Climate models and deniers

After some introspective posts, it is time for Science!  I was going to write this one after the IPCC published the full version of the fifth report, so that my shotgun would be heavily loaded before I blast the sceptics.  But then I don’t need to wait as it’s clear from the preliminary release that their tactic is to be either to make things up, or pick on semantic discrepancies between paragraphs.  They are not going to be dismissing entire technical appendices with well supported and referenced research, which is what would be required to have any credibility.

Here, I am not a member of the general public, I have an excellent, broad, and continuous scientific education, and I am not thrown by their diversionary tactics.  There is a lot of talk about models too.  ‘The models didn’t show this.’  ‘The models did show that.’  All I ever hear is, ‘I don’t know anything about models’.

Gran Turismo 5, the game, on the PS3, is a model.  It models cars on racetracks, and it does it rather well.  It includes many relevant factors to a car at speed on tarmac.  Sprung mass, unsprung mass, drag coefficient, torque, grip, gradients, and many others I don’t know about because I am not an automotive engineer.  Within the model, things seem to be very realistic; if you lap an Aston Martin Vantage it will beat a Focus ST.  However, if you go to a real track featured in the game and drive both those cars, then it will not match the model.  The lap times would be different, and the gap between the cars different, and other variables would be off to.  Corner entry speeds, brake points, the gear you take a particular section in.  The game may be similar, but it would not be identical to reality.  The Aston would still win though.

One of the main features of the Gran Turismo series is the customisation of the cars.  These modifications produce realistic changes to the performance of the car, at least within the context of the game.  Changing the damper rate or rebound rate, adding boost pressure or reducing weight, will make the car faster, but possibly harder to drive.  A big spoiler on the back will break up the air flow and reduce drag, which increases top speed, but it also shifts the centre of gravity back, so the front wheels have less weight on them and will under steer more.  When braking, remembering that a car is a weight on springs, the centre of gravity will shift forward, but it now has further to go than before you modified the car, which may make it a bit awkward under braking.  Do this to the real car and the discrepancy between the game and reality grows, as the model only takes some of this in to account, while nature takes everything in to account in real time, at the atomic level.

This is one of the roles of the model in science, to allow you to play around and experiment with factors that you cannot in real life.  Even if we had access to a vast range of exotic cars and a track, I doubt any of us would pound round and round, tweaking the car till we had a huge crash, then revising our modifications, and carrying on again.

But in the model it is quick and simple to try out different scenarios.  Change the inputs and see how it affects the output; see what small changes produce dramatic differences.  It allows the principles that underlie the model to be explored.  Although all of the maths in the model is of human invention, the interaction of multiple layers of maths is beyond the ability of any human to instinctively understand; you need a computer to do it.  The maths may produce results that those who created the maths could never have expected.

The Mars lander, Curiosity, had the most accurate delivery system ever devised, with a landing area only 4 miles by 12 miles.  It eventually landed roughly one mile from the nominal bulls eye in that area.  After a flight of three hundred and fifty two million miles, that is a truly astonishing feat.  And it is all down to Newton, and very clever people.  The mission team used his equations of motion and gravity to plot the course that the craft should fly on its way to Gale crater.  An equation is a model; you put in variables measured from the real world, and it will give you a prediction that applies in the real world.  In the case of Newton these predictions are extremely accurate, if you stay well below the speed of light.  But the model is totally wrong.  Newton assumes there is a force of gravity; that satellites orbit due to an attractive force.  They don’t; they orbit because they are travelling in a straight line on a curved space time.  Einstein showed us that both mass and energy distort space time.  The mass of the Earth bends space time such that if you get above the atmosphere and travel at the right speed you never stop going round.  And you don’t have to keep turning left; you go in a straight line and the universe turns left for you.

That is the other use for models in science.  It lets you predict a system using a simpler method, while ignoring that actual mechanism found in the universe, so you can have more confidence in your predictions.  You could run a Mars mission using general relativity, and it would give a slightly more accurate prediction, but the error inherent in your flight hardware would eliminate that advantage.  Better to use the easier mathematics; less likely to make mistakes and easier to double check.  It cost two and a half billion dollars.  It’s important to get the sums right.

There is a third option, of a model which is accurate and uses the actual mechanism that the universe utilises, but we can never really tell we are in that situation.  We can only say the mechanism hasn’t been proved wrong yet.  See my post on the scientific method.

Climate models are in the Gran Turismo mould of things.  They simulate, or emulate parts of the factors involved in Earth’s climate.  Some are simpler than others, but none are detailed enough to be expected to make stunningly accurate predictions.  A scan through the TOP500, the list of fastest supercomputers, will show that systems used, and sometimes dedicated, to climate modelling, are regularly in the top twenty.  There are systems that model as many climate factors as possible, over the whole globe, but at a low resolution.  Other systems concentrate on one aspect of climate at higher resolutions.  But been able to do everything at high resolution for the whole world is a technology we do not have, and it may be more than a decade away; and we’re not a decade from the exaflop.  Sometime in the mid to late 2020’s someone will build a quarter billion dollar machine that can simultaneously handle all the different factors of climate that we have a mathematical representation for at a one kilometre or better resolution.  Probably.

Till then we model the parts separately, or globally, with fewer factors included.  Our approximation is more approximate.  But that isn’t a waste of time and effort.  Remember a model is there to explore the underlying principle in the thing been modelled; to try and find factors that are minor and factors that are powerful.  If you take ten models and run one hundred different scenarios through them, you get one thousand different answers.  None of which may be the correct answer; the truly accurate answer.  That doesn’t matter, because what it gives you is a data set, and analysing that will give you insight.

Anyone who has ever crashed in Gran Turismo will know that the model completely ignores damage.  Now the problem with that is that it introduces deep flaws in to the model and allows it to predict outcomes that will never ever happen.  If you are approaching a chicane, then instead of applying the brakes like the other cars, it is possible to plough in to the field ahead of you and use them as your brakes.  Your car slows much quicker than it would under braking, and you can go from eighth to third in half a second.  I know enough about cars to understand this is nonsense.  Climatologists understand that their models will give them nonsense outputs sometimes.  Or at least outputs that may teach them something about their model, but don’t directly apply to the Earth itself.

 

The deception employed by the sceptic is to exaggerate the capacity of the model, and then use the models inability to meet this imagined standard, to decry them.  The general public presume that if a sceptic credits a model with great predictive power, then it must have such, as they wouldn’t pay is such a compliment otherwise.  Having setup this falsehood, they then use the falsehood as evidence of the model being wrong.  But it’s false, so it’s not evidence.  The public have difficulty spotting this, and the public is where we draw our elected officials from, which adds to the problem.

All the different models and variables and runs, give us data.  Data gives us trends and indications.  The indication is warming.

It doesn’t matter if the models did or did not predict ‘the pause’.  The models can emulate ‘the pause’, so we can justifiably say we understand it.  As an engineer it’s clear to me that if you add energy to a system, but it doesn’t warm up, but it does have a heat sink, that there is no mystery where the heat has gone.  The climate has a heat sink, we call it ocean.  The atmosphere has not carried on warming, but the water has, especially the deep water.  We have directly measured this; it is not a supposition.

The biggest disappointment, in terms of those who argue against climate change, is that the physics is crushingly simple; near infrared, far infrared, harmonic frequencies, and reflectivity.  I have, in the past explained it to people with no real scientific education, people who think the greenhouse effect is something that might happen in the future.  The greenhouse effect has been present for four billion years and is the single most important contributor to Earth having life, as it keeps our water wet, as opposed to solid.  Normally I can make them understand that the Earth absorbs energy from the sun, and radiates energy too.  Some of the energy is stored here for a while, but there is a constant turnover.  The amount stored is determined by the thermal equilibrium of our climate.

The thermal equilibrium of a system is set by the amount and ratio of ingredients in the system.  Humans are changing some of these ingredients.  This will change the equilibrium.  We strongly believe it will change it in the positive direction; hotter.  We know that if we don’t change it, then things will be alright, but if we do change it we don’t know, because of the short comings of our models, that it will still be alright.  I fail to see the rational argument in knowingly changing the thermal equilibrium of our planet, whist simultaneously saying that we needn’t worry, and things will be fine, when you cannot possibly know that.

If not changing it is the best option, then changing it as little as possible seems then next best option.  Now the deniers like to say there are natural systems at play here that have very large effects, so let us leave the evidence aside for a moment and say that is true.  We cannot do anything about the natural systems, they must run their course, be we can affect the human contribution because we are the humans.  Having accepted that a small change is better than a large change, we should minimise our climate altering emissions wherever possible.

Anyway, oxidising carbon is a rubbish way to power a civilisation.  The rest of the universe runs off binding energy and so should we, whether it is the binding energy in our own reactors, or the reactor we orbit.

I should say here that I am very positive about the future.  The doom mongers are as wrong as the deniers.  Humans will adapt.  We are a very adaptable species; our technology makes us more so.  We will still be here in a few centuries, but the world will be different.  Some will not be pleased with the changes.  Parts of the Earth that are currently heavily populated may become uninhabited.  If a billion people choose to migrate then it will cause trouble, although there is nothing you can do to stop them, except moan afterwards.  Those of us who understood the physics all along will say ‘we told you so’.

Some describe this current era of Earth’s history as the Anthropocene, a recognition that human activity is dominating even geological activity in shaping habitat.  We are a powerful specie, and if we acknowledge that we are altering the climate to our detriment, we are only one step away from acknowledging we can alter it to our benefit.  Obviously, as I love science, and I love technology, for me geoengineering is an easy sell.  In fact I have a feeling that an industrial civilization of eleven billion cannot exist without it.  It would be an extension of the power, data, transport, and agriculture that such a civilisation would require.

The Earth has always had a very stable climate, compared to the rest of the solar system.  This is because it has a thermostat which runs via geological and biological processes, but it takes tens of millennium to respond to any changes.  We are getting near the point where we need to take over that role ourselves, and thanks to our models, we are getting near the point that we could.