Climate Modelling

[madd0ct0r]

just for fun.

I want to try to produce a model of an Iceworld (think Hoth, but in the 40k Universe), then crash a spaceship on it.

The wreckage will produce heat at a constant rate.

grid will be three times as wide as high, and wrapped.

Each iteration the following energy should be added to each cell:

Solar - an incoming constant reduced exponentially with distance from equator, modified by that cell's albedo and the possibility of cloud cover (which here acts as an albedo modifier).

Bonus heat for the three cells that are the wreckage area.

and subtract Heat radiated (as a function of the cells temperature (ie hotter areas radiate more, again modified by cloud cover)

Each iteration, each cell will be averaged by the ones around it.


Once a cell hits zero degrees (they start pretty low) it transforms to liquid. I'm only really looking at the top 10m of the surface here, so that seems reasonable.

The number / percentage of surface water cells are used to generate the probability of clouds anywhere on the planet (I'm assuming high winds and a decent timestep)


anything I've missed?

[Ziggy Stardust]

I'm a bit confused as to what type of model you are trying to get here. A model of the planet's climate? Then what does the spaceship have to do with it? A model of how the spaceship effects the local environment through heat waste?

[Surlethe]

That's a pretty hot spaceship, to be able to affect an entire planet's climate.

[madd0ct0r]

well it is 40K. I've handwaved it as a warp reactor leak steadily converting matter around it to radioactive isotopes.

All I want to do is to let the planet run for a while, and see if it stabilizes at a at certain temperature. Then dump the sapceship in, and see if it can melt a small sea in the ice around it, and at what point equilibrium is restored.

I was hoping to do it in excel, but it looks like I might have to break out the C++.

[Ziggy Stardust]

I still don't get what you are looking for. Why do you want to model a 40k spaceship melting a hole in the ice? If this is for a story or a game or fluff or something, then is it important to have a complicated model? Just make up an arbitrary timeline. What is the point of this model? What are you even TRYING to model? I mean, break out the C++? What are you going to program? What on Earth are you even talking about?

You need to give people some better information here, beyond "Model a spaceship on an ice world!!!!1"

[Number Theoretic]

I could see a possible point in doing so: If the spaceship contains a really huge energy source that is now, after the crash, malfunctioning and bleeding out its energy as heat, a temperature model can answer you the following questions:

1) How big will the area around the spaceship that gets thawed by the waste heat?

2) Does this area spreads over the entire globe or will it reach a maximum size at some point and then start to shrink again, which brings us to the next question,

3) Does the wreck releases enough heat to eventually change the planet's climate on a global scale?

4) How fast does the heat spread from the spaceship? Can be important if there is a plot that is somehow dependent on that.

Concerning the tools to model this: there are better ones than Excel or C++. I would pick a combination of Sage and IPython+pylab or Matlab. Another excellent multi-domain modeling language is Modelica but support for partial differential equations is limited. With Python oder Matlab, i think, you can interface some Finite Element solvers, but i have to admit i haven't tried that.

[madd0ct0r]

I still don't get what you are looking for. Why do you want to model a 40k spaceship melting a hole in the ice? If this is for a story or a game or fluff or something, then is it important to have a complicated model? Just make up an arbitrary timeline. What is the point of this model? What are you even TRYING to model? I mean, break out the C++? What are you going to program? What on Earth are you even talking about?

You need to give people some better information here, beyond "Model a spaceship on an ice world!!!!1"

What more information do you need?
It's a vague background investigation I want to do. I'm writing a 40k Aquanautica supplement based around a campaign on this planet.
The planet's an ice world, with the only thing of note being the wreck of the starship from the crusade, and the larger irradiated lake/sea that's formed around it.

I'm pretty sure I've been pretty clear about what I'm trying to do.

This is 1) - to check what amount of heat output would be needed to meet the timescale
2) see what shape the sea ends up
3) all of number theoretic's questions
4) because I enjoy doing this sort of thing.

Yes I could hand wave it, set it as background or do lots of other things, but it's a few hours work and it's interesting.

What am I going to program?

OK. Start with an array 100x300, proceed as described in the OP.

It's an iterative timestep model with feedback loops - I'm trying to 'solve it' by letting it run and see what happens, instead of creating the partial differential equations (difficult) and solving them (very difficult). It's a crude, brute force way of doing it but for this low level it's fine, and intresting to watch.

My only worry is not having down graphics in C++ before, hence looking longingly at excel.

[Ziggy Stardust]

This is 1) - to check what amount of heat output would be needed to meet the timescale
2) see what shape the sea ends up
3) all of number theoretic's questions
4) because I enjoy doing this sort of thing.

Okay, from my point of view, there are just so many variables involved I don't know how you could model it without handwaving things anyway.

What do we know about waste heat form a warp engine? Very little (maybe some passing references in the fluff or rules that could be used to post upper/lower limits). So what kind of engine are we going to use as a stand-in? Is the engine fully functional, or melting down? How fast? How big is the spaceship? What kind (since this is 40K, there will be major differences depending on the faction)? Is it the entire spaceship relatively intact, or is it broken up? To what degree is it broken up? Is the wreckage irradiated? What about other sources of energy beside the engine? How large is the impact crater? What impact does unspent ordinance have? Or detonated ordinance? How OLD is the wreck? Are there still fires burning, or has it been there for a thousand years?

Where on the planet did the ship crash? Polar, equatorial, somewhere in between? What is the terrain like? Is this tundra, ice sheet atop an ocean like the arctic, glacier over land, a continental mass like antarctica? A giant iceberg? Is this the planet approximately earth-sized? How far from the sun? What are the parameters of its orbit? Weather? What effect do continued snow/precipitation, seasonal cycles, freezing, etc. have?

In what time frame is it possible for the engine to continue to dump enough heat into the environment to maintain an irradiated lake without it being refrozen? On Earth, if you bomb a spot through the ice in the Arctic, a year hence you would never find it, because the ice isn't permanent, it is constantly reforming, moving with ocean currents, etc.

etc. etc.

It just seems to me that in order to actually frame any possible mathematical model, you are going to need to simplify things and make so many assumptions that you are basically handwaving the entire thing ANYWAY, so why go through such a convoluted process when (since this is a game anyway) you can declare the environment to be X by fiat?

That's just my take on this. I am not saying you shouldn't do this, because you say you enjoy doing this, but it just seems to me that there are a lot of things you have to take into account for this model to work.

[Number Theoretic]

Of course, there are indeed a lot of assumptions that have to be made (or to be examined in separate models and then use their answers as input to create the initial conditions of the heat spread model). But apart from the fun in modeling, i see another reason why one want to explicitly model this: Consistency of the observed environmental phenomena with the stated initial situation. Also, it is possible that the models gives you some surprising answers, that you couldn't have guessed before explicitly modeling it. And maybe these surprising answers make the story even more interesting.

[Ziggy Stardust]

Fair enough. I was being a bit harsh with my earlier posts, when I think about it more I actually do think it's an interesting situation to model. I still stand by my last post, in that the vast number of variables make it an unnecessarily complex model. In any case, here is a link that might help.