PDA

View Full Version : [Award Winner] Using tectonic plates to draw a world map

dhalsimrocks
06-11-2008, 10:59 AM
I looked through all the other threads in this forum and didn't see anything related to it, so I thought I'd make my very first post here a tutorial.

WARNING - This tutorial gets somewhat technical

If you're anything like me (a science geek), your fantasy worlds have to be based on some degree of scientific fact, rather than total randomness. I've been working on a fantasy world for going on 13 years now, and I lost the original world map I had drawn. So I recently found myself thinking of making a new one from scratch.

Having learned about the prehistoric supercontinent, Pangea, many years ago, I thought about creating my new map by drawing a supercontinent and imagining how it would break up and form my world map. I researched previous supercontinents of Earth, such as Pangea, Laurasia, Gondwana and even older, Rodinia. Eventually, my research took me to geological basics: tectonic plates.

I then realized my method would not be in creating a fictional supercontinent and breaking it up, but by determining the existing tectonic plates of my fictional world and letting them decide what it would look like, along with a little deliberation on my part. The end result was pleasantly surprising. Now, on to the "How".

Step 1 - Drawing the Tectonic Plates

Note - please read the entire tutorial before proceeding with this step.

As a point of reference, examine the Earth's tectonic plates:

http://en.wikipedia.org/wiki/Image:Plates_tect2_en.svg

Keep this open as I'll be referring to it later in this guide.

Notice that there are 7 large, major plates and 8 small, minor plates (there are actually more minor plates, but for all intents and purposes in this guide, we'll say 8 ). Also notice that their shapes are not very erratic. There are very few inward curves.

So start by drawing a similar number of large and small plates. This is a very easy step, actually. Just as in the photo of the Earth's plates, the borders do not need to have a lot of randomness, like a coastline would. And keep in mind that the shape of the tectonic plate does not necessarily determine the shape of the continent (but it can).

The next step is to determine the direction of the plates' movement. This page shows the general movement direction of Earth's plates:

http://education.sdsc.edu/optiputer/flash/plateArrows.htm

Notice that there is a mix of directions. You should try for a similar mix. These plate directions will be a large influence on the terrain of your final map, so use some deliberation here, but make a few of them random (whatever direction comes off the top of your head) to make things interesting.

Here is an example of the plates and their directions for the map I recently worked on:

4271

Before we continue, I will need to digress and give a basic geology lesson.

dhalsimrocks
06-11-2008, 11:30 AM
Digression - Basic Tectonic Plate Geology

The movement of Earth's tectonic plates has an enormous effect on its terrain. Most important to this are the boundaries of the plates, and there are two factors that come into play: the type of crust at the boundaries and the type of boundary.

There are two types of crust: oceanic and continental. Oceanic crust is exactly that: crust that forms ocean floors. Continental crust is that which creates landmasses, although not all continental crust is above water.

There are 3 types of plate boundaries:

Transform boundaries, where two plates slide past each other (rub the palms of your hands together for a visual aid). California's San Adreas Fault is an example of a transform boundary. Transform boundaries produce many earthquakes.

Divergent boundaries, where two plates are moving away from each other, creating oceanic ridges and continental rifts. The Atlantic Ocean was produced by divergent plate boundaries between the North American + South American plates and the Eurasian + African plates. They were once joined together in the Pangea supercontinent, but the plates movement spread them apart and the divergent boundary created an ocean. Africa's Great Rift Valley is also being caused by a divergent boundary.

Convergent boundaries, where two plates are colliding. This creates mountains and/or oceanic trenches or island chains, depending on the types of crust that are colliding. Convergent boundaries create a process known as "subduction", where the crust of one plate slides underneath the crust of the other. This creates volcanic activity along the line of subduction. This will be important later.

Transform boundaries have a lesser impact on terrain than divergent or convergent boundaries, although it is important to remember that they do cause earthquakes!

Divergent boundaries occur between two continental crusts or two oceanic When two oceanic crusts meet at a divergent boundary, it creates an oceanic ridge, not to mention new crust. These are basically crust formation points, and you should have one or two on your map. Example: The Mid-Atlantic Ridge.

When two continental crusts diverge, you get a rift valley, and over time, they separate into two landmasses. Example: Africa's Great Rift Valley.

Convergent boundaries between two oceanic crusts eventually create chains of islands (usually in an arc). The subduction causes volcanic activity that eventually builds up into islands. Examples: Aleutian islands, Phillipines.

Convergent boundaries between oceanic crust and continental crust creates volcanic mountain ranges on the edge of the continental crust, because of subduction. Examples: Andes Mountains in South America, Cascade and Sierra Nevada mountains in the USA.

Convergent boundaries between continental crusts creates dramatic mountain ranges as the landmasses smash into each other and buckle the landscape. One crust can subduct beneath the other. The classic example is The Himalayas. This was caused when the Indian Plate crashed into the Eurasian Plate at a high rate of speed (still very slow, however). The Indian Plate is slowly subducting beneath the Eurasian and is creating the tallest mountains in the world, as well as the Tibetan Plateau (the highest in the world).

For a visual of most of the scenarios:

http://en.wikipedia.org/wiki/Image:Tectonic_plate_boundaries.png

Also, see the image of the Earth's plates.

So now that we know how the different boundaries and crusts effect the landscape, we can take the next step in producing our map.

ravells
06-11-2008, 11:55 AM
This is a very interesting and educational read! Thank you very much for posting! (I'll delete this message after a few days so as not to interrupt the flow of your tutorial).

jfrazierjr
06-11-2008, 12:05 PM
Thanks for the info AND the links. Though I will proabaly not delve this deep, it is nice to have an idea of what might be going on in the real world. I think for me, this will mainly help influence the idea behind why islands are where they are and how many of them there are.

I know some people over on the NBOS forums have had some debate about how most mapping programs that generate worlds don't take this into account and the general thing is that you would need hugh housepower or tons of time to even approximate getting this created right. Mainly, this is directed at Fractal World Explorer or Fractal Terrains autogeneration functions. For the most part, the people on that forum are fairly agnostic in how real geology affects world creation. Most people have the attitude (which I mostly have) that "its just a fantasy world, so as long as the rivers go down hill, thats probably good enough for me."

Joe

dhalsimrocks
06-11-2008, 12:24 PM
Step 2 - Drawing the Continental Crust

It is now time to decide where the continents will be and give them their shapes.

When I did this, I made a few copies of the tectonic plate sketch on which to draw my continents and landmasses.

But where to start?

One good place would be to find your divergent boundaries. This is a possible location for an oceanic ridge/crust formation point. This is also a possible location for two continents that were one time joined together, but are now separated, like South America and Africa. So you might want these two continents to have coastlines that would mostly "fit together", but separated by a body of water.

Once you have those coastlines drawn, think about the rest of the continent. The shape can be entirely up to you and can have a lot of randomness.

Next, think about your Oceanic Plates. When looking at the Earth's plates, you will notice that only two major plates have no continental crust at all: the Pacific Plate and the Nazca Plate. One of your large plates could be entirely oceanic, and would therefore be a location for one of your oceans, if not the largest. It would also be a good location for a "Pacific Ring/Ring of Fire" type scenario, where most of the plate's boundaries have a lot of volcanic and seismic activity, not to mention some of the deepest trenches where there is the most subduction.

Then look at your convergent boundaries. They will be the locations of either Himalaya type mountain ranges, Andes or Cascade volcanic mountain ranges, or volcanic island arcs. Decide what you would like to have and where. This can help you decide where your oceanic and continental crusts will be.

Finally, think about whether or not you want any continental divergent boundaries, where there may be a large rift in the continent (and a new plate boundary forming).

Most of this step is up to you, but is also largely determined by how you drew your tectonic plates and what direction they are moving. Take some time with this step and make it your own.

Here is what my map looked like after adding the continents (it's not the finished product):

4273

Continents A and B are separated by an ocean created by a divergent plate boundary. They were once together and as such, they would "fit" if placed side by side. Both would also fit with the southern continent, which I forgot to label, as that is also a divergent boundary.

A<->G, E<->B, F<->D and C<->D are all continental convergent boundaries and are creating large mountain ranges.

Notice the island arcs between all of the oceanic convergent boundaries.

All the way to the north is a continental rift caused by a continental divergent boundary. (x) will one day be its own landmass and tectonic plate.

The island between G and (x) has a ridge of mountains caused by a convergent boundary and subduction. B has one as well, but that's from a very old subduction that is not current.

Now, I'm sure a legitimate geologist would look at this and tell me a thousand little things I did wrong, but it has enough scientific base for me, and it has resulted in a realistic looking map. And what's more is that there was enough randomness involved that I didn't know exactly what it was going to look like in the end, so it was just a bit of a surprise.

I still have some more work and detailing to do before going on to editing this in the GIMP according to RobA's excellent guide.

Anyway, even if you don't use this whole method to create world maps, I do hope that there is information here that could be helpful in determining at least some of the terrain on your maps or help shape continents.

RPMiller
06-11-2008, 12:29 PM
I looked through all the other threads in this forum and didn't see anything related to it, so I thought I'd make my very first post here a tutorial.
Post #14 in this thread. ;)

I will definitely be reading through this thread with great interest. Thank you very much and have some rep! Oh, and Welcome to the Guild! :D

Karro
06-11-2008, 12:54 PM
In my worldmap (http://www.cartographersguild.com/showthread.php?t=2110), I'm using largely the same approach. I had a few basic ideas about where I wanted continents to go, and a few shapes I wanted to use based on some old maps I had done as a child, but mostly I wanted to fashion the world from scratch using some basic understanding of plate tectonics.

I'll be working on this more after I finish my little side-jaunt (http://www.cartographersguild.com/showthread.php?t=2089)learning how to use GIMP.

Redrobes
06-11-2008, 09:12 PM
Well done - this is shaping up to be a great thread and read. Keep it up ! Have some more rep...

HandsomeRob
06-12-2008, 09:37 AM
It may be important to remember that plate motion can't really be described by simply drawing an arrow on the plate and saying "it moves that way." Actual plate motion is actually a rotation around a point on the earth's surface. What you really need to know for this exercise is relative motion; it is completely possible to say that plate 1 is moving west (with respect to plate 2) and at the same time moving east (with respect to plate 3). This situation happens all the time.

Here is a quick tutorial on velocity space analysis:
Say you have a situation like in the first attachment here. Plates A and C are moving toward each other at a rate of 4 cm/yr. Plates A and B are moving away from each other at a rate of 3 cm/yr. What is the motion between plates B and C?

Construct a velocity space diagram to find out (see second attachment). Since plate B is moving west at 3 cm/yr (with respect to plate A), place a point for B 3 units to the west of A. Since plate C is moving south at 4 cm/yr (again, with respect to plate A), place a point for C 4 units to the south of A. Now it should be clear that plate C is moving southeast at 5 cm/yr with respect to plate B, as the point for C is 5 units to the southeast of the point for B.

This means that the B/C boundary is convergent, with an element of right-lateral strike-slip motion as well. The third attachment here shows the solution.

I hope some of that a) makes sense and b) helps.
-Rob

dhalsimrocks
06-12-2008, 10:05 AM
It may be important to remember that plate motion can't really be described by simply drawing an arrow on the plate and saying "it moves that way." Actual plate motion is actually a rotation around a point on the earth's surface. What you really need to know for this exercise is relative motion; it is completely possible to say that plate 1 is moving west (with respect to plate 2) and at the same time moving east (with respect to plate 3). This situation happens all the time.

Excellent information to add to this! Thanks! The plate tectonics rabbit hole goes pretty deep if you really get into it. And there is a lot more that what I said to be considered if you want total accuracy. (which is precisely what I meant when I said a legitimate geologist could find a thousand things I did wrong. ;) )

I initially entertained the thought of taking the rotational motion into consideration, but my desire to have a new map overtook my desire for scientific accuracy, and decided that it would add considerable complexity, and thus time, to my project. So that's why I stopped at the point of drawing an arrow on the plate and saying "it moves that way".

I take that back, I did take rotational motion into consideration for a few of the plates on my map, but I didn't note it. In particular, my southernmost plate is rotating anti clockwise.

Karro
06-12-2008, 10:22 AM
I think this can get even more complicated, based on an observation of real-world plate. Many of the plates have highly irregular shapes, and those with plate-peninsulas seem to have different parts of the plates moving in somewhat different directions relative to a neighboring plates. I assume this is do to the interaction of the forces from the main part of the continent's rotational velocity (which I believe is influenced by subterranean magma currents) and the friction and pressure caused by collision with neighboring plates. In some of these scenarios, it looks like a larger plate may be torn assunder by another plate smashing through it.

dhalsimrocks
06-12-2008, 10:34 AM
There are also several dozen or so more smaller plates on the Earth than that image I referenced in my first post. It's a lot more complex than "7 big ones and 8 small ones".

Edward Protera
07-30-2008, 06:37 PM
:!: Thank you thank you thank you for writing this tutorial.

This may be exactly what I need to get through those first few steps. I love this idea to death... xD

Anyway, as if I haven't said it enough already, thanks. :3

ravells
07-30-2008, 06:57 PM
If you like the tutorial do not forget to rate it! To rate the thread, look at the dark bar at the top of the thread - I'm guessing you might give it five stars!

Edward Protera
07-30-2008, 09:29 PM
If you like the tutorial do not forget to rate it! To rate the thread, look at the dark bar at the top of the thread - I'm guessing you might give it five stars!

Thank you for the tip, sir. I'll eventually get the hang of this place... :3

Also, I'm coming upon a couple of plate interactions you don't cover here... would you happen to know what happens when plates move perpendicular to each other? I mean, I can guess, obviously, but...

jfrazierjr
07-30-2008, 09:46 PM
Thank you for the tip, sir. I'll eventually get the hang of this place... :3

Also, I'm coming upon a couple of plate interactions you don't cover here... would you happen to know what happens when plates move perpendicular to each other? I mean, I can guess, obviously, but...

Bad earthquakes! But seriously, I have no idea (other than what I said in jest, though it is correct) I know somewhere on the forums, there is a link to a GREAT resource about how tectonic plate movement works. There is also a resource somewhere on steve jackson games website(might be the one I reference linked to above) about world building that includes how geology/vulcanology (and perhaps some other *ology stuff) works in relationship to building worlds from scratch.

su_liam
07-31-2008, 03:30 AM
Yeah. To complicate things further, plates don't only translate and rotate, they deform to some degree. More mountains eventually, or maybe, new baby plates.

RobA
07-31-2008, 09:04 AM
Obligatory Wikipedia references:
http://en.wikipedia.org/wiki/Fault_%28geology%29#Fault_types
http://en.wikipedia.org/wiki/Plate_tectonics

-Rob A>

Midgardsormr
07-31-2008, 09:38 PM
And once again: Expeditious Retreat Press (http://e23.sjgames.com/credits.html?t=publisher&n=Expeditious%20Retreat%20Press)

Look for A Magical Society: Guide to Mapping about 3/4 of the way down the page.

If you enjoy the sample chapter, please consider supporting Expeditious Retreat Press by buying an entire book. It'd be a shame for such a wonderful publisher to go under from lack of sales. I can recommend any of the Magical Society books, especially Western Europe and Ecology and Culture.

Arkkeeper
12-08-2008, 06:44 PM
Subscribed and watching closely. this reminded me that one of my Overlays needs to be Tectonic movement although I think I have that under (hold on)... Oh Yes! My geology and Structure Map Overlay...

Greason Wolfe
12-09-2008, 01:52 PM
A nice tutorial. It happens that I've experimented with the same process a few times in the past with moderate success.

Lately, however, I've been taking a slightly different approach. Instead of laying out the plates on my own, I've been letting something like FT Pro generate random land masses and oceans and then trying to work out sensible plates from those results. It's made things interesting so far and, in some instances, let me identify dozens upon dozens of potential plates both above and below sea level.

overlordchuck
03-12-2009, 08:19 AM
This tutorial is magnificent. Repped.

macbeth
03-13-2010, 11:06 AM
Great tutorial. I'm using it in a slightly different fashion as well. Instead of starting at the plates, I started at my super-continent, drew a few random lines for where the future boundaries would be, then drew the continents as they drifted apart. This lets me control the number of continents and the basic shape and gives me the location of my divergent boundaries and my plate movement directions. Then I'll use the knowledge here to find out where my convergent boundaries are and where my mountain ranges and island chains form.

The only thing I didn't see you address is how mountain ranges, when they meet the ocean, don't just stop. They continue on into the ocean and decrease in elevation and flatten out, and the peeks become first a peninsula, then islands in an island chain until the range eventually flattens out and decreases in elevation enough to fall completely below sea level. This is somewhat apparent in southwestern Argentina and northwestern Mexico.

Zavael
04-21-2010, 04:54 PM
ha, I didnt know that somebody will take this approach to create worlds too :)
but.. U wrote that U want it becouse of some facts rather than randomness. Its nice but as soon as u put the continents as u like, it started to take a path of total randomness

Instead of this (Im not forcing u to study the geology any deeper) u could start (like the earth did) with one big Pangea and then split it up as the plates are moving, so u are not controling the continent count, not form and not the size of them. And it does not need to study nothing more than u did before.

One another big plus of this is that u can set the climate on all the continents based on the initial position of the continent, of its moving and the final position, becasue it is not true, that at the Equatorial it has to be a desert (I know there are more things that affect climate like the Gulf stream, but for start it is enough)... just look at the world map and u see for example that the greenland is totaly frozen while skandinavia isnt and so on...
http://i22.tinypic.com/dbqvbm.jpg
this goes for the seas like mediterran or lakes too...

but u have all my respect for not just set the continents and hills as u want and u did nice research of how it is really done
if I finish my map, I give it somewhere here and all the progress from the start, maybe once ;)

Rhotherian
05-24-2010, 11:54 AM

I used it to draw my latest map (infact, you can see the plate boundaries showing through from the oposite side of the page). I'm really pleased with the way it came out. Thanks again!! ^^

snoopy
11-15-2011, 12:31 AM
Great tutorial. I'm using it in a slightly different fashion as well. Instead of starting at the plates, I started at my super-continent, drew a few random lines for where the future boundaries would be, then drew the continents as they drifted apart. This lets me control the number of continents and the basic shape and gives me the location of my divergent boundaries and my plate movement directions. Then I'll use the knowledge here to find out where my convergent boundaries are and where my mountain ranges and island chains form.

The only thing I didn't see you address is how mountain ranges, when they meet the ocean, don't just stop. They continue on into the ocean and decrease in elevation and flatten out, and the peeks become first a peninsula, then islands in an island chain until the range eventually flattens out and decreases in elevation enough to fall completely below sea level. This is somewhat apparent in southwestern Argentina and northwestern Mexico.
I'd like to see an example of this. Though I don't find myself minding inaccurate maps in novels, I find that when I consider to creating one I become obsessed with ensuring that the details are as accurate as humanly possible. I've tried to find videos on the change in shape and position of Earth's tectonic plates themselves, but they only ever show the movement of the continents. I'm not sure how to deal with the change in shape of the actual plates.

In the time of Pangaea, I imagine that there were more plates containing just large expanses of ocean opposite Pangaea. How did these disappear/shrink/change shape over time? Did they just get mostly or entirely subducted beneath other plates? The more I think about it, the more questions I have and the harder time I have beginning.

So I guess what it all boils down to is this: To anyone who has used the Pangaea/plate movement over time method, how did you determine how the plates would change shape and size as well as position? I hope this question is clear, although I am not sure that it is.

HBrown
02-09-2012, 07:19 PM
In the time of Pangaea, I imagine that there were more plates containing just large expanses of ocean opposite Pangaea. How did these disappear/shrink/change shape over time? Did they just get mostly or entirely subducted beneath other plates? The more I think about it, the more questions I have and the harder time I have beginning.

I am no geologist, but I would think that some of this is obvious. Where plates diverge, that mid-ocean ridge that forms acretes to the plates on either side. Where places converge, one is subducted and its edge is lost, as is the original shape. But the side that glides over is deformed too, but crumpling, in the process that pushes up those mountain chains. Transverse boundaries are not unaffected, either. Anyone who has flown over the relevant parts of California has seen how very straight the San Andreas fault line is (by natural standards, at least). If you live in the San Francisco area you can drive along a string of lakes on the peninsula, reservoirs for the city, that are linear and run in a straight line on top of the fault. Everything to the west of the fault is crustal material that has been shoved off the continental plate and onto the oceanic place.

That's the easy part. The harder part is deciding how to model it in your map. Since most of us don't have a supercomputer at our disposal, we'll have to wing it. Mountains pile up where plates collide, and the boundaries of the colliding plates erode. Unless we're trying to do a series of maps accurately portraying the last few hundred million years of our imaginary worlds, that's probably enough.

I like this approach and intend to try it, in that mythical day when I get some time. I think I would add a rotational component to the plates and perhaps even play with some simplified programmatic model to see what kind of terrain I can auto-generate. Given that I don't have that supercomputer, any terrain generated would be coarse, but might serve as a starting point for a final map.

Thanks for this thread! If someone would be kind enough to tell me how to rep something, I'd do it in a flash.

Mustak
04-17-2012, 10:28 AM
Very interesting.

I will use it to define my world map.

I can imagine the convection zones (http://en.wikipedia.org/wiki/Convection) (from top) to define the direction of the plates.