Poor old Baroness Verma

Further to yesterday's revelation that the government has abandoned the global temperature record as evidence of manmade climate change comes this new exchange between our two ennobled gladiators:

Lord Donoughue (Labour): To ask Her Majesty’s Government, further to the Written Answer by Baroness Verma on 8 November 2012 (WA 224–5) stating that “the temperature rise since about 1880 is statistically significant” and the Written Answer by Baroness Verma on 21 January (WA 99) stating that the Government do not use “purely statistical models” to analyse global temperatures, whether they will reconsider the earlier assertion that the rise in global temperatures since 1880 is “statistically significant”.

Baroness Verma (Conservative):  With regards to the Written Answer I gave the Noble Lord on 8 November 2012 (Official Report, Column WA 224-5), I have nothing further to add beyond my previous answers on this subject.

I feel for Baroness Verma. Parliament has been misled and she has been made to look thoroughly dishonest. However, I think readers should be in little doubt that she is just a mouthpiece here, faithfully repeating the story she is told by the civil service. It would be ridiculous to suggest that she offered up her own opinion on statistical significance in the temperature records. I therefore have no reason to doubt her sincerity or her honesty.

But at this point there needs to be some accountability. What we cannot have is the noble baroness  pretending that everything is well in the world. Someone told her that the warming since 1880 was statistically significant. That was wrong. Who takes the blame?

Lord Donoughue (Labour)

To ask Her Majesty’s Government, further to the Written Answer by Baroness Verma on 8 November 2012 (WA 224–5) stating that “the temperature rise since about 1880 is statistically significant” and the Written Answer by Baroness Verma on 21 January (WA 99) stating that the Government do not use “purely statistical models” to analyse global temperatures, whether they will reconsider the earlier assertion that the rise in global temperatures since 1880 is “statistically significant”.

• Hansard source (Citation: HL Deb, 5 February 2014, c56W)

Baroness Verma (Conservative)

With regards to the Written Answer I gave the Noble Lord on 8 November 2012 (Official Report, Column WA 224-5), I have nothing further to add beyond my previous answers on this subject.

Comments

[Mydogsgotnonose]

Government and legislatures are becoming seriously split. Those in the pay of big carbon are under orders to keep the scam going. However, an increasing number of independent minds are realising just how deep the corruption has become.

The foot-soldiers like Verma are mere pawns in a big game, exemplified by Krebs yesterday on TV trying to maintain the mass hypnosis of the Public.

[johanna]

Baroness Verma is just reading the script which has been vetted by her uber-Minister's office and provided by his Department.

Still, if you want the prestige (and perhaps $$) that go with being a junior Minister, you have to be prepared to take the rough with the smooth.

[steveta_uk]

Just as I said on the previous Verma thread.

Her previous answer reads as she "will no longer answer questions about statistical models".

[mike fowle]

I have nothing further to add is standard Whitehall code for you've got me bang to rights, gov.

[Richard Drake]

Lord Donoughue is doing a sterling job.

[tallbloke]

Are the emails between Baroness Verma and her shadowy advisers not FOIable then? Open govt Dave?? Hmm?

[Don Keiller]

Given Baroness Verma's last stonewalling response "I have nothing further to add beyond my previous answers on this subject." Where does Lord Donoughue go from here?

It is clear that the puppet-masters pulling Verma's strings have decided that "enough is enough" and that they will not say anything more on a subject that may well be self-incriminating.

What mechanisms exist to force an actual informative response?

[agouts]

I am intrigued to know to what extent it is considered acceptable for a government minister to refuse to answer a perfectly legitimate question in Parliament. Where indeed does Donoughue go from here? What recourse is open to him?

[Martyn]

Its the civil service wot did it!

But who in the civil service?

[JamesG]

Well it was always clear that the scientists distrust observations anyway, and now we know from the AR5 commitee grilling of Allen, Hoskins and Stott that they they distrust models too. All we have left is their expert opinion with large uncertainty bars - which with a couple of quid might almost buy you a coffee.

[Headless Chicken]

Who takes the blame?

Might as well be me. get blamed for everything else.

Wish I'd never visited the grassy knoll

[Sherlock1]

'Poor old Baroness Verma'...

Er - no. She professes to be the government's 'spokesperson' on this subject - so she has to take the flack when questions get near the truth...

[David Ashton]

Off topic, please delete if you think inappropriate here. I have just watched the last ten minutes of Daily Politics, Andrew Neil's guests were Geoffrey Lean and Lord May. Lord May commented that there were no/few scientists working on the sceptical side, mentioned a BBC meeting in the past in which Bjorn Lomborg was presenting sceptical view point. He then called Lomborg a charlatan. I don't agree with everything Lomborg writes, but a "charlatan", surely not.

[A C Osborn]

But the temperature rise since 1880 is Statistically Significant. It shows that it warmed since 1880.
What it doesn't show is any attribution.

[diogenese2]

David Ashton @ 1.27 : Lomberg's field is Political Science and Economics. His "climate" chapters in "The Sceptical Environmentalist" are based on the IPCC SAR & TAR of 1996 & 2001, in which he does not dispute anything in the "attribution" working group but strongly disputes the "impacts" conclusions. The I-pod recording in not yet available so I have not yet seen the context in which Lomberg's work was raised, but Lord May is an ecologist so one would perhaps expect a somewhat more nuanced critique of Lomberg than "he is a charlatan".

[David S]

Shame AN didn't ask him whether he considered ad hominem arguments had any place in a serious debate.

[Roy]

Martyn wrote:

Its the civil service wot did it!

But who in the civil service?

Somebody who will get a knighthood or a similar honour when he/she retires.

[David Ashton]

Diogenes2, thanks for your comments but I am fully aware of Lomborg's position, on climate change, and his academic background, the TSE is on my bookshelf having been read cover to cover. His work was not discussed on the DP, Lord May simply commented that Lomborg had attended a BBC meeting and then proceeded to call him a Charlatan. As you say he doesn't dispute the AGW theory (or didn't when he wrote the book) , I believe wrongly, but then he isn't a climate scientist. He used his skills to show that the solutions being proposed we're worse than the threat. He certainly isn't a Charlatan.

[Peter C]

@ Don K "What mechanisms exist to force an actual informative response?"

In the UK, none. In fact one of the 'arts of politics' is to never give an honest or truly informative answer, after all, you might be held to account. Better to give a non-answer that sounds good but intrinsically means nothing, even better to give an answer that has multitudinous possible interpretations. That said, one only has to listen to any parliamentary exchange to realise that the most common tactic is simply to answer a different question than was actually asked then stonewall as having already answered.

All great fun for them, but then they wonder why the public has little more than contempt for them. There was an exchange between Cameron and Milliband yesterday where Cameron claimed his government outspent the last Labour government on flood protection and Milliband claimed it was untrue. The actuality was that in raw pounds Cameron was right, but in inflation adjusted pounds Milliband was right. And the essence? Tory spend £4.2 bn, Labour (adjusted) spend £4.5 bn. In other words, they each spent near enough the same but bringing it into the argument nicely distracted from any debate as to how well or how appropriately the money was spent, never mind if it was too much or too little.

[Frank]

Bish and Doug: I think you are over-emphasizing the technical inaccuracies that have been made by the government and the IPCC with regard to statistical models. "Statistical significance" can be established with physical models and there is nothing wrong with this approach to science. Indeed, most scientific progress has been accomplished using physical, not statistical models. Consider an object falling through the atmosphere. Physics tells us to analyze a time series for the distance fallen using a quadratic model, but air resistance (which varies with velocity squared) needs to be taken into account. Some domains are characterized by a quadratic relationship and some show a linear relationship (terminal velocity, where the force of gravity is compensated by air resistance. If our time series contained noise similar to that in temperature data, it is unlikely that we would learn much from purely statistical analysis of the data. The standard approach would be to construct a hypothesis that the distance fallen is the result of a constant acceleration opposed by a force that varies with velocity squared, the density of the atmosphere (r) and the projected area of the object (A): d2z/dt2 = g - (kAr/m)*(ds/dt)^2 With some experimental testing, this hypothesis would turn into a established theory with values for g and k. More experiments would tell us how closely theory can be trusted to agree with observations and what situations it gives good or bad predictions. Does it work well for objects falling through water? What happens at the transition between laminar and turbulent flow?

In a prophetic 1991 paper titled "Chaos, spontaneous climatic variations and detection of the greenhouse effect", Ed Lorenz predicted the dilemma that confronts climate science today: 1) We don't have enough data on decadal natural variation in temperature to permit purely statistical models to detect AGW. 2) Physical models are an acceptable substitute ONLY IF those models have not been "tuned" to match the historical record. One can't tune the models to match the historical record when we don't know how how big a role natural variation played in creating that record! Since models with high climate sensitivity invariably have high sensitivity to aerosols and models with low climate sensitivity have low sensitivity to aerosols, unambiguous evidence exists that models have been tuned (intentionally or unintentionally) to match the 20th century temperature record. (From a pragmatic perspective, what government would continue to fund a model that didn't reproduce the historical record?) Therefore, neither statistical NOR physical models are suitable for attribution studies. Lorenz's paper could be a powerful weapon against the consensus.

http://eaps4.mit.edu/research/Lorenz/Chaos_spontaneous_greenhouse_1991.pdf

In my "model" for an object falling through the atmosphere, there is only one adjustable parameter, the coefficient of drag (k). It is easy to validate that model by varying the mass of the falling object (m), its area (A) and the density of the air (r). If a single value for the coefficient of drag (k) doesn't work for all situations, the model is invalid (for those situations, at a minimum). Climate models contain several dozen parameters that describe subgrid processes (cloud formation, precipitation, vertical thermal diffusion in the ocean, etc.) These parameters can be optimized one at a time in the IPCC's sophisticated climate models or explored via ensembles of simplified climate models (Stainforth et al). The work with ensembles of simplified models failed to uncover an optimum set of parameters that reproduce all of the features of our climate (temperature, precipitation, seasonal changes, TOA fluxes, cloud cover) AT THE SAME TIME. The output from the IPCC's models doesn't explore the full range of future climates that are compatible with the laws of physics.

[Billy Liar]

Martyn wrote:

Its the civil service wot did it!

But who in the civil service?

Somebody who will get a knighthood or a similar honour when he/she retires.

Feb 6, 2014 at 3:31 PM | Roy

Somebody already did ... woe, woe and thrice woe!

[ptw]

With some experimental testing, this hypothesis would turn into a established theory with values for g and k. More experiments would tell us how closely theory can be trusted to agree with observations and what situations it gives good or bad predictions

experiments? ah
So you need to "observe" and "measure" variables..That's where statistics enter the scene.

[John M]

Here in the States, she would have the option of "pleading the fifth".

[Martyn]

We must be fast appoaching the the time when its more honourably not to be included on the upper end of the Queen's Honours list. Words that I would have associated with people at that end of the list like integrity and soundness seem to have been superseded in the recent years with words like notoriety and shame.

[Frank]

ptw: Finding an appropriate statistical model for a time series is quite different from using statistics to determine how well observations fit a physical model and to establish a 95% confidence interval for parameters (like g) used by a physical model. Science usually progresses by making a hypothesis - often by guessing at a physical model - and testing that hypothesis; not by searching for a statistical model. Doug's emphasis on the mistakes being made with statistical models might be considered misdirected, because the climate science consensus relies mostly on physical models. Baroness Verma has already publicly admitted relying on physical models, so her answer was correct: "[There is no need] to add beyond my previous answers on this subject."

Above, I combined separate theories of gravity (z = -0.5*g*t^2) and air resistance (F = kAr*v^2). These theories have been tested and validated separately, not as a combination. Likewise, theories covering light absorption, thermal infrared emission, fluid dynamics, conservation of energy, etc. have been throughly tested as separate phenomena and could be combined into a climate model that shouldn't require further validation (except for the software that implements the theories). The problems arise from: 1) uncertainty in the parameters used to approximate sub-grid processes (cloud formation, precipitation, thermal diffusion of heat) and 2) the fact that important convective phenomena in the atmosphere and eddies in the ocean occur on scales too small for large grid cells. (Grid cells are getting smaller, so problem 2) will diminish.) A generally accepted range exists for many parameters based on experimental or observational evidence, but the total uncertainty inherent in several dozen parameters that are reasonably well constrained can be very large. When the full "parameter space" is explored with ensembles of climate models by randomly varying parameter within their accepted range, the range of possible future climates associated with 2XCO2 expands significantly.