the 1990s. This paradigm was inspired in large part by the rapid change of the North Pacific climate around 1977 [e.g., Kerr, 1992] and the identification of other abrupt shifts in association with the Pacific Decadal Oscillation (PDO) [Mantua et al., 1997].” [http://www.beringclimate.noaa.gov/regimes/Regime_shift_algorithm.pdf]
Pacific Regime Shifts
Hare and Mantua, 2000 (“Empirical evidence for North Pacific regime shifts in 1977 and 1989”): “It is now widely accepted that a climatic regime shift transpired in the North Pacific Ocean in the winter of 1976–77. This regime shift has had far reaching consequences for the large marine ecosystems of the North Pacific. Despite the strength and scope of the changes initiated by the shift, it was 10–15 years before it was fully recognized. Subsequent research has suggested that this event was not unique in the historical record but merely the latest in a succession of climatic regime shifts. In this study, we assembled 100 environmental time series, 31 climatic and 69 biological, to determine if there is evidence for common regime signals in the 1965–1997 period of record. Our analysis reproduces previously documented features of the 1977 regime shift, and identifies a further shift in 1989 in some components of the North Pacific ecosystem. The 1989 changes were neither as pervasive as the 1977 changes nor did they signal a simple return to pre-1977 conditions.”
[http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6V7B-41FTS3S-2…]
Overland et al “North Pacific regime shifts: Definitions, issues and recent transitions”
[http://www.pmel.noaa.gov/foci/publications/2008/overN667.pdf]: “climate variables for the North Pacific display shifts near 1977, 1989 and 1998.”
The following figure from the above paper show analysis of PDO and Victoria Index using the Rodionov regime detection algorithm. A regime shift is also detected around 1947-48.
The following figure shows regime shift detection for the summer PDO, showing shifts at 1948, 1976 and 1998.
[http://www.beringclimate.noaa.gov/data/Images/PDOs_FigRegime.html]
(For detailed information on the 1976/77 climate shift,
see: http://www.appinsys.com/GlobalWarming/The1976-78ClimateShift.htm)
Regime Shift Detection in Annual Temperature Anomaly Data
The NOAA Bering Climate web site provides the algorithm for regime shift detection developed by Sergei Rodionov [http://www.beringclimate.noaa.gov/regimes/index.html]. The following analyses use the Excel VBA regime change algorithm version 3.2 from this web site.
The following figure shows the regime analysis of the HadCRUT3 annual global annual average temperature anomaly data from the Met Office Hadley Centre for 1895 to 2009 [http://hadobs.metoffice.com/hadcrut3/diagnostics/global/nh+sh/annual].
The analysis was run based on the mean using a significance level of 0.1, cut-off length of 10 and Huber weight parameter of 2 using red noise IP4 subsample size 6. Regime changes are identified in 1902, 1914, 1926, 1937, 1946, 1957, 1977, 1987, and 1997. Running the analysis based on the variance rather than the mean results in regime changes in the bold years listed above.
Regime Shift Relationship to Solar Cycle
The NASA Solar Physics web site provides the following figure showing sunspot area.
[http://solarscience.msfc.nasa.gov/SunspotCycle.shtml]
The following figure compares the Hadley (HadCrut3) monthly global average temperature (from [http://hadobs.metoffice.com/hadcrut3/diagnostics/global/nh+sh/]) overlaid with the regime change line (red line) shown previously, along with the sunspot area since 1900. The sunspot cycle is approximately 11 years. The sun’s magnetic field reverses with each sunspot cycle and thus after two sunspot cycles the magnetic field has completed a cycle – a Hale Cycle – and is back to where it started. Thus a complete magnetic sunspot cycle is approximately 22 years. The figure marks the onset of odd-numbered cycles with a vertical red line, even-numbered cycles with a green line.
From the figure above it can be seen that the regime changes correspond to the onset of solar cycles and occur when the “butterfly” is at its widest. The most significant warming regime shifts occur at the start of odd-numbered cycles (1937, 1957, 1977, 1997). Each odd-numbered cycle (red lines above) has resulted in a temperature-increase regime shift. Even-numbered cycles (green lines above) have been inconsistent, with some resulting in temperature-decrease regime shifts (1902, 1946) or minor temperature-increase shifts (1926, 1987).
An unusual one is the 1957 – 1966 cycle, which in the monthly data shown above visually looks like a temperature-increase shift in 1957 followed by a temperature-decrease shift in 1964 but the regime detection algorithm did not identify it. This is likely due to the use of annually averaged data in the regime detection algorithm.
The following figure shows the relative polarity of the Sun’s magnetic poles for recent sunspot cycles along with the solar magnetic flux [www.bu.edu/csp/nas/IHY_MagField.ppt]. The regime change periods are highlighted by the red and green boxes. Each one occurs on as the solar cycle is accelerating. The onset of an odd-numbered sunspot cycle (1977-78, 1997-98) results in the relative alignment of the Earth’s and the Sun’s magnetic fields (positive North pole on the Sun) allowing greater penetration of the geomagnetic storms into the Earth’s atmosphere. “Twenty times more solar particles cross the Earth’s leaky magnetic shield when the sun’s magnetic field is aligned with that of the Earth compared to when the two magnetic fields are oppositely directed” [http://www.nasa.gov/mission_pages/themis/news/themis_leaky_shield.html]
The following figure shows the longitudinally averaged solar magnetic field. This “magnetic butterfly diagram” shows that the sunspots are involved with transporting the field in its reversal. The Earth’s temperature regime shifts are indicated with the superimposed boxes – red on odd numbered solar cycles, green on even.
[http://solarphysics.livingreviews.org/open?pubNo=lrsp-2010-1&page=articlesu8.html]
The Earth’s temperature regime shift occurs as the solar magnetic field begins its reversal.
Solar Cycle 24
Solar cycle 24 is in its initial stage after getting off to a late start. An El Nino occurred in the first part of 2010. This may be the start of the next regime shift.
Climate Regime Shifts
[last update: 2010/07/04]
|
“The notion that climate variations often occur in the form of ‘‘regimes’’ began to become appreciated in the 1990s. This paradigm was inspired in large part by the rapid change of the North Pacific climate around 1977 [e.g., Kerr, 1992] and the identification of other abrupt shifts in association with the Pacific Decadal Oscillation (PDO) [Mantua et al., 1997].” [http://www.beringclimate.noaa.gov/regimes/Regime_shift_algorithm.pdf]
|
|
Pacific Regime Shifts
Hare and Mantua, 2000 (“Empirical evidence for North Pacific regime shifts in 1977 and 1989”): “It is now widely accepted that a climatic regime shift transpired in the North Pacific Ocean in the winter of 1976–77. This regime shift has had far reaching consequences for the large marine ecosystems of the North Pacific. Despite the strength and scope of the changes initiated by the shift, it was 10–15 years before it was fully recognized. Subsequent research has suggested that this event was not unique in the historical record but merely the latest in a succession of climatic regime shifts. In this study, we assembled 100 environmental time series, 31 climatic and 69 biological, to determine if there is evidence for common regime signals in the 1965–1997 period of record. Our analysis reproduces previously documented features of the 1977 regime shift, and identifies a further shift in 1989 in some components of the North Pacific ecosystem. The 1989 changes were neither as pervasive as the 1977 changes nor did they signal a simple return to pre-1977 conditions.” [http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6V7B-41FTS3S-2…]
Overland et al “North Pacific regime shifts: Definitions, issues and recent transitions” [http://www.pmel.noaa.gov/foci/publications/2008/overN667.pdf]: “climate variables for the North Pacific display shifts near 1977, 1989 and 1998.”
The following figure from the above paper show analysis of PDO and Victoria Index using the Rodionov regime detection algorithm. A regime shift is also detected around 1947-48.
The following figure shows regime shift detection for the summer PDO, showing shifts at 1948, 1976 and 1998. [http://www.beringclimate.noaa.gov/data/Images/PDOs_FigRegime.html]
(For detailed information on the 1976/77 climate shift, see: http://www.appinsys.com/GlobalWarming/The1976-78ClimateShift.htm)
|
|
Regime Shift Detection in Annual Temperature Anomaly Data
The NOAA Bering Climate web site provides the algorithm for regime shift detection developed by Sergei Rodionov [http://www.beringclimate.noaa.gov/regimes/index.html]. The following analyses use the Excel VBA regime change algorithm version 3.2 from this web site.
The following figure shows the regime analysis of the HadCRUT3 annual global annual average temperature anomaly data from the Met Office Hadley Centre for 1895 to 2009 [http://hadobs.metoffice.com/hadcrut3/diagnostics/global/nh+sh/annual].
The analysis was run based on the mean using a significance level of 0.1, cut-off length of 10 and Huber weight parameter of 2 using red noise IP4 subsample size 6. Regime changes are identified in 1902, 1914, 1926, 1937, 1946, 1957, 1977, 1987, and 1997. Running the analysis based on the variance rather than the mean results in regime changes in the bold years listed above.
|
|
Regime Shift Relationship to Solar Cycle
The NASA Solar Physics web site provides the following figure showing sunspot area. [http://solarscience.msfc.nasa.gov/SunspotCycle.shtml]
The following figure compares the Hadley (HadCrut3) monthly global average temperature (from [http://hadobs.metoffice.com/hadcrut3/diagnostics/global/nh+sh/]) overlaid with the regime change line (red line) shown previously, along with the sunspot area since 1900. The sunspot cycle is approximately 11 years. The sun’s magnetic field reverses with each sunspot cycle and thus after two sunspot cycles the magnetic field has completed a cycle – a Hale Cycle – and is back to where it started. Thus a complete magnetic sunspot cycle is approximately 22 years. The figure marks the onset of odd-numbered cycles with a vertical red line, even-numbered cycles with a green line.
From the figure above it can be seen that the regime changes correspond to the onset of solar cycles and occur when the “butterfly” is at its widest. The most significant warming regime shifts occur at the start of odd-numbered cycles (1937, 1957, 1977, 1997). Each odd-numbered cycle (red lines above) has resulted in a temperature-increase regime shift. Even-numbered cycles (green lines above) have been inconsistent, with some resulting in temperature-decrease regime shifts (1902, 1946) or minor temperature-increase shifts (1926, 1987).
An unusual one is the 1957 – 1966 cycle, which in the monthly data shown above visually looks like a temperature-increase shift in 1957 followed by a temperature-decrease shift in 1964 but the regime detection algorithm did not identify it. This is likely due to the use of annually averaged data in the regime detection algorithm.
The following figure shows the relative polarity of the Sun’s magnetic poles for recent sunspot cycles along with the solar magnetic flux [www.bu.edu/csp/nas/IHY_MagField.ppt]. The regime change periods are highlighted by the red and green boxes. Each one occurs on as the solar cycle is accelerating. The onset of an odd-numbered sunspot cycle (1977-78, 1997-98) results in the relative alignment of the Earth’s and the Sun’s magnetic fields (positive North pole on the Sun) allowing greater penetration of the geomagnetic storms into the Earth’s atmosphere. “Twenty times more solar particles cross the Earth’s leaky magnetic shield when the sun’s magnetic field is aligned with that of the Earth compared to when the two magnetic fields are oppositely directed” [http://www.nasa.gov/mission_pages/themis/news/themis_leaky_shield.html]
The following figure shows the longitudinally averaged solar magnetic field. This “magnetic butterfly diagram” shows that the sunspots are involved with transporting the field in its reversal. The Earth’s temperature regime shifts are indicated with the superimposed boxes – red on odd numbered solar cycles, green on even. [http://solarphysics.livingreviews.org/open?pubNo=lrsp-2010-1&page=articlesu8.html]
The Earth’s temperature regime shift occurs as the solar magnetic field begins its reversal.
|
|
Solar Cycle 24
Solar cycle 24 is in its initial stage after getting off to a late start. An El Nino occurred in the first part of 2010. This may be the start of the next regime shift.
|
Stephen Wilde says:
July 6, 2010 at 11:05 am
Leif, You shifted the ground.
Such mutterings are just sophisticated debating technique [you a lawyer, by chance?] and do not bring anything to the table.
“no numbers, no equations, no quantifications” is what you referred to.
“A logical coherence describing and linking varied observations, according with ongoing events and hopefully providing some predictive skill and all without any obvious abuse of the basic laws of physics.” is what I contend is far better.
If your prediction cannot be [or is not] quantified in some way, it cannot be compared with what happens and a skill score cannot be evaluated.
Though both would be nice in an ideal world but in the absence of data beggars cannot be choosers as Basil points out.
In the absence of data you cannot assert anything at all.
It’s just that the world is refusing to cooperate in the way you would like (so far at least).
The ‘world’ will only pay attention to numbers, not to hand waving, no matter how coherent the waver thinks he is.
Introducing the issue of CFCs and anthropogenic CO2 to explain away that observation
Explain ‘away’ is inappropriate as those issues do explain the observations [per my daughter-in-law Signe who knows more about this than you and I combined, http://www.leif.org/EOS/nature04746.pdf or would want to know]. The ‘away’ bit is only if you don’t like the result.
“The fact that the energy of a photon is most often transferred by collisions to other molecules does not really matter, since it means that collisions can also excite molecules that will sometime emit a photon ”
In a discussion of CO2 I find this a bit of a gloss without mention of the relative emissivities-the warm surface is 1000 times that of CO2, and H2O vapor, twice that of CO2 and, moreover, having a broad spectrum of possible emission frequencies rather than a discrete few.
In sum, CO2 heats the surrounding gas in the presence of IR flux(incoming or outgoing) but at it’s low partial pressure may thereafter be ignored.
http://www.leif.org/EOS/Nature/nature04746.pdf works better
Stephen Wilde says:
This is for you:
http://www.rexresearch.com/piccardi/piccardi2.pdf
Reinforcing Steven’s earlier comment:
The problem with this analysis is that a statistical model that assumes constant regimes is fitted to the HadCRUT3 data which has a long term linear trend (real or “adjusted”, it doesn’t matter). The residual sum-of-squares will always be improved by splitting a region into two smaller regions that are each flatter than the whole. This process will go on until the cutoff is reached. That cutoff is set at 10 years, which is roughly the length of the solar cycles that Alan is trying to align the temperature data to.
A more careful analysis would allow a linear trend within each region (the cause could be changes in TSI, random walk, feedbacks, etc.). I did that analysis. To avoid overfitting, I picked the set of breakpoint locations that minimized BIC. Ignoring the AR1 component that is undeniably there in the climate data, four breakpoints are identified: 1864, 1907, 1945, 1963. Adding in 1 year lagged data to correct for autocorrelation, a single breakpoint is identified: 1963.
Both of these fits, but especially the simpler one, look more faithful to the data than the one that Alan proposes. ANOVA proves that they are both better fits than Alan’s model (p < 0.001). This shouldn't be much of a surprise, since it's easier to fit a trend with one sloped line than with even a bunch of horizontal lines.
As always, the code for my analysis is available. Feel free to criticize, correct, or extend it.
Enneagram :
I don’t think that is for me. Sounds too much like homeopathy.
Leif :
So you aver that at all times the upward energy flux from Earth to space is wholly unaffected by solar variability ?
and that the only source of variations in upward flux is internal Earth system variability ?
and that therefore any temperature changes in the stratosphere are wholly due to changes occurring below with no room at all for such changes being effected or affected by changes occurring above ?
I find that very hard to believe in a universe where everything is in constant motion. Such assertions are extraordinary and require extraordinary evidence. Please supply it.
@Enneagram,
From the chapter you posted:
Why did you post this? It looks like homeopathy quackery.
Leif,
I like Signe’s paper. It appropriately expresses the levels of uncertainty and recognises the natural components in ozone variability. It avoids the normal alarmist contentions that the observed changes were solely or even primarily anthropogenic so she is excluded from my critical comments.
What her paper is notable for is in acknowledging a failure to clearly demonstrate any measurable distinction between ozone depletion or ozone recovery from anthropogenic as compared to natural causes.
Thus the CFC/ozone depletion issue suffers from the same defects that you find so damning against me namely “no numbers, no equations, no quantifications”.
My observation that the observed changes in ozone quantities could have been overwhelmingly from natural causes remains standing as a potential explanation.
Whoops, I should have said no ‘meaningful’ numbers equations or quantifications. There is a lot of numbers but nothing that actually resolves what we need to know namely the relative contributions of all the influences on ozone quantities.
Stephen Wilde: I have come to believe that you intentionally use terms as you see fit, and not as is generally accepted, to confuse those who read your comments and to make it difficult (impossible) for those who wish to debate a topic with you since your understandings, representations, and uses of terms are constantly shifting.
Example: You replied, “I use the term PDO in the general sense that has entered common currency but I am aware of the more restricted definition that you use.”
What does “PDO in the general sense” mean, Stephen? A basin wide phenomenon? The low frequency, multidecadal component of ENSO? The “restricted definition” I use is the one that is accepted, Stephen. Here’s a link to the JISAO definition:
http://jisao.washington.edu/pdo/
Here’s a link to the Wikipedia definition:
http://en.wikipedia.org/wiki/Pacific_decadal_oscillation
Here’s one NOAA webpage that defines it that same way:
http://www.nwfsc.noaa.gov/research/divisions/fed/oeip/ca-pdo.cfm
Please define the way in which you use PDO, and then google scholar “Pacific Decadal Oscillation” and provide links to studies that use it in the same way that you use “PDO in the general sense that has entered common currency”. You’ve got 67oo plus papers to sort through so you should be able to come up with a few.
You wrote, “I propose that the main underlying cause is changes in the winds above the equatorial oceans as they respond to latitudinal shifts in the global air circulation systems as per my hypotheses. Those air circulation shifts being the result of an interplay between oceanic and solar cycles.”
And without data to support your proposal, you’re speculating, which is a nice way of saying you’re guessing.
R. Craigen says:
July 6, 2010 at 10:53 am
Steven’s point in his first post is that the choice of algorithm and minimum segment length determine the outcome. A model that thinks segments have constant temperature is going to keep chopping up a temperature record that shows a linear trend, until it is stopped by the cutoff. Alan’s analysis actually stop short of this, but this only happens because 1957–1967 and 1967–1977 happen to have the same means and any other split of 1957–1977 violates the minimum segment length. I don’t (and can’t) deny a solar effect, but I think it mostly fine-tunes where the segment boundaries lie on the scale of one or two years.
The above is just my interpretation, but in an earlier comment I describe the more robust analysis. In summary, the solar correlation disappears. There’s still a lot of room for improvement, using a method that explicitly handles some measure of what the sun is doing.
tallbloke: You replied, “Whatever it is that caused increasing cloudiness from 1998 would have to be a contender for an important role I would have thought.”
Does this mean we’re done arguing? I’m not sure I got my money’s worth.
The ISCCP data is hotly contested, as you’re aware, due to missing data over the Southern Indian Ocean prior to 1998, and due to the influence of volcanic aerosols.
Regards
R. Craigen says:
July 6, 2010 at 10:53 am
Knowledge of mechanisms commonly comes long after knowledge of laws concerning the behavior of related variables.
Ding Ding.
Still true for Newtonian gravity after 400 hundred years. And some of the ‘laws’ around it are pretty much approximate ad hoc mathematical happenstance too.
We have much to learn.
Bob Tisdale says:
July 6, 2010 at 3:35 pm (Edit)
tallbloke: You replied, “Whatever it is that caused increasing cloudiness from 1998 would have to be a contender for an important role I would have thought.”
Does this mean we’re done arguing? I’m not sure I got my money’s worth.
The ISCCP data is hotly contested, as you’re aware, due to missing data over the Southern Indian Ocean prior to 1998, and due to the influence of volcanic aerosols.
I would argue longer, but I took milady out for her birthday dinner and I’m in no fit state.
Palle et al provide an independent take on cloudiness from the end of ’98, but unfortunately, not before. Clever idea, measuring earthshine onto the moon. Not such a good idea partaking of too much of the moonshine on the Earth though… hic.
Doug S
July 6, 2010 at 9:17 am
Reading through the material I found the General expectation value formula particularly interesting:
[..]
If I understand the basics of this correctly we would expect the energy transfer from CO2 other systems to decrease with altitude in the atmosphere due to decreasing densities. The more important piece of the puzzle is to define the operator function that couples CO2 to other neighboring systems
Doug, it’s not a classical mechanical equation – the operator function is a quantum operator, and the density p refers to density of quantum states.
(This is actually the topic of a historic paper “The Quantum Theory of the Emission and Absorption of Radiation” P. A. M. Dirac 1927)
With respect to CO2, in my understand, as the atmospheric density decreases with altitude, one would expect the number of collisions to decrease, increasing the potential emitting time of excited CO2 molecules before a collision occurs.. Therefore, per photon absorption, the probability of a kinetic transfer, rather than an emissive one, would be expected to decrease.
Stephen Wilde: You wrote, “Nonetheless I have said several times that there are lots of ways to falsify my propositions.”
Please detail the ways in which you believe your conjectures can be falsified.
Heuristic “regime shift” algorithms are a simplistic crutch for those unprepared to do serious physical or phenomenological signal analysis.
Bob Tisdale says (July 6, 3:23 am):
“The assumption you’re making is that the (approx.) 60-year cycle persists back in time beyond the instrument temperature record. Paleoclimatological reconstructions of the AMO suggest it does not”
Thanks for the links Bob. I read the Gray et al 2004 paper you linked to. It states: “We present a tree-ring based reconstruction of the
Atlantic Multidecadal Oscillation (AMO) which demonstrates that strong, low-frequency (60–100 yr) variability in basin-wide (0–70N) sea surface temperatures (SSTs) has been a consistent feature of North Atlantic climate for the past five centuries.”
The paleo reconstruction conducted in that paper uses tree ring proxies from various locations in southeast US, northern and southern Europe and the Middle East, where the temperature data do not match the AMO. Combining these different cycles results in the masking of individual cycles. I don’t think this is a very robust reconstruction of the AMO.
Take a look at pages 52 and 53 here, referencing work by the same authors, where they show the tree ring proxies at several locations in the US (and most of these areas have a strong correlation to the AMO). What you will see is that the 60 year cycle is persistent back to 1600, although it occasionally shifts by 30 years. (And they state: “Strong evidence for multidecadal (30-70 yr) persistence and cross-regional synchrony”)
P.S. Don’t anybody tell anybody that I visited Realclimate today. Cos I might start losing my Scep Cred.
Ta.
And, after all the debate, both here and and on Realclimate, I’ve come to the conclusion that my original Photon Football idea is actually a pretty good model. Because the number of photons re-emitted is equal to the number of photons absorbed, even if there’s a bit of bouncy-bouncy in between photon absorptions and emissions.
I’ll start work on it in the morning, if there isn’t any World Cup football to watch.
Unless Leif tells me that Germany aren’t going to win the World Cup…
gary gulrud
July 6, 2010 at 7:42 am
Gary – some kind of reference or link would be welcome.
There seems to be a misunderstanding of what a regime shift algorithm does.
Maud Kipz (July 6, 1:49pm) says: “The residual sum-of-squares will always be improved by splitting a region into two smaller regions that are each flatter than the whole. This process will go on until the cutoff is reached.”
And (July 6, 3:33pm) ” A model that thinks segments have constant temperature is going to keep chopping up a temperature record that shows a linear trend, until it is stopped by the cutoff.”
A regime shift algorithm is not a linear trend algorithm. See http://www.appinsys.com/GlobalWarming/RegimeParams.htm – the last 2 examples show analyses using a cutoff length of 4. The algorithm does not chop it up smaller and smaller as Maud says.
Your (Maud) “more robust” algorithm is a slope-based algorithm (not a regime-based algorithm) – these are two different things. And apparently it missed the slope change at 1878. This figure plots your identified slope change points as well as the 1978 slope change: http://www.appinsys.com/GlobalWarming/SlopeChange.jpg on the HadCrut3 anomaly data.
idlex says: (July 6, 2010 at 3:37 am) And since we seem to have more and more of this sort of incomplete “science”, we have more and more true believers in all sorts of things.
A profound observation, idlex; rather stunning in the simplicity of expression used to highlight a great truth which the world would do well to heed and think on.
Thanks for the guidance on the quantum operator oneuniverse, I guess it is abundantly clear that I am at best a C student of physics. Still, I’m left with one nagging thought after reading through all of the very good posts here. The science of CAGW seems to be far from settled.
Stephen Wilde says:
July 6, 2010 at 2:05 pm
Leif : So you aver that at all times the upward energy flux from Earth to space is wholly unaffected by solar variability ?
This is typical of your style. You might as well have said: “turn of the Sun and see what you get”. The flux from the Earth to Space is equal to the flux from space [including the Sun]. If the input varies so does the output.
temperature changes in the stratosphere are wholly due to changes occurring below with no room at all for such changes being effected or affected by changes occurring above ?
Changes in the stratosphere is very much influenced by incoming solar radiation, but not by temperatures above, simply because the amount heat is too low up there as the density falls by a factor of a thousand for each 50 km gain in altitude.
I find that very hard to believe in a universe where everything is in constant motion.
Galileo knew Newton’s first law: “an object in motion tends to stay in motion”
Maud Kipz says:
July 6, 2010 at 2:09 pm
Stephen Wilde says:
July 6, 2010 at 2:21 pm
I like Signe’s paper. It appropriately expresses the levels of uncertainty and recognises the natural components in ozone variability.
You take that to far. She does not think that things are so uncertain that one cannot make any conclusions.
Thus the CFC/ozone depletion issue suffers from the same defects
As she explains it to me, the working hypothesis that explains the data is that the CFC have effect and that there is now recovery. How strong that is is muddled a bit by solar cycle effects and we need another cycle to be able to decide for sure.
Bob Tisdale says:
July 6, 2010 at 3:27 pm
Stephen Wilde: I have come to believe that you intentionally use terms as you see fit, and not as is generally accepted, to confuse those who read your comments and to make it difficult (impossible) for those who wish to debate a topic with you since your understandings, representations, and uses of terms are constantly shifting.
I’ll second that. My example is ‘turbulence in the sun”
tallbloke says:
July 6, 2010 at 4:41 pm
We have much to learn.
You can start right now.