Climate Classification System

[RonHiler]

In an effort to find a good climate classification system (which I need to have in order to do the next part of the race creation screens, which are all about habitibility ranges), I ran across this site:

http://en.wikipedia.org/wiki/K%C3%B6...classification

This could be a good starting place for us. We would need to be able to take the data from the world builder and assign a climate classification to each map point.

In the race builder screens, we'd have the broad catagories (tropical rain forest, tropical monsoon, savanna, etc.) and allow the user to set thier race's habitibility for that type of catagory.

I haven't fully digested the site yet, so there may be complications. One that I already see is that some of the classifications depend on how much rain falls (or sometimes temperatures) during certain parts of the year, which we won't have from our world builder, so we'll have to figure out some sort of adaptation for that.

Anyway, look it over, and if you have any comments or suggestions, or if you have a better idea or reference, please do let me know.

[RonHiler]

Even better:

http://en.wikipedia.org/wiki/Holdridge_Life_Zones

I like this one better because it doesn't rely on differences between monthly averages. It is only based on lattitude (which we have) [or alternately, we could use the mean temperature], precipitation (which we have), and something called the Potential evapotranspiration ratio (which, from what I can gather, I think we can calcuate from evaporation and biomass).

Also its classifications are less obscure. Everyone knows what a tropical rain forest is, or a subtropical desert, and so on. On the somewhat bad side, there are 38 of them! I'm not sure I am wanting quite that many types of terrain, but it might be okay, depending on how we organize them.

[Pix]

I remember the Koppen system from college. A couple of points--it has 20-30 zones, similar to the other one. Note that all the zones have names as well as the letter classifications, which themselves are meaningful and generally translate directly into a fairly obvious translation (rainy in the summer, rainy in the winter, rainy all year round, or whatever).

[RonHiler]

Alright, after much thought, I may have a system that will work for us. Tell me what you guys think.

At first I was just going to list all 38 (or whatever it is) types of terrain, and let the user set their tolerances for that type (in terms of temperature and water), and we'd set the population growth of a city based on those values.

I had three problems with that system, however. The first is it's pretty cumbersome (I mean, setting 38 terrain types just the way you want them would be a serious chore). The second was it didn't make a whole lot of sense. It seems logical that if you could tolerate a Boreal desert and a Warm Temperate desert pretty well, then you should be pretty tolerant to a Cool Temperate desert as well (which is right in between them). But with the aforementioned system, there is the possibility that wouldn't be the case, which is a bit silly (or we'd have to come up with a system that would allow the program to set sympathetic values, but good god I don't even want to think about the complexity of it, not to mention it would be totally obscure to the player, which is bad). And the third is that all those classifications, while they're fairly common named, don't always make a ton of sense to a casual user. I mean, what really is the difference between a Boreal desert and a Cool Temperate desert? Or Boreal Dry Scrub, Cool Temperate Steppe, and Warm Temperate Dry Forest? It is, to be sure, a bit obscure.

But then I thought of a possible solution to all three problems. The Holdridge Life Zone classification system (which is what we are using) is based on three variables. Tempature, annual rainfall, and the potential evapotranspiration ratio (PET) [which is essentially just evaporation plus uptake of water by biomass].

So, what if we give the user those three variables and allow them to set, for each, a low end, a high end, and move around the range? So in total they'd have 9 variables to play with, which is totally reasonable. And just so they know what those variables do, we'd give them a graphic (which is my version of the Holdridge graph) like the one I put together this morning (I'm not QUITE done with it, but pretty close)

www.rjcyberware.com/Graphics/misc/Lifezones.png

We'd overlay this with a green zone showing where their best habitats are (and then ring that with a yellow and red zone showing lesser habitats). That way they can see the results of what they are doing when they change their temps or rainfall or PET tolerances, but they don't have to worry so much about what a Boreal desert is.

What do y'all think? Is it a workable solution? Do you guys think everyone would understand what is going on?

[RonHiler]

Hey guys,

I'm still working on the habitibility stuff. The holdup at this point is that the six lines we have (two each for tempature, rain, and evaporation) can intersect in a great variety of ways, and determining which intersection points to use for the graphic overlay is pretty tricky, and there are as many as six intersection points (and as few as three). It's like one of those horrendous logic puzzles I used to play as a kid, hehe. But anyway, I'll figure it out.

I've spent a great deal of time and energy working on this system, and I hope I'm not making a mistake. In a lot of ways, the system works very well for us (we have all three variables inherent in the world creation proccess, it accounts for similar terrain synergies, and so on as I outlined in my other post on the subject).

But it does seem to have a couple of deficiencies as well. There are two problems I see in the chart.

One of them is the forests. According to the chart, almost everything with temps > 3 degrees C with more than 25 cm of rainfall per year is forest. This is somewhat offset by the min evap bar (which, if it's more than about 4, will shift us over to scrub and desert area). But even so, I'm a bit worried that vast chunks of our landmasses are going to be classified as some form of forest. It doesn't seem particularly balanced to me.

Another thing that concerns me is the lack of some terrain types. Where, for instance, are the jungles? What about the swamps? Or the grasslands? Essentially, at this point, we have forests (in various forms), scrub, desert, and tundra. Which seems to leave a lot of terrain types out of the equation. Should we take this chart as a base terrain type and then try to add subterrain types on top of them? Could jungle be a subterrain of one of the wet or rain forests? Grasslands as part of the Steppe terrain? If so, how do we assign these subtypes?

Obviously there are (unfortunately) some unanswered questions here. If you guys have any thoughts, I'd love to hear them. I probably won't try to tackle these issues with this build (as it's already taken as long as the first three builds combined, and I'm still not even done with it), so I'll probably just stop with what we have at this point, but perhaps we can make another pass at the habitibility stuff in the M5 build (along with the first part of the world building code).

[Pix]

I took a course a long time ago on geography, so some of this is vaguely familiar.

IIRC grasslands are generally an intermediate stage--given time and sufficient water, forests will tend to spread into them. You might consider the range of scrub/steppe/dry forest/moist forest to be potential grasslands, perhaps picking an arbitrary region on the edge of a forest to become grasslands. Jungles might be an alternate type of or alternative to wet/rain forest (I'm trying to think--I think Civilization had jungles but no rain forests) and swamps would be in a similar position, but also relating to groundwater percent coverage. A swamp could be either a grassland or a forest that has a lot of surface water. Go back and see if you can find a map showing this classification system applied to the surface of the world, and if so, you might be able to make some judicious guesses.

[RonHiler]

I took a course a long time ago on geography, so some of this is vaguely familiar.

IIRC grasslands are generally an intermediate stage--given time and sufficient water, forests will tend to spread into them. You might consider the range of scrub/steppe/dry forest/moist forest to be potential grasslands, perhaps picking an arbitrary region on the edge of a forest to become grasslands.

Ah, I didn't know that. Cool. That will definitely help, and seems like a reasonable idea.

Jungles might be an alternate type of or alternative to wet/rain forest (I'm trying to think--I think Civilization had jungles but no rain forests) and swamps would be in a similar position, but also relating to groundwater percent coverage. A swamp could be either a grassland or a forest that has a lot of surface water.

Yeah, that's kind of what I was figuring too. That we would assign the base terrain type strictly on the chart as we have it, then go back over it and reassign some regions based on their base types and a bit of extra data (like, as you say, ground water, maybe elevation, and so on). I just don't know how many other types we need and how exactly to assign them. Be prepared to delve into the depths of your memory of those classes Pix , I'll probably be asking you a lot of questions on the subject when we get to that point.

I think it's a good plan to wait until the world builder is on its first pass before taking a second pass at this habitibility stuff so we'll have a better idea of what kinds of variables we have available to us (it's one thing to have them down on paper, but entirely another when you have it working in code).

Go back and see if you can find a map showing this classification system applied to the surface of the world, and if so, you might be able to make some judicious guesses.

Aye, there are such maps, I ran across them several times during my initial research, so for sure I'll be going back to them and checking it out.