Wednesday, December 04, 2013
Positively Against Positive Feedback
John Brignell makes an interesting point about the climate alarmists' claims about positive feedback:
It is the vital secret ingredient that is difficult for most people to understand. It is therefore more valuable to the computer-aided conjurer than all the smoke. mirrors and props are to the stage magician.
It is the vital secret ingredient that is difficult for most people to understand. It is therefore more valuable to the computer-aided conjurer than all the smoke. mirrors and props are to the stage magician.
Feedback is a potent component of system design, creation and description. Within certain restrictions, negative feedback reduces gain and is a stabilising mechanism, while positive feedback increases gain and can lead to autonomous instability. We recently noted that the climate alarmists had almost exclusively concentrated on proposing mechanisms of the positive sort. In fact, both the theory and the modelling on which climate alarmism is based are highly dependent on the selection of putative positive feedback mechanisms and the corresponding relative diminution of negative ones.
Even the some of the simplest of digital computer models defy convenient forms of stability analysis: it is so easy for such systems to include non-linearity, memory, thresholds etc. We are then reduced to testing with simple trial input signals (such as pulse or step). The warmist modellers are so secretive that we do not know what tests, if any, have been applied. The simplest explanation of the common behaviour of the climate models is that they are all inherently unstable, yielding a ramp-like output, almost regardless of input. Ancient modellers, especially those who go as far back as the days of analogue computers, will be familiar with such behaviour in erroneous models of the real world.
Forty years ago, we were trying to teach students not to pre-load computer models with assumptions based their own prejudices, which would predetermine the outcome. Now, professors are doing just that and thereby achieving world fame. Appropriate point for a timely repetition ...
----------------------------------------------------------------------------------
I would go a bit further.
Negative feedback is a common feature of any complex mechanical system. It simple means that when the machine is getting outside its safe design parameters it activates some system of returning it to normal. A valve on a steam engine which opens at a particular pressure is one such.
It also applies to living systems. I once gave this lecture to a doctor who said "You're talking about endocrinology", which I hadn't known.
It should also be applied to social systems - which is why I like the US constitutional respect for "checks and balances". This is also part of why free speech and democracy are good things - in a dictatorship if things go wrong nobody but the dictator can change it and has no incetive to admit getting anything wrong, indeed they have an incentive to suppress any voice of sanity, whereas in a democracy voters notice (moreso when free speech prevails) and there is the pressure to fix it. That is why, in a democracy with free speech any really destructive doctrine will quickly fail.
Whereas where there are overwhelming barriers to new parties as in Brtain today & the USSR under Stalin (where new parties are disenfranchised by a corrupt electoral system or shot respectively) and state ownership of the media, destructive lunacies which have ideological approval (catastrophic warming & Lysenkoism respectively) become rigidly enforced. Indeed once they have become official truth and the governing class is sufficiently isolated rom reality, a positive feedback system develops.
In machinery there is a technical name for a device designed to use positive feedback. A bomb. In nature positive feedback occurs in forest fires and avalanches (at least until they run out of wood or snow). In medicine it is known as cancer.
You get the point. This ought to be a guiding principle in establishing any political governing system. Fire and government are dangerous servants and the latter is the more dangerous. The Norse understood this when they made Loki (Fire) the smartest but most untrustworthy of their gods.
And so back to climate science.
The climate alarmists claim positive feedback is dominant in the runaway warming they threaten. CO2 rise, they claim, will cause temperature rises of about 0.5C to 1 degree but "climate forcings" will multiply that by any number they feel like. No evidence of that has ever been produced - it is simply part of their computer models.
Any system governed by positive feedback is, as demonstrated, inherently unstable. But we have had climate on this planet for 4,500 million years and never once seen a runaway greenhouse effect. There have been periods of cooling - once with the planet completely or almost completely covered in ice which means there may be a net positive feedback effect from cooling (more ice means more sunlight reflected into space).
But with 4,500 million years without a heating explosion it is simply impossible that climate feedback on heating can be positive. It can be neutral, in which case we will get that 0.5-1C warming but odds are better in it being somewhere on the negative axis which means less, perhaps much less, than that.
Labels: British politics, global warming, Science/technology
