DESIGN: Terrain Values

[MrBaggins]

Still all just supposition...

Show me a map, or scientific reference stating that grasslands are solely human derived... and I'll agree with you.

Everything I've read about grasslands says they are formed mainly due to climate... not human influence, except in large deforestation.

Give referential proof, Gilgamensch

[Gilgamensch]

Again:

I am not saying it doesn't exist.

All historic references/books/whatever are saying that wherever humans went, they irrigated since they climbed of the trees and changed from collectors to 'farmers'. This is where we are starting the game (actually a few thousand years after ~4000 BC, they started around 10.000 years ago)

Only people with better irrigation/farming was able to expand.

You agree?

So if it is like this, the idea of your grasslands isn't the same of history. The Biomes are just generalisations. Nothing else.

We use plains, did you see any reference for plains in the Biome-definition? No, they do not exist, so shall we remove them then from the game?

[MrBaggins]

Originally posted by Gilgamensch
We use plains, did you see any reference for plains in the Biome-definition? No, they do not exist, so shall we remove them then from the game?

Yes I did, actually... They are a subclass of Temperate Grassland... just a short variety... so technically theres no such thing as Plains. However, for our purposes, it's a useful terrain type to differentiate between high precipitation and low precipitation grassland.

http://www.runet.edu/~swoodwar/CLASSES/GEOG235/biomes/tempgrass/prairie.html

Short Grasslands occur where there are the same climatic conditions that taller grasslands exist, but have less precipitation.. typically only 10-12 inches.

[MrBaggins]

(and I think this stuff we're finding will be fabulous to assist in map generation)

[Gilgamensch]

(and I think this stuff we're finding will be fabulous to assist in map generation)

not sure, it could be, at least there are some 'rules' in there.

But if plains are a subclass, shall we now get rid of plains?

[MrBaggins]

No... because its acceptable to model low and high precipitation grassland... they support different crop types. Low precipitation land is better for (e.g.) cattle grassing, and high precipitation better for (e.g.) thirsty crops, such as corn.

[Gilgamensch]

Still don't get the relation to map-creation

[MrBaggins]

You create the continents... shift them, to create mountain ranges... create weather patterns, which will be altered by the mountains, hills and oceans to create more rain in certain places...

The places that have most rain would be Jungle (and swamp?), then next Forests, then grasslands then plains and lastly desert.

[Gilgamensch]

would make sense, but how do you 'determine' wind paterns? Or shall we just assume it is from west to east (like in SMAC)?

[MrBaggins]

Yep (with one strong central flow over large bodies of water)... and I believe reversed near the equator, with a mirror image south of the equator.

I'm not an expert though...

[Gilgamensch]

That brings me to one question though:

Can (and if how) we determine the equator?

Brings me to another point:

One thing I never liked that the poles are always at 0 and max. Can we not randomise it? Would bring a bit more surprise in, me thinks

For South of equator: IIRC you are right (Coriolis-force)

[Maquiladora]

Also the doughnut map has the same composition with the poles and climates etc, surely it would be different? Same goes for Neptune map. Perhaps you could exclude the poles from random flat maps, make it feel like a small scenario, instead of trying to look like the whole world.

[MrBaggins]

True... that climate would only exist for standard worlds

You'd not use it for flat maps.

I have no idea what you'd use for neptun and torus... since they won't really exist...

[Locutus]

The major problem in this terrain discussion (esp. Plains vs Grassland) is that the current terrain types basically don't make much sense. The designers (and basically these are Sid Meier & co, since Activision just copied their system from Civ1/Civ2) mixed up several ways of viewing terrain in a very confusing and random way: elevation, soil type, climate and vegetation.

If you look only at elevation, you get terrain types like Mountains, Hills, Plains, Coastal (and then some sea terrains). (If you want to properly model real terrain there should probably be even several more terrains in between: High Mountains, Low Hills, High Plains, etc.) In Civ and CtP, other types like Forest, Swamp, Desert, Grassland, etc are simply implied to be similar to Plains.

If you look only at vegetation, you get terrain types like Jungle, Forest, Grassland, Plains, Desert, Swamp. In Civ and CtP, Hills and Mountains and the like are simply implied to have a certain standard type of vegetation (Mountains are bare, Hills are covered in grass, Coasts are bare beaches).

If you look only at climate, you get terrain types like Jungle, Desert, Tundra, etc. Common terrain types like Hills, Grassland, Plains, etc are simply implied to have a moderate climate (though at least CtP differientates in different types of Hills and Mountains).

Finally, if you look only at soil type, you get terrain types like Rocks (Hills/Mountains), Clay (Forest), Sand (Plains, Desert), Loess (Grassland, Swamp), etc. They're usually not represented directly in CtP and Civ, but are implied.

In the real world of course, these factors (and others) can be combined almost without limit to create hundreds of different types of terrain: you can have low and flat terrain with a clay soil, lightly forested and a mild climate: the terrain you'll find in the Netherlands. Or you can have the same terrain but with a hot climate and you have something you might find in a Mediterrean country, or (more accurately) Mesopotamia before climate change and deforestation turned much of it into desert. You could also have high hills/low mountains, rocky soil, hot climate and little to no vegetation: fits much of inland Turkey and Iran. Change the climate to cold and you get Siberia. Or change the vegetation to dense woodland and you find yourself in South-East Asia or Northern Italy, depending on the type of wood (and climate, since that's very much connected to vegetation of course).

I'm not saying we should model such a complex system in CtP (though they can be useful considerations for the map generator), but we should realize how random our current system basically is, and possibly consider adding a few more terrain types like Forested Hills, Flood Plains, Wooded Plains/Open Woodland, etc.

[The Big Mc]

Originally posted by Locutus
I'm not saying we should model such a complex system in CtP (though they can be useful considerations for the map generator), but we should realize how random our current system basically is, and possibly consider adding a few more terrain types like Forested Hills, Flood Plains, Wooded Plains/Open Woodland, etc.

We are doing this to the underwater at the moment

as for land the only new one i know of is my post eruption volcano