Building Models

Ricky Gervais didn’t mince words at the Golden Globes:

Delingpole: Ricky Gervais Deserves a Medal for Roasting the Wankerati at the Golden Globes

At the same event, Jennifer Aniston read out a message from Russell Crowe:

“Make no mistake, the tragedy unfolding in Australia is climate change-based,” he wrote about his home country. “We need to act based on science, move our global workforce to renewable energy and respect our planet for the unique and amazing place it is. That way we all have a future. Thank you. “

How much does Russell Crowe know about climate science? I’ve no idea. Maybe he knows a lot more than I do.

But I’ve been trying to find out more about it. So for the past two years I’ve been building my own climate simulation model, slowly adding new elements to it, to try to make it more realistic.

My latest model has a polyhedral Earth with 80 triangular faces, and I’ve drawn up an outline map of its continents, in approximately the right places, using 24 of the triangles. I’ve also got a Keplerian model of the Earth’s elliptical orbit around the Sun, which I use to find the intensity of sunlight throughout the year on each of the triangles. The air temperature above these triangles is pretty much completely defined by the albedo of the surface – how much of this sunshine is absorbed at the surface, and how much is reflected back into space. So I can calculate air temperatures throughout the year.

I then simulate the start of an ice age by dropping water on all 80 triangles. If the air temperature is above 0º C, it falls as rain. If below 0º C, it falls as snow. I assume that snow melts when it lands on the oceans. Using a continental albedo of 0.2 the snow only settles at high latitudes, as shown on the map below of all 80 triangular faces:

With a slightly more reflective albedo of 0.3 (granite rock), the snow settles further south:

The numbers on the triangles give snow depth in metres. If albedo is raised to 0.4 (desert sand), the snow settles as far south as 30º N:

Once the snow starts settling it gradually gets deeper. And the surface rocks beneath the insulating snow gradually warm up, and begin to melt the overlying snow. And so even though it continues to snow, the snow sheets gradually thin and melt. In this model the snow only gets to be about 600 – 700 metres deep at high latitudes near the poles. But it all gradually melts away as the underlying surface rock heats up. It melts slowly at high latitudes, quickly at low latitudes. After it has all melted, there’s a short pause (an interglacial period) while the surface rocks cool down again, after which the snow starts to settle and build up again.

It’s a very simple numerical model. And it runs on my home computer. And it has all sorts of missing elements (e.g. air movement, ocean currents, to name just two). But it’s already proving instructive.

For example, most maps of global glaciation during the last ice age show the snow confined to high latitudes above 50º N. But I’m seeing snow settling much further south. Might the discrepancy be because, while ice sheets several kilometres thick will leave moraines and drumlins as evidence of their past presence, snow sheets a few hundred metres deep probably will not leave much evidence? Perhaps, as my model improves, the snow will become confined to high latitudes. But right now I’m inclined to think it may have got much further south. And I have good reason to believe so: my own calculations show it.

And I can’t see any other way to form an opinion about the Earth’s climate, except by building my own model. It’s like making a paper plane to see if it will fly: how else do you find out? If you haven’t got your own model, you’ll simply be relying on someone else’s. How do you know if they’ve got it right?

Maybe Russell Crowe has got his own much better climate simulation model running on a supercomputer, and when he’s not making movies he spends his time tinkering with it, adding mountains and tectonic plates and volcanoes? Maybe Jennifer Aniston has got an even better one than his?

But I suspect that neither of them has any model at all, and they are just repeating what they’ve been told to say. But isn’t that what actors are supposed to do: read from a script?