It looks like you're new here. If you want to get involved, click one of these buttons!

- All Categories 2.4K
- Chat 505
- Study Groups 21
- Petri Nets 9
- Epidemiology 4
- Leaf Modeling 2
- Review Sections 9
- MIT 2020: Programming with Categories 51
- MIT 2020: Lectures 20
- MIT 2020: Exercises 25
- Baez ACT 2019: Online Course 339
- Baez ACT 2019: Lectures 79
- Baez ACT 2019: Exercises 149
- Baez ACT 2019: Chat 50
- UCR ACT Seminar 4
- General 75
- Azimuth Code Project 111
- Statistical methods 4
- Drafts 10
- Math Syntax Demos 15
- Wiki - Latest Changes 3
- Strategy 113
- Azimuth Project 1.1K
- - Spam 1
- News and Information 148
- Azimuth Blog 149
- - Conventions and Policies 21
- - Questions 43
- Azimuth Wiki 719

## Comments

About that caveat I mentioned above:

Following up on this gap, I extended the derivation to formally evaluate the velocity.

So from:

$ \zeta(t) = \sin( \sqrt{A} \sum_{i=1}^{i=N} k_i \sin(\omega_i t) + \theta_0 ) $

and the third simplified Laplace equation

$ \frac {\partial v}{\partial t} =-\frac {1}{a} \frac {\partial }{\partial \varphi } (g\zeta +U)$

we can derive

$ \frac {\partial v}{\partial t} = \cos( \sqrt{A} \sum_{i=1}^{i=N} k_i \sin(\omega_i t) + \theta_0 ) $

So the

accelerationof wind, not the velocity, is what obeys the Sturm-Liouville equation. A derivative preserves the periods of the Fourier components, but not the amplitudes, so what we see is a differently shaped envelope for QBO -- i.e. one that is more spiky due to the time derivative applied.A single lunar Draconic tidal term of 27.212 days multiplied by a yearly modulation peaked at a specific season is enough to capture the QBO acceleration time-series (with a correlation coefficient of 0.35 considering the data waveform has

no filtering appliedand very little optimization went into the fit):This makes sense as the acceleration of wind is simply a $ F = ma $ response to a gravitational forcing.

Incidentally, I submitted this research to the AGU meeting for this fall and will find out if the abstract gets accepted in October.

`About that [caveat I mentioned above](discussion/comment/15472/#Comment_15472): > "The caveat that I expressed is that my derivation was for delta variations in atmospheric pressure, whereas QBO is a measure of velocity. I brushed this aside for the moment because the two - velocity and pressure - are closely related at the differential level (for example see Bernoulli's equation)." Following up on this gap, I extended the derivation to formally evaluate the velocity. So from: $ \zeta(t) = \sin( \sqrt{A} \sum_{i=1}^{i=N} k_i \sin(\omega_i t) + \theta_0 ) $ and [the third simplified Laplace equation](http://contextearth.com/2016/07/04/alternate-simplification-of-qbo-from-laplaces-tidal-equations/) $ \frac {\partial v}{\partial t} =-\frac {1}{a} \frac {\partial }{\partial \varphi } (g\zeta +U)$ we can derive $ \frac {\partial v}{\partial t} = \cos( \sqrt{A} \sum_{i=1}^{i=N} k_i \sin(\omega_i t) + \theta_0 ) $ So the *acceleration* of wind, not the velocity, is what obeys the Sturm-Liouville equation. A derivative preserves the periods of the Fourier components, but not the amplitudes, so what we see is a differently shaped envelope for QBO -- i.e. one that is more spiky due to the time derivative applied. A single lunar Draconic tidal term of 27.212 days multiplied by a yearly modulation peaked at a specific season is enough to capture the QBO acceleration time-series (with a correlation coefficient of 0.35 considering the data waveform has *no filtering applied* and very little optimization went into the fit): ![qboAccel](http://imageshack.com/a/img923/9315/Zgjszt.png) This makes sense as the acceleration of wind is simply a $ F = ma $ response to a gravitational forcing. Incidentally, I submitted this research to the AGU meeting for this fall and will find out if the abstract gets accepted in October.`

nad said:

That's a property of discrete periodic signals. A model of seasonally modulated tidal periods puts an emphasis at the same time each year. As the tidal signal will either constructively or destructively interfere at a specific time, it will look like the cycles will maintain a set period until the lunar periods make it go out-of-phase and therefore "flip" the sign of the interference. A model of analog cycles that reproduces the discrete cycles will generate an

apparent2.37 year period, along with the associated higher frequency harmonics to create this particular erratic yet periodic discrete signal. That's just how Fourier series work.Here is another fit, where I modeled a very sharp seasonal modulation, and then smoothed it with an damped exponential. Again the

accelerationof QBO defined according to Laplace's tidal equations matches most of the observed delta spikes.That is using only one lunar factor, due to the draconic/nodal tide. Adding the other tidal factors will improve this fit even more.

To get back to the QBO

velocity, all we need to do is integrate this waveform. And this will naturally generate a square-wave like oscillation reminiscent of QBO, because a square-wave will generate +/- delta functions upon differentiation, spaced just as the acceleration waveform shows!Now, the natural question is: Can the seasonal modulation be this sharp? Certainly. If one looks at a histogram of tropical storms, it has a very narrow distribution, sharply centered on September 10. The width of this peak fits within a single 27 day tidal cycle!

I don't know why it's this sharp but that's what the historical data shows.

Another recent paper shows the same sharp anomalies in atmospheric gravity waves over the city of Prague

Putting this in some perspective, here are a couple of take-away bullets:

## QBO

consensusexplanation requires several paragraphs, a scaled lab model, and numerical simulation to demonstrate.## Ocean Tides

So what's wrong with this picture? Why is the consensus theory for tides so intuitively simple, yet the QBO so horribly complex?

I think the answer is that the AGW denier Richard Lindzen fell well short in setting up a foundational theory for QBO. The actually model is likely much simpler to explain and to mathematically express than we have been lead to believe. That shouldn't be surprising if one finds out how many times Lindzen's theories have been debunked over the years. I hate to say this but Lindzen may give astronomer Thomas Gold a run for his money in terms of failed theories.

Updated my blog here as well.

`nad said: > "I wrote Ansatz because the 2.4 Period Appears to me to be a lot for having those rather long strictly 2 year periods before a "flip" of phase." That's a property of discrete periodic signals. A model of seasonally modulated tidal periods puts an emphasis at the same time each year. As the tidal signal will either constructively or destructively interfere at a specific time, it will look like the cycles will maintain a set period until the lunar periods make it go out-of-phase and therefore "flip" the sign of the interference. A model of analog cycles that reproduces the discrete cycles will generate an *apparent* 2.37 year period, along with the associated higher frequency harmonics to create this particular erratic yet periodic discrete signal. That's just how Fourier series work. Here is another fit, where I modeled a very sharp seasonal modulation, and then smoothed it with an damped exponential. Again the *acceleration* of QBO defined according to Laplace's tidal equations matches most of the observed delta spikes. ![qbop](http://imageshack.com/a/img923/7199/zqEfww.png) That is using only one lunar factor, due to the draconic/nodal tide. Adding the other tidal factors will improve this fit even more. To get back to the QBO *velocity*, all we need to do is integrate this waveform. And this will naturally generate a square-wave like oscillation reminiscent of QBO, because a square-wave will generate +/- delta functions upon differentiation, spaced just as the acceleration waveform shows! Now, the natural question is: Can the seasonal modulation be this sharp? Certainly. If one looks at a histogram of tropical storms, it has a very narrow distribution, sharply centered on September 10. The width of this peak fits within a single 27 day tidal cycle! ![storm](http://imageshack.com/a/img921/3905/QnqHEU.gif) I don't know why it's this sharp but that's what the historical data shows. Another [recent paper](http://contextearth.com/kramer2016/) shows the same sharp anomalies in atmospheric gravity waves over the city of Prague ![gw](http://imageshack.com/a/img922/1163/QwNzzC.png) > Kramer, Ricarda, Sabine Wüst, and Michael Bittner. "Investigation of gravity wave activity based on operational radiosonde data from 13 years (1997-2009): Climatology and possible induced variability." Journal of Atmospheric and Solar-Terrestrial Physics 140 (2016): 23-33. --- --- Putting this in some perspective, here are a couple of take-away bullets: # QBO - The QBO is arguably the most fundamental global behavior in atmospheric sciences. - Yet, the *consensus* explanation requires [several paragraphs](http://www.goes-r.gov/users/comet/tropical/textbook_2nd_edition/print_4.htm#page_2.2.0), a [scaled lab model](http://www.goes-r.gov/users/comet/tropical/textbook_2nd_edition/print_4.htm#ch4r159), and numerical simulation to demonstrate. --- # Ocean Tides - The ocean tide is a fundamental behavior in geophysics. - An explanation takes a sentence and knowledge of the lunar and solar orbit. --- So what's wrong with this picture? Why is the consensus theory for tides so intuitively simple, yet the QBO so horribly complex? I think the answer is that the AGW denier Richard Lindzen fell well short in setting up a foundational theory for QBO. The actually model is likely much simpler to explain and to mathematically express than we have been lead to believe. That shouldn't be surprising if one finds out how many times Lindzen's theories have been debunked over the years. I hate to say this but Lindzen may give astronomer Thomas Gold a run for his money in terms of failed theories. Updated my blog [here](http://contextearth.com/2016/08/23/qbo-model-final-stretch/) as well.`

Interesting that the storm histogram I posted was just tweeted

The sharp peak is explained here

So from what I can understand, the circulation along the equator diminishes in the late summer, allowing a strong buildup toward favorable thermodynamics, eventually reaching a tipping point?

`Interesting that the storm histogram I posted was just tweeted ![tweet](http://imageshack.com/a/img923/563/EN31bf.png) The sharp peak is explained [here](http://www.noaa.gov/stories/peak-of-hurricane-season-why-now) > "This minimum in the shear combines with favorable thermodynamics – ocean temperatures in the deep tropics that increase with each day of summer sun, warmer air temperatures, and increasing atmospheric moisture. When the dynamics and thermodynamics are in sync, as they often are from mid-August through early October, disturbances like African tropical waves can easily strengthen. " So from what I can understand, the circulation along the equator diminishes in the late summer, allowing a strong buildup toward favorable thermodynamics, eventually reaching a tipping point?`

I read this. It made me ask how it might relate to an explanation I just read (no idea where) that the current deluges in the US come from the El Nino build-up of hot water in the Pacific migrating over the North Pole?

`I read this. It made me ask how it might relate to an explanation I just read (no idea where) that the current deluges in the US come from the El Nino build-up of hot water in the Pacific migrating over the North Pole?`

Jim, you may be right on that interpretation. Any kind of tipping point would likely show a sharp peak.

`Jim, you may be right on that interpretation. Any kind of tipping point would likely show a sharp peak.`

In the link you provided in 202 it is written:

This is my biggest problem with the theories about QBO. If I look at the images than the QBO seems to be a result of a downward movement not of an upward movement.

`In the <a href="http://www.goes-r.gov/users/comet/tropical/textbook_2nd_edition/print_4.htm#page_2.2.0">link</a> you provided in 202 it is written: >Fig. 4.56. Schematic of the relationship of Kelvin waves and Rossby-gravity waves in the equatorial zone. In all of the diagrams in this figure, the black line shows the vertical profile of the near equatorial zonal wind (westerlies are positive) in the upper troposphere and lower stratosphere. The Kelvin waves (red) are propagating upwards through the atmosphere transporting westerly momentum. The Rossby-gravity waves (blue) are also propagating upwards, taking easterly momentum with them. This is my biggest problem with the theories about QBO. If I look at the images than the QBO seems to be a result of a downward movement not of an upward movement.`

This upward and downward movement may be simply a phase relationship between the stratified layers. Almost all of the momentum is tied up in the horizontal wind, and so the vertical momentum transfer is minimal in the stratosphere. Dropping in to the troposphere, there is more drag so the wind speed will always lag there. I should add a drag term to the equations to see how it changes. It should get weaker with an accompanying phase lag.

Hard to understand these vertical movements when the horizontal movements are that strong.

`This upward and downward movement may be simply a phase relationship between the stratified layers. Almost all of the momentum is tied up in the horizontal wind, and so the vertical momentum transfer is minimal in the stratosphere. Dropping in to the troposphere, there is more drag so the wind speed will always lag there. I should add a drag term to the equations to see how it changes. It should get weaker with an accompanying phase lag. Hard to understand these vertical movements when the horizontal movements are that strong.`

I don't understand what you mean by that. It is the wind direction which is moving downward. That is an eastward SOI will eventually morph downward into an eastward QBO wind a.s.o.

`>This upward and downward movement may be simply a phase relationship between the stratified layers. I don't understand what you mean by that. It is the wind direction which is moving downward. That is an eastward SOI will eventually morph downward into an eastward QBO wind a.s.o.`

The stratosphere is called the stratosphere because the layers are stratified and the 30 hPa layer is considered the strongest QBO effect. The other layers are solved from the spatial standing wave which forms according to Laplaces equation and the lower and upper boundary conditions. I don't consider that too interesting because the first-order effect is still the strong oscillation at 30 hPa.

E-M waveguides work on this same principle. The metallic surfaces show a skin effect and set the boundary conditions. Yet the interesting behavior is really in the inner region of the waveguide.

If someone else wants to provide a different physical meaning to this, they can. I prefer to work out the governing wave equations and let that do the explaining.

Btw, I wrote this new post Saturday morning to help understand what I am trying to do

http://contextearth.com/2016/09/03/geophysical-fluid-dynamics-first-and-then-cfd/

`The stratosphere is called the stratosphere because the layers are stratified and the 30 hPa layer is considered the strongest QBO effect. The other layers are solved from the spatial standing wave which forms according to Laplaces equation and the lower and upper boundary conditions. I don't consider that too interesting because the first-order effect is still the strong oscillation at 30 hPa. E-M waveguides work on this same principle. The metallic surfaces show a skin effect and set the boundary conditions. Yet the interesting behavior is really in the inner region of the waveguide. If someone else wants to provide a different physical meaning to this, they can. I prefer to work out the governing wave equations and let that do the explaining. Btw, I wrote this new post Saturday morning to help understand what I am trying to do http://contextearth.com/2016/09/03/geophysical-fluid-dynamics-first-and-then-cfd/`

I have no idea how much stratification is going on in the atmossphere. I dont know but I could imagine that this Laplace equations have some connections with a set of equations which describe thicker layers (which may eventually be stratified) similar to how Eulerian fluids appear in Navier-Stokes (a very good overview is given by Tim van Beeks posts). I didnt fully read what you wrote in your post but it seems to be related to this fluid dynamics mechanism of simplification.

It may of course be the case that the images I saw or my perception of them were somehow erranous, but what I saw is a clear downward movement. So I currently imagine the QBO as being somekind of vortex line whose timeevolution is going down just as in the video you showed in 190. Hence in order to describe that the Laplace equations seem not sufficient.

`>The stratosphere is called the stratosphere because the layers are stratified and the 30 hPa layer is considered the strongest QBO effect. I have no idea how much stratification is going on in the atmossphere. I dont know but I could imagine that this Laplace equations have some connections with a set of equations which describe thicker layers (which may eventually be stratified) similar to how Eulerian fluids appear in Navier-Stokes (a very good overview is given by <a href="https://johncarlosbaez.wordpress.com/2012/05/30/fluid-flows-and-infinite-dimensional-manifolds-part-3/">Tim van Beeks posts</a>). I didnt fully read what you wrote in your post but it seems to be related to this fluid dynamics mechanism of simplification. It may of course be the case that the images I saw or my perception of them were somehow erranous, but what I saw is a clear downward movement. So I currently imagine the QBO as being somekind of vortex line whose timeevolution is going down just as in the video you showed in <a href="https://forum.azimuthproject.org/discussion/comment/15466/#Comment_15466">190.</a> Hence in order to describe that the Laplace equations seem not sufficient.`

The Laplace tidal equations (derived from a linearization of the primitive equations describing fluid flow) provided the breakthrough which allowed ocean tides to be mathematically modeled. Not widely appreciated, but it's the horizontal gravitational force that causes the vertical displacement of these tides.

from http://faculty.washington.edu/luanne/pages/ocean420/notes/TidesIntro.pdf

Yet practically speaking, the equations need to be tweaked quite a bit because ocean tides show significant spatial variability. The position of the moon in terms of the tropical or synodic month (27.322 days) plays a significant role as this establishes the longitudinal placement (Pacific, Atlantic, etc) of the strongest instantaneous tidal forces.

On the other hand, the QBO is almost completely longititudinally uniform. In this case, the draconic or nodal month (27.212 days) is the significant driver. This generates a horizontal or latitudinal forcing across the equator. So if the moon and sun are in max nodal excursions with respect to the equator, this will affect a change in the QBO direction, due to the cross-terms in the Laplace tidal equations:

I am not doing anything out of the ordinary, but simply placing a latitudinal-directed cyclic forcing in Laplace's equations and then reducing the set for a small angle approximation -- which I cover here:

http://contextearth.com/2016/07/04/alternate-simplification-of-qbo-from-laplaces-tidal-equations/

This is as solid a theory as I can devise from scratch and it matches the empirical observations of the QBO period precisely. In other words, if the nodal lunar month differed from 27.212 days by much more than a fraction of a percent, the entire theory would be invalidated (e.g. apply instead the tropical lunar month of 27.322 days and the empirical agreement is much worse).

Same goes for oceanic tides; if the period of the tides differed much from what is theoretically predicted by the lunisolar periods, the entire theory of a gravitational forcing leading to bulging tides would be invalidated.

So its essentially the same mechanism for oceanic tides applied to another fluid, leading to stratospheric tides.

The only problem is that you won't find this tidal derivation of QBO anywhere in the research literature. Do a Google scholar search on QBO with the additional terms Draconic and lunar and you won't find much of anything. The scientific consensus is that the period of QBO arises as some sort of emergent phenomenon best described as an internal system resonance. Lindzen long ago considered lunar forcing and rejected the possibility, and so this is how the standard QBO model evolved.

In retrospect, where I think scientists such as Lindzen went wrong wrt QBO is in focusing on the QBO

velocityand not the QBOacceleration. Acceleration is the operable characteristic, as that is what is observed in response to a Newtonian forcing in a non-viscous setting. And acceleration is what is formulated in Laplace's equations.`The Laplace tidal equations (derived from a linearization of the [primitive equations](https://en.wikipedia.org/wiki/Primitive_equations) describing fluid flow) provided the breakthrough which allowed ocean tides to be mathematically modeled. Not widely appreciated, but it's the horizontal gravitational force that causes the vertical displacement of these tides. > "The actual imbalance force, called the tide-generating force (TGF) has horizontal components (called tractive force) as well as the vertical component, and is distributed as shown below (this is a hypothetical case exaggerated to show the direction of TGF). Because it is difficult to generate vertical movement of water in the ocean, tractive forces are much more important for the generation of tides than the vertical component of TGF. That the moon’s gravity is stronger on one side of the earth than the other is crucial to the argument we just made. In fact, physicists generalize from this and call any deformational force that results from gravity varying from one point of an object to another “tidal”. " ![luanne](http://imageshack.com/a/img921/9549/jEqXWx.png) from http://faculty.washington.edu/luanne/pages/ocean420/notes/TidesIntro.pdf Yet practically speaking, the equations need to be tweaked quite a bit because ocean tides show significant spatial variability. The position of the moon in terms of the tropical or synodic month (27.322 days) plays a significant role as this establishes the longitudinal placement (Pacific, Atlantic, etc) of the strongest instantaneous tidal forces. On the other hand, the QBO is almost completely longititudinally uniform. In this case, the draconic or nodal month (27.212 days) is the significant driver. This generates a horizontal or latitudinal forcing across the equator. So if the moon and sun are in max nodal excursions with respect to the equator, this will affect a change in the QBO direction, due to the cross-terms in the Laplace tidal equations: ![config](http://imageshack.com/a/img922/7527/zm7pUf.png) I am not doing anything out of the ordinary, but simply placing a latitudinal-directed cyclic forcing in Laplace's equations and then reducing the set for a small angle approximation -- which I cover here: http://contextearth.com/2016/07/04/alternate-simplification-of-qbo-from-laplaces-tidal-equations/ This is as solid a theory as I can devise from scratch and it matches the empirical observations of the QBO period precisely. In other words, if the nodal lunar month differed from 27.212 days by much more than a fraction of a percent, the entire theory would be invalidated (e.g. apply instead the tropical lunar month of 27.322 days and the empirical agreement is much worse). Same goes for oceanic tides; if the period of the tides differed much from what is theoretically predicted by the lunisolar periods, the entire theory of a gravitational forcing leading to bulging tides would be invalidated. So its essentially the same mechanism for oceanic tides applied to another fluid, leading to stratospheric tides. The only problem is that you won't find this tidal derivation of QBO anywhere in the research literature. Do a [Google scholar search](https://scholar.google.com/scholar?q=%22quasi-biennial+oscillation%22+lunar+draconic&btnG=&hl=en&as_sdt=0%2C24) on QBO with the additional terms Draconic and lunar and you won't find much of anything. The scientific consensus is that the period of QBO arises as some sort of emergent phenomenon best described as an internal system resonance. Lindzen long ago considered lunar forcing and rejected the possibility, and so this is how the standard QBO model evolved. In retrospect, where I think scientists such as Lindzen went wrong wrt QBO is in focusing on the QBO *velocity* and not the QBO *acceleration*. Acceleration is the operable characteristic, as that is what is observed in response to a Newtonian forcing in a non-viscous setting. And acceleration is what is formulated in Laplace's equations.`

A somewhat forgotten paper by Glenn Brier from 1989 demonstrates that ENSO is not as random as some think

http://contextEarth.com/2016/09/09/obscure-paper-on-enso-determinism/

Brier finds 6.75 and 14 year periods in the historical data going back to the 1525. My model finds the strongest periods at 6.5 and 14 years from 1880 until current.

Brier's paper only has 10 citations. That's probably why I didn't uncover it until now, and only thought to look more deeply because Brier always has interesting ideas.

`A somewhat forgotten paper by Glenn Brier from 1989 demonstrates that ENSO is not as random as some think http://contextEarth.com/2016/09/09/obscure-paper-on-enso-determinism/ Brier finds 6.75 and 14 year periods in the historical data going back to the 1525. My model finds the strongest periods at 6.5 and 14 years from 1880 until current. Brier's paper only has 10 citations. That's probably why I didn't uncover it until now, and only thought to look more deeply because Brier always has interesting ideas.`

I wrote over a year ago that I have cracked the ENSO code, which I have. It has now gone over a year and I have improved my data and method, but I still have more ways to improve my results. Here is my new updated result, where the training period is from 1979 up to the end of 2004, the testing period is from 2005 up to the end of 2014 and the forecast from 2015 up to 2022.

This is a zoomed up version from 2014 to 2022. As you can see the green line is the real ENSO value and the red line is the ENSO value from the Neural network, which is an average from 9 independent ensembles. As you can see the current La Niña which NOAA has written off, is going to deepen with the strongest period around Feb, Mar or Apr, which is then followed by an uptick for the next following 12 months. I missed the size of the resent El Niño. I cannot differentiate between strong and weak El Niños at the moment but I think this is something I can correct in my next update. I also plan to make shorter forecast for say 1 year, 2 years and 4 years where I also plan to include real ENSO data up to the current date and combined this with the result from my NN. As of now, I use no ENSO data as input in my NN, which some may find hard to believe. I use only data from the lunar cycles, variations in Earth’s magnetic field and variations in solar wind as inputs.

`I wrote over a year ago that I have cracked the ENSO code, which I have. It has now gone over a year and I have improved my data and method, but I still have more ways to improve my results. Here is my new updated result, where the training period is from 1979 up to the end of 2004, the testing period is from 2005 up to the end of 2014 and the forecast from 2015 up to 2022. <img class="aligncenter size-full wp-image-508" src="http://www.coolingnews.com/wp-content/uploads/2016/08/ENSO-ANN-big.jpg" alt="ENSO-ANN-big" width="1003" height="596" srcset="http://www.coolingnews.com/wp-content/uploads/2016/08/ENSO-ANN-big.jpg 1003w, http://www.coolingnews.com/wp-content/uploads/2016/08/ENSO-ANN-big-300x178.jpg 300w, http://www.coolingnews.com/wp-content/uploads/2016/08/ENSO-ANN-big-768x456.jpg 768w" sizes="(max-width: 1003px) 100vw, 1003px" /> This is a zoomed up version from 2014 to 2022. As you can see the green line is the real ENSO value and the red line is the ENSO value from the Neural network, which is an average from 9 independent ensembles. As you can see the current La Niña which NOAA has written off, is going to deepen with the strongest period around Feb, Mar or Apr, which is then followed by an uptick for the next following 12 months. I missed the size of the resent El Niño. I cannot differentiate between strong and weak El Niños at the moment but I think this is something I can correct in my next update. I also plan to make shorter forecast for say 1 year, 2 years and 4 years where I also plan to include real ENSO data up to the current date and combined this with the result from my NN. As of now, I use no ENSO data as input in my NN, which some may find hard to believe. I use only data from the lunar cycles, variations in Earth’s magnetic field and variations in solar wind as inputs. <img class="aligncenter size-full wp-image-509" src="http://www.coolingnews.com/wp-content/uploads/2016/08/ENSO-forecast-9-july-16.jpg" alt="ENSO-forecast-9-july-16" width="952" height="548" srcset="http://www.coolingnews.com/wp-content/uploads/2016/08/ENSO-forecast-9-july-16.jpg 952w, http://www.coolingnews.com/wp-content/uploads/2016/08/ENSO-forecast-9-july-16-300x173.jpg 300w, http://www.coolingnews.com/wp-content/uploads/2016/08/ENSO-forecast-9-july-16-768x442.jpg 768w" sizes="(max-width: 952px) 100vw, 952px" />`

Per, Your work doesn't pass the first sanity check. All you have to do is backcast it to years before 1980. There are 100 years of ENSO data prior to 1980 to validate against. If you can do this and your results show a good correlation, I will be pleasantly surprised.

Otherwise, I believe you may be a victim of a massive over-fitting exercise. Beginners always go through the overfitting stage, but usually get straightened out when they extrapolate to the out-of-band intervals.

The other curious statement in your claims is that you say that you can't get the amplitude right on projection tests, yet your amplitude looks very good on your test interval, 2005-2014. That makes no sense unless you were actually fitting against your test interval. That's no good and it makes it look like you are cooking the books.

My strong recommendation is don't wait for the future when you can learn from the past. I am going all the way back to 1880 and unless you do that too, I remain suspicious of your results.

`Per, Your work doesn't pass the first sanity check. All you have to do is backcast it to years before 1980. There are 100 years of ENSO data prior to 1980 to validate against. If you can do this and your results show a good correlation, I will be pleasantly surprised. Otherwise, I believe you may be a victim of a massive over-fitting exercise. Beginners always go through the overfitting stage, but usually get straightened out when they extrapolate to the out-of-band intervals. The other curious statement in your claims is that you say that you can't get the amplitude right on projection tests, yet your amplitude looks very good on your test interval, 2005-2014. That makes no sense unless you were actually fitting against your test interval. That's no good and it makes it look like you are cooking the books. My strong recommendation is don't wait for the future when you can learn from the past. I am going all the way back to 1880 and unless you do that too, I remain suspicious of your results.`

Per said:

How can you apply only this data to project into the future given that "variations in Earth’s magnetic field and variations in solar wind" are impossible to predict?

I can understand projecting lunar cycles, which can only mean that the lunar cycles are the primary factors while the magnetic and solar wind are secondary factors in obtaining a good fit. As you can't possibly know what the solar wind or magnetic field is in 5 years time, your projection can only use lunar cycles.

Now you really have to backproject your model to 1880, as you can't use an excuse of not having adequate historical magnetic field and solar wind data. All the lunar cycle data is deterministically known back to 1880 and well before that.

I know that you presented your results at this questionable climate conference a few weeks ago but we aren't as gullible as the AGW skeptic community, and so will challenge you on your assertions.

`Per said: > " I use only data from the lunar cycles, variations in Earth’s magnetic field and variations in solar wind as inputs." How can you apply only this data to project into the future given that "variations in Earth’s magnetic field and variations in solar wind" are impossible to predict? I can understand projecting lunar cycles, which can only mean that the lunar cycles are the primary factors while the magnetic and solar wind are secondary factors in obtaining a good fit. As you can't possibly know what the solar wind or magnetic field is in 5 years time, your projection can only use lunar cycles. Now you really have to backproject your model to 1880, as you can't use an excuse of not having adequate historical magnetic field and solar wind data. All the lunar cycle data is deterministically known back to 1880 and well before that. I know that you presented your results at this questionable [climate conference](https://geoethic.com/london-conference-2016/) a few weeks ago but we aren't as gullible as the AGW skeptic community, and so will challenge you on your assertions.`

Besides Per there is another character named Plum (I kid you not) who is trying to fit cycles in the temperature and claims to use 89 (!) sinusoidal factors. All these are arbitrary apparently. Problem is that one can fit just about anything with that many factors. That's over at Clive Best's blog if you want to check it out.

Whether Per is something something similar here, I don't know. The problem with Per is that he is intentionally being obtuse on what he is doing. He first posted here probably over a year ago, and he still hasn't given any details. I do marvel at his fit above -- eyeballing it, I wouldn't be surprised if his correlation coefficient is at least 0.995. Rare to find anything that good considering all the uncertainty involved.

All I know is that we won't get anywhere unless we deal in good faith and are able to share AND reproduce scientific results.

`Besides Per there is another character named Plum (I kid you not) who is trying to fit cycles in the temperature and claims to use 89 (!) sinusoidal factors. All these are arbitrary apparently. Problem is that one can fit just about anything with that many factors. That's over at Clive Best's blog if you want to check it out. Whether Per is something something similar here, I don't know. The problem with Per is that he is intentionally being obtuse on what he is doing. He first posted here probably over a year ago, and he still hasn't given any details. I do marvel at his fit above -- eyeballing it, I wouldn't be surprised if his correlation coefficient is at least 0.995. Rare to find anything that good considering all the uncertainty involved. All I know is that we won't get anywhere unless we deal in good faith and are able to share AND reproduce scientific results.`

Hi WebHubTel I try to answer your accusations, despite calling me a liar, which is somewhat amusing as you seem very interested in what I’m doing. But, first I shall like to explain what I’ve done. Over 4 years ago I discovered with the help of my neural network that ENSO variation was influence by strong tidal gravitational forces. Since then I’ve refined and experimented and included solar electromagnetic data which are also correlated to ENSO.

My goal is to make long range ENSO predictions as good as possible. I’ve now created a sharp and effective tool to investigate weak correlations which with normal statistical methods are almost impossible to discover. There are a few steps still I plan to make which would resolve both the strength of El Niños problem and the locations. In other words, whether it is a modoki type El Niño, an eastern placement or somewhere in between. Your claim that we don’t know the future values of the values of Earth’s magnetic values and the solar wind is correct. However, the expected overall trend is known. So what I have done is that I have used different trend lines for these data series and added stochastic noise to these data coherent to each other’s. Each ensemble is independent to each other as they use unique generated solar data and also unique in-data into the neurons which are randomly selected with unique random seeds. Of course in the training part there is a tendency for overfitting, but that is an integrated part of how neural network works. This problem is eliminated for the forecasts as I use unique randomized input data and unique solar forecast data for each ensemble data.

I see that you work with QBO which cyclicity is connected directly to the Moon cycles and much easier to work with than it is with ENSO. You seem to get entangled with Laplace transforms and differential equations related to QBO.

The fact that I have identified not only some obscure triggers for ENSO variations, but the main underlining drivers for ENSO variability and the climate community has not, raises some interesting questions. Why haven’t they done this and what does this tell you about the credibility of the GCM climate models?

`Hi WebHubTel I try to answer your accusations, despite calling me a liar, which is somewhat amusing as you seem very interested in what I’m doing. But, first I shall like to explain what I’ve done. Over 4 years ago I discovered with the help of my neural network that ENSO variation was influence by strong tidal gravitational forces. Since then I’ve refined and experimented and included solar electromagnetic data which are also correlated to ENSO. My goal is to make long range ENSO predictions as good as possible. I’ve now created a sharp and effective tool to investigate weak correlations which with normal statistical methods are almost impossible to discover. There are a few steps still I plan to make which would resolve both the strength of El Niños problem and the locations. In other words, whether it is a modoki type El Niño, an eastern placement or somewhere in between. Your claim that we don’t know the future values of the values of Earth’s magnetic values and the solar wind is correct. However, the expected overall trend is known. So what I have done is that I have used different trend lines for these data series and added stochastic noise to these data coherent to each other’s. Each ensemble is independent to each other as they use unique generated solar data and also unique in-data into the neurons which are randomly selected with unique random seeds. Of course in the training part there is a tendency for overfitting, but that is an integrated part of how neural network works. This problem is eliminated for the forecasts as I use unique randomized input data and unique solar forecast data for each ensemble data. I see that you work with QBO which cyclicity is connected directly to the Moon cycles and much easier to work with than it is with ENSO. You seem to get entangled with Laplace transforms and differential equations related to QBO. The fact that I have identified not only some obscure triggers for ENSO variations, but the main underlining drivers for ENSO variability and the climate community has not, raises some interesting questions. Why haven’t they done this and what does this tell you about the credibility of the GCM climate models?`

per, You still haven't answered why you have not tried to fit your model to ENSO data in the interval 1880 to 1980. That is perfectly good data to test against.

As I said before, I wouldn't be surprised if your correlation coefficient isn't at least 0.995 in the chart you posted over the 30-year fitting interval. You must work with dozens of degrees of freedom in your model to get that kind of agreement. One would think that the noise in the data itself would limit how close a fit one can get. Consider that the SOI dipole pair of Tahiti and Darwin doesn't have a correlation coefficient much better than -0.80 and that should be -1 if it accurately reflects an actual dipole behavior. The reason it doesn't come close is measurement noise, as for example, there may be a hurricane in Tahiti that isn't impacting Darwin and that will throw off the differential readings for at least a month.

If I had a correlation coefficient greater than 0.99, I would

immediatelytry it out on the 1880 to 1980 data set. Even if it wasn't still 0.99, and moved down around 0.7 to 0.8, it would indicate which of the factors were the most important.Yet it really bothers me that you have been working on this for how many months or years, and you still haven't applied the 1880 to 1980 data yet? I did that from day one ! There are really no controlled experiments you can do in climate science, and so all you have to work with is empirical data. For someone not to use the available data indicates to me that you may be naive about the nature of the behavior, or are trying to hide something.

Hints:

I can't think of three more important requirements in research analysis, and the fact that you are doing none of these has me frustrated.

If you don't do these, people will continue to assume that you are:

`per, You still haven't answered why you have not tried to fit your model to ENSO data in the interval 1880 to 1980. That is perfectly good data to test against. As I said before, I wouldn't be surprised if your correlation coefficient isn't at least 0.995 [in the chart you posted](15536/#Comment_15536) over the 30-year fitting interval. You must work with dozens of degrees of freedom in your model to get that kind of agreement. One would think that the noise in the data itself would limit how close a fit one can get. Consider that the SOI dipole pair of Tahiti and Darwin doesn't have a correlation coefficient much better than -0.80 and that should be -1 if it accurately reflects an actual dipole behavior. The reason it doesn't come close is measurement noise, as for example, there may be a hurricane in Tahiti that isn't impacting Darwin and that will throw off the differential readings for at least a month. If I had a correlation coefficient greater than 0.99, I would _immediately_ try it out on the 1880 to 1980 data set. Even if it wasn't still 0.99, and moved down around 0.7 to 0.8, it would indicate which of the factors were the most important. Yet it really bothers me that you have been working on this for how many months or years, and you still haven't applied the 1880 to 1980 data yet? I did that from day one ! There are really no controlled experiments you can do in climate science, and so all you have to work with is empirical data. For someone not to use the available data indicates to me that you may be naive about the nature of the behavior, or are trying to hide something. Hints: * Use all available data * Explain exactly what you are doing so someone else can reproduce it * Apply a standard metric to your result, for example a correlation coefficient and number of degrees of freedom I can't think of three more important requirements in research analysis, and the fact that you are doing none of these has me frustrated. If you don't do these, people will continue to assume that you are: * Cherry picking * Appealing to personal trust * Working with smoke and mirrors`

If Per Strandberg has the solution to ENSO he really ought to try it on the 1880-1980 interval.

I bring that up again, because I decided to test our model against the ENSO proxy record that goes back to the year 1651. The complete write-up is here:

http://contextearth.com/2016/09/27/enso-proxy-revisited/

The significance of the fit over the earlier time spans is not as good as for the recent instrumental record, yet it also appears extremely unlikely to be the result of chance. What I did was to take about 500 sets of synthetic red noise time-series, and then statistically compare the agreement of the model to the real data and to the synthetic data. Since red noise is often thought to characterize ENSO, this is a good way to compensate for over-fitting -- since over-fitting will improve a fit to red noise just as does to the real data, this technique has discriminatory power in separating significance from chance.

Note that the four spans making up the proxy record are all in the high correlation coefficient region. Pulling four straight samples from a set of Monte Carlo red noise runs having each a correlation coefficient above 0.7 is unlikely. This indicates that the model to the current instrumental record likely applies to the past (and seeing one of the fits above 0.8 is a 1 in 500 likelihood just by itself).

This is a confidence booster in substantiating that we are on the right track with the ENSO model.

`If Per Strandberg has the solution to ENSO he really ought to try it on the 1880-1980 interval. I bring that up again, because I decided to test our model against the ENSO proxy record that goes back to the year 1651. The complete write-up is here: http://contextearth.com/2016/09/27/enso-proxy-revisited/ The significance of the fit over the earlier time spans is not as good as for the recent instrumental record, yet it also appears extremely unlikely to be the result of chance. What I did was to take about 500 sets of synthetic red noise time-series, and then statistically compare the agreement of the model to the real data and to the synthetic data. Since red noise is often thought to characterize ENSO, this is a good way to compensate for over-fitting -- since over-fitting will improve a fit to red noise just as does to the real data, this technique has discriminatory power in separating significance from chance. ![red](http://imagizer.imageshack.us/a/img921/9971/orP6oE.png) Note that the four spans making up the proxy record are all in the high correlation coefficient region. Pulling four straight samples from a set of Monte Carlo red noise runs having each a correlation coefficient above 0.7 is unlikely. This indicates that the model to the current instrumental record likely applies to the past (and seeing one of the fits above 0.8 is a 1 in 500 likelihood just by itself). This is a confidence booster in substantiating that we are on the right track with the ENSO model.`

I am not going to make any more comments on this thread and so go to the following for further ENSO and QBO discussion:

https://forum.azimuthproject.org/discussion/1471/qbo-and-enso#latest

As a parting comment, I've been getting trackbacks on my blog from other blogs that Per Strandberg has been contributing to. All I can say is that people such as Per and others that are trying to model ENSO are pushing an agenda. They claim that they are doing science but are clearly using that as a smokescreen to promote their ideas pro-Brexit, anti-Clinton, anti-AGW, and anti-renewables.

EDIT: I keep getting trackbacks with comments on how rude I am. Suffice to say that these people are paranoid in the extreme. The commentary is delusional on a Trump level -- here is an example:

Apparently there are "agents" and "elites" out there that are infiltrating AGW denier blogs such as WUWT and Climate Etc trying to make them less denialist. ... what? I found pages of these comments, warning about "dark agents from California" working to subvert society.

`I am not going to make any more comments on this thread and so go to the following for further ENSO and QBO discussion: https://forum.azimuthproject.org/discussion/1471/qbo-and-enso#latest As a parting comment, I've been getting trackbacks on my blog from other blogs that Per Strandberg has been contributing to. All I can say is that people such as Per and others that are trying to model ENSO are pushing an agenda. They claim that they are doing science but are clearly using that as a smokescreen to promote their ideas pro-Brexit, anti-Clinton, anti-AGW, and anti-renewables. EDIT: I keep getting trackbacks with comments on how rude I am. Suffice to say that these people are paranoid in the extreme. The commentary is delusional on a Trump level -- here is an example: >"So far as I understand wuwt & ce exist to help one or a few individual elites protect their financial investments from political change (by having a small crew of agents apply crude tools like harassment to herd readers onto an oversimplified lukewarm path)." Apparently there are "agents" and "elites" out there that are infiltrating AGW denier blogs such as WUWT and Climate Etc trying to make them less denialist. ... what? I found pages of these comments, warning about "dark agents from California" working to subvert society.`