Jump to content
Sign in to follow this  
Paul

Manmade Climate Change Discussion

Recommended Posts

Mont Pelerin Society Revealed As Home To Leading Pushers Of Climate Science Denial

 

THERE’S a popular talking point coming from climate change denialists that all people who accept the science and the need to act on it are somehow blinded by faith.

 

In Australia, climate science contrarian columnists can barely touch their keyboards without typing out the words "global warming faith" or explaining how human-caused global warming is some sort of "new religion".

 

http://www.desmogblog.com/2014/01/15/exclusive-mont-pelerin-society-revealed-home-leading-pushers-climate-science-denial

Share this post


Link to post
Share on other sites

Regarding climate modelling what is your take on Tamino's posts on smoothing, #1484, as this is way outside my area of expertise. Bearing in mind of course you are not a fan of Tamino.Posted Image

 

Finally got around to reading this. It's definetely interesting. The main problem for least-squares regression has always been end-points. This is a consequence of the maths. For instance a polynomial of degree one is a linear trend, and we all know the problem of cherry-picking end points to demonstate a fictitious trend one way of the other.

 

For instance, here's some of my code - you should be able to use Excel's macro features to step through the code if you want, that shows how the calculation can actually be done: at least it was my approach.

Option ExplicitPublic Type POINT    X As Single    Y As SingleEnd TypePublic Function Trend(Data() As POINT, ByVal Degree As Long, Optional ByRef equation As String) As POINT()    'degree 1 = straight line y=a+bx    'degree n = polynomials!!    Dim a() As Single    Dim Ai() As Single    Dim B() As Single    Dim P() As Single    Dim SigmaA() As Single    Dim SigmaP() As Single    Dim PointCount As Long    Dim MaxTerm As Long    Dim m As Long, n As Long    Dim i As Long, j As Long    Dim Ret() As POINT    Degree = Degree + 1    MaxTerm = (2 * (Degree - 1))    PointCount = UBound(Data) + 1    ReDim SigmaA(MaxTerm - 1)    ReDim SigmaP(MaxTerm - 1)    ' Get the coefficients lists for matrices A, and P    For m = 0 To (MaxTerm - 1)	    For n = 0 To (PointCount - 1)		    SigmaA(m) = SigmaA(m) + (Data(n).X ^ (m + 1))		    SigmaP(m) = SigmaP(m) + ((Data(n).X ^ m) * Data(n).Y)	    Next    Next    ' Create Matrix A, and fill in the coefficients    ' Note the matrix is twice as wide in order to augment the idetity matrix    ReDim a(Degree - 1, Degree - 1)    For i = 0 To (Degree - 1)	    For j = 0 To (Degree - 1)		    If i = 0 And j = 0 Then			    a(i, j) = PointCount		    Else			   a(i, j) = SigmaA((i + j) - 1)		    End If	    Next    Next    ' Create Matrix P, and fill in the coefficients    ReDim P(Degree - 1, 0)    For i = 0 To (Degree - 1)	    P(i, 0) = SigmaP(i)    Next    ' We have A, and P of AB=P, so we can solve B because B=AiP    Ai = MxInverse(a)    B = MxMultiplyCV(Ai, P)    ' Now we solve the equations and generate the list of points    PointCount = PointCount - 1    ReDim Ret(PointCount)    ' Work out non exponential first term    For i = 0 To PointCount	    Ret(i).X = Data(i).X	    Ret(i).Y = B(0, 0)    Next    ' Work out other exponential terms including exp 1    For i = 0 To PointCount	    For j = 1 To Degree - 1		    Ret(i).Y = Ret(i).Y + (B(j, 0) * Ret(i).X ^ j)	    Next    Next    equation = "y=" & Format$(B(0, 0), "0.000")    For j = 1 To Degree - 1	    If Sgn(Format$(B(j, 0), "0.000")) = 1 Or Sgn(Format$(B(j, 0), "0.000")) = 0 Then		    equation = equation & "+" & Format$(B(j, 0), "0.000") & "x^" & j	    Else		    equation = equation & Format$(B(j, 0), "0.000") & "x^" & j	    End If    Next    Trend = RetEnd FunctionPublic Function MxMultiplyCV(Matrix1() As Single, ColumnVector() As Single) As Single()    Dim i As Long    Dim j As Long    Dim Rows As Long    Dim Cols As Long    Dim Ret() As Single    Rows = UBound(Matrix1, 1)    Cols = UBound(Matrix1, 2)   	 ReDim Ret(UBound(ColumnVector, 1), 0) 'returns a column vector    For i = 0 To Rows	    For j = 0 To Cols		    Ret(i, 0) = Ret(i, 0) + (Matrix1(i, j) * ColumnVector(j, 0))	    Next    Next    MxMultiplyCV = RetEnd FunctionPublic Function MxInverse(Matrix() As Single) As Single()    Dim i As Long    Dim j As Long    Dim Rows As Long    Dim Cols As Long    Dim Tmp() As Single    Dim Ret() As Single    Dim Degree As Long    Tmp = Matrix    Rows = UBound(Tmp, 1)    Cols = UBound(Tmp, 2)    Degree = Cols + 1    'Augment Identity matrix onto matrix M to get [M|I]    ReDim Preserve Tmp(Rows, (Degree * 2) - 1)    For i = Degree To (Degree * 2) - 1	    Tmp(i Mod Degree, i) = 1    Next    ' Now find the inverse using Gauss-Jordan Elimination which should get us [I|A-1]    MxGaussJordan Tmp    ' Copy the inverse (A-1) part to array to return    ReDim Ret(Rows, Cols)    For i = 0 To Rows	    For j = Degree To (Degree * 2) - 1		    Ret(i, j - Degree) = Tmp(i, j)	    Next    Next    MxInverse = RetEnd FunctionPublic Sub MxGaussJordan(Matrix() As Single)    Dim Rows As Long    Dim Cols As Long    Dim P As Long    Dim i As Long    Dim j As Long    Dim m As Single    Dim d As Single    Dim Pivot As Single    Rows = UBound(Matrix, 1)    Cols = UBound(Matrix, 2)    ' Reduce so we get the leading diagonal    For P = 0 To Rows	    Pivot = Matrix(P, P)	    For i = 0 To Rows		    If Not P = i Then			    m = Matrix(i, P) / Pivot			    For j = 0 To Cols				    Matrix(i, j) = Matrix(i, j) + (Matrix(P, j) * -m)			    Next		    End If	    Next    Next    'Divide through to get the identity matrix    'Note: the identity matrix may have very small values (close to zero)    'because of the way floating points are stored.    For i = 0 To Rows	    d = Matrix(i, i)	    For j = 0 To Cols		    Matrix(i, j) = Matrix(i, j) / d	    Next    NextEnd Sub

Now whilst polynomials (including degree one - a straight line) are increasingly unwarranted for doing any sort of trend analysis, Tamino didn't mention what they are useful for. They are incredibly useful in approximation theory. See here: https://en.wikipedia.org/wiki/Approximation_theory Perhaps an easier way to think about this is that when one does Fourier analysis you decompose a series into a number sine waves (or cosine waves) and using the most significant - those with the greatest amplitude - you can approximate functions: that's to say you can get, more or less, a quantitative (and perhaps qualitative) similar version that is less complex of the original series. This is how CDs and DVDs work.

Edited by Sparkicle

Share this post


Link to post
Share on other sites

 

Finally got around to reading this. It's definetely interesting. The main problem for least-squares regression has always been end-points. This is a consequence of the maths. For instance a polynomial of degree one is a linear trend, and we all know the problem of cherry-picking end points to demonstate a fictitious trend one way of the other.

Yes that's very true and been the cause of countless arguments.

 

Will now spend some time going through your code but not sure I'll be able to get my head around it. But thanks very much Sparks for the time, effort and explanation. Much appreciated by one who has always found this a tad mind boggling and the latter isn't getting any sharper I'm afraid. 

EDIT

When some years ago I studied Darwin for a while I made the mistake of buying R.A.Fisher's definitive book, "The genetical Theory of Natural Selection". Oops.

Edited by knocker

Share this post


Link to post
Share on other sites

No problems. Essentially what's going on in is

 

(i) Create Vandermonde matrix from series data (matrix A) - this is the matrix form of the partial derivatives of the residuals.

(ii) Create geometric progression column vector (matrix B )

(iii) Since A * B = P, we can solve to get B.

(iv) Matrices are not skew fields (are not rings that support division) so we can't simply say P/A = B , so we must find the inverse of A, A-1, giving us P * A-1 = B

(v) The resulting numbers in B are the terms of the polynomial - so degree one gives us = y=mx +c

 

At least I think that's right - this is from memory unfortunately. I haven't looked at this stuff in detail for years!!

 

Hope this helps - at least a little. Also, some 15 years down the line, I would never write this code like this, anymore: I would implement it using rational arithmetic to get better stability for the Guass Jordan reduction, and needless to say, it would be object-oriented, in line with the programming world's latest fad. At the very least, if you use this, change every instance of the word 'single' to 'double'

 

Personally, I am great fan of polynomials, but, as Tamino says, you have to very careful when you use them. I opted for a 9-poly because the end-points looked about right - really, no other reason. But if you look really carefully at my last adventure, you can see that the poly does not centre at the start - ie it's approximation is too warm at the beginning - although the right hand end-point looks good, if, maybe it's a little low. Effectively, it's a judgement call, and qualititative nature of the results - a reducing trend in the 30 year running standard deviation, is unaffected, so I went for it.

Edited by Sparkicle

Share this post


Link to post
Share on other sites

Thanks again. I think I'm getting some of that but one thing I can say for certain I have now got the Occluded Brain.

Share this post


Link to post
Share on other sites

OSLO, Jan 15 (Reuters) - Governments may have to extract vast amounts of greenhouse gases from the air by 2100 to achieve a target for limiting global warming, backed by trillion-dollar shifts towards clean energy, a draft U.N. report showed on Wednesday.

 

A 29-page summary for policymakers, seen by Reuters, says most scenarios show that rising world emissions will have to plunge by 40 to 70 percent between 2010 and 2050 to give a good chance of restricting warming to U.N. targets.

 

http://www.reuters.com/article/2014/01/15/climate-solutions-idUSL5N0KP1OI20140115

Share this post


Link to post
Share on other sites

It will no longer be just the atmospheric GHG's by 2100 though? The changes to the Arctic/Greenland and Antarctica will also play a part in the 'new climate' by that time?

 

I really do not know the impact that an ice free Arctic, from July through October, will bring to our climate system but I suspect that such a change would be hard to undo?

Share this post


Link to post
Share on other sites

http://forum.netweather.tv/topic/76448-scepticism-of-man-made-climate-change/page-49#entry2896905

 

He's only dangerous because people might be inclined to listen to him. And he keeps odd company.

 

 

“A disservice to the scientific methodâ€: climate scientists take on Richard Lindzen

 

A team of UK climate experts has published a critique of a talk given by climate skeptic scientist Richard Lindzen a few weeks ago. The event was organised by the Campaign To Repeal the Climate Change Act.

 

Lindzen, a Professor of Meteorology at the Massachusetts Institute of Technology, was speaking at the House of Commons in a meeting chaired by Christopher Monckton. While the authors of the critique could agree with Lindzen on some grounds, they also found some pretty glaring inaccuracies in his talk.

 

 

http://www.carbonbrief.org/blog/2012/04/climate-scientists-take-on-lindzen/

 

And if the devil could cast his net over this lot.

 

http://www.desmogblog.com/richard-lindzen

 

The Weekly Standard's Lindzen puff piece exemplifies the conservative media's climate failures
The Weekly Standard suggests we should gamble our future on the climate scientist who's been the wrongest, longest
 

 

Edited by knocker

Share this post


Link to post
Share on other sites

The ex-TV presenter Anthony Watts is all excited again. This time of a new paper, "Rate of Tree Carbon accumulation increases continuously with tree size", (link below). For some reason he seems to have effortlessly seamed into the field of dendrochronology and sees it as another opportunity to have another pop at Michael Mann, Personally I think he's up a tree, or is it off his tree? Can't help thinking of the famous quote, "Who will rid me of this rambling nutter".

 

 

From the “trees aren’t linear instruments and the Liebigs Law department†and the Smithsonian Tropical Research Institute, comes this story that suggests the older trees are, the less linear their tree ring growth might be, which has implications for “paleoclimatology†and Mann’s hockey stick temperature reconstructions from tree rings.

 

Anyway HotWhopper thought it worth a comment.

 

Running (tree) rings around Anthony Watts - Ignoramus Extraordinaire

Anthony Watts has a tendency to put his foot in his mouth any time he lets his fingers touch his keyboard. (Ha ha - do you like the contortionist imagery?)  This time he's showing off his ignoramus side again in a WUWT article with the headline: Bad news for Michael Mann’s ‘treemometers’ ? (Archived here.)

Share this post


Link to post
Share on other sites

Thanks again. I think I'm getting some of that but one thing I can say for certain I have now got the Occluded Brain.

 

One last link, and then that's it. Here's a page that does it in a similar way, but uses the matrix transpose. This should help out a lot more than my rambling posts!

 

http://mathworld.wolfram.com/LeastSquaresFittingPolynomial.html

 

The idea is sound though: a polynomial really is a generalisation of the the degree one linear line. It also has caveats which my code doesn't such as checking whether a matrix is invertible (my code doesn't check for this) etc etc

Share this post


Link to post
Share on other sites

Watts writes only:From the “settled science†department. It seems even Dr. Kevin Trenberth is now admitting to the cyclic influences of the AMO and PDO on global climate. Neither “carbon†nor “carbon dioxide†is mentioned in this article that cites Trenberth as saying: “The 1997 to ’98 El Niño event was a trigger for the changes in the Pacific, and I think that’s very probably the beginning of the hiatus,â€This is significant, as it represents a coming to terms with “the pause†not only by Nature, but by Trenberth too.- Followed by a lengthy quote from Nature.I find it curious this can provoke a foam flecked piece such as linked above.

Why cant you stick to your own thread. I and others can do it and I'm sure you can. Edited by Devonian

Share this post


Link to post
Share on other sites

I can't see the problem ? The planet warmed at a rapid rate and then slowed down that rate of warming as the oceans 'recharged' themselves? now we're ready for another bout of rapid warming.

 

This bout of warming has the damaged Arctic to aid it by offering up more open water/land than was available to soak up heat the last time we saw a warming spurt.

 

Had we seen global cooling then there might be more to talk about but the planet continued to warm up even through the 'hiatus' so all we have are two different rates of warming, fast and not as fast?

Share this post


Link to post
Share on other sites

I can't see the problem ? The planet warmed at a rapid rate and then slowed down that rate of warming as the oceans 'recharged' themselves? now we're ready for another bout of rapid warming.This bout of warming has the damaged Arctic to aid it by offering up more open water/land than was available to soak up heat the last time we saw a warming spurt.Had we seen global cooling then there might be more to talk about but the planet continued to warm up even through the 'hiatus' so all we have are two different rates of warming, fast and not as fast?

I think there is a underlying net anthro forcing and the climate will average to the sum of that and all the other forcings. As ever the question is how big that sum is, and how it will grow. I suspect '98 pushed us a bit fast and recently we're a bit behind but there is no reason why, looking at the various IPCC summaries, we might not see several decades of modest warming and still see a lot by 2100 as the big A really kicks in? Edited by Devonian

Share this post


Link to post
Share on other sites

Am I missing something because I thought that's what And Then there's Physics was saying. His quick calculation opting for a sooner rather later return to warming.

 

 

So, is it possible that the current slowdown/pause could persist until 2030? Well, the energy imbalance in 1998 was probably around 0.6 Wm-2. If we follow a BAU emission scenario, atmospheric CO2 could increase to around 450ppm by 2030. Using ΔF = 5.35 ln (C/Co) would suggest that anthropogenic forcings would increase by about 1 Wm-2 if a surface temperature standstill also persisted until 2030. So, we’d have an energy imbalance of between 1.5 and 2 Wm-2 and the oceans would be absorbing significantly more energy than they are today. I guess it’s possible, but it seems somewhat unlikely. The oceans would need to absorb an ever increasing fraction of the excess energy.

 

So, the basic idea seems plausible, but even from the article itself it appears that it’s not accepted by all. However, it would seem unlikely that one could maintain the current conditions for much longer unless the planet is able to have a significantly higher energy imbalance than today without driving faster surface warming. I’ve written this quite quickly, so there’s probably much more that could be said and there may be some things I’ve missed, or mis-interpreted. The article itself, however, ends with

Edited by knocker

Share this post


Link to post
Share on other sites

Dr. Andrew Dessler has been one of the most valuable players  on the climate research team for some time. On thursday he testified before the Senate Environment and Public Works Committee, Barbara Boxer’s domain.

 

http://www.epw.senate.gov/public/index.cfm?FuseAction=Files.View&FileStore_id=26edecac-2c6f-4f8e-ab90-962a7d074d06

Share this post


Link to post
Share on other sites

KL says.

Well you could blow me down with a feather

 

Or maybe the Solar Wind?

Edited by knocker

Share this post


Link to post
Share on other sites

It's odd to find the other place now holding up a beeb presentation when ,only a few days ago, they were calling it all sorts of names for having it's agenda formed by climate scientists?

 

Anyhoos, yet again we have folk telling us it's all 40 years away and that we live in a very different world today ( of our own making) so not to look at the L.I.A. as any kind of reference point for what to expect.....

 

No mention, by the solar scientists, of the other, terrestrial impacts, that were helping form the L.I.A. I note???

 

Meanwhile , back on planet earth.......

Share this post


Link to post
Share on other sites

http://forum.netweather.tv/topic/76448-scepticism-of-man-made-climate-change/page-49#entry2898634

 

 

With button pushing phrases, a hint of conspiracy theorising, some bog standard stereotyping and repetition for effect you dismiss the work of an entire science. Phew, a right tour de force.

 

My advice? Start a blog. Call it, um, I dunno, What Us Want Trashing, write the same stuff as in your post twice a day but just use different words (ideally, get some other people to do the same just for a change) and, eventually, the message might get through...

Edited by Devonian

Share this post


Link to post
Share on other sites
Guest
This topic is now closed to further replies.
Sign in to follow this  

×
×
  • Create New...