Unnamed Tectonic Start

[kirkspencer]

The original map of this world (not the first in this thread) was a supercontinent breakup done in g.plates. The only reason for the latitudinal lines now is as guides for the current (air and water) reversals. If you've not used g.plates, it's using a globular map and I highly recommend it (as it's been recommended to me) for tectonics. If you're wanting to get some idea of the globe's appearance, I recommend gprojector.

Yes, doing an imaginary world and ignoring tectonics is easier. I've build a few worlds with my imagination and poor artistry. I've learned a lot more and gotten much better each time I return to this site and try something harder. The first time was a beginner's grasp of gimp and inkscape. Last time it was a much better mastery of climate. This time I aim to make it tectonic plate movement. And I look forward to the serendipitous other things I'll learn.

[kirkspencer]

Well that's annoying. Running this in Wilbur will have to wait till I can get to my windows box.

Which means i've had time to look at my map and ponder things, and in turn ask a few questions both for education and a better map.

Points of clarification, first. I remembered that those are plates, not continents.

With that in mind, am I correct in assuming that where these plates are running across oceanic plates that island arcs would develop? Oh, dear that needs expansion. Assuming the plate has 100 miles or more of continental shelf leading. Because I know that where the continental shelf is pretty close to the subduction zone we get mountains like the Andes.

second question/request for opinions

I've made a marker of island "dots" at the long continent to the northeast. I'm open to possibilities. What happened is that for a range of 30 million years or so we has this contact between Three and Two (plates of the supercontinent). As you can see by the first snapshot from g.plates, Three and Two experienced a rift along the long suture/border they shared, and Three began a steady march westward. (Transform fault through what had been a long, shared mountain range. I'm still pondering how I want to model that at this scale.) The first map actually shows where Two breaks free of the eastern mass (and the transform along the bulge of Fives is yet another zone of interest) and starts being pulled by the subduction zone created south of Two and east of Three.

2-3_initial.jpg

During the initial separation, Three was headed westward and was almost clear of Two when it started south. (Three is moving almost twice as fast as Two if that matters to anyone.) Without impact rules in place, the southwestern corner of Two and the eastern tip of Three co-locate. Neither is sufficiently blocked to kill either of the driving subduction zones.

The second image shows us the point of maximum overlap - again noting this is two graphic overlays so there's no impact consequences built in. The thing I want to point out here is that for the remainder of the passage Three's rotation east is just fast enough that Two doesn't overlap any more. Instead there's repeated transform and short approaches (convergences) until Three is completely clear of Two's pathway.

2-3_max_overlap.jpg

So I've got either a convergence/transform boundary that eventually becomes a transform boundary, OR I can choose this as my uncommon (I get to do it once) continent/continent subduction (India plate into Asia forming himalayas). I like the latter, as Two gets these huge mountains as it heads into the antarctic zone (leading to the mountains holding up the sky, or whatever legend you choose). My problem is what does that tip of Three really look like after that? Eaten, of course, but am I going to get a chain of volcanic mountain/islands? If so, on which boundary? What happens to the transition border where Two did NOT eat Three? At base it's a transform boundary, but it's a transform perpendicular to a subduction zone.

There are a lot of other places with fun puzzles like this, but I'll save them. The only thing I want to add is unrelated to the issue in actual tectonics, but I think is necessary to note for artistic choice reasons. That whole subduction zone has already crossed the south spin pole at final map time. Last picture is the mass that's so distorted by the equirectangular above. That point of impact is the almost-right angle to the bottom right, and the direction of movement (and so the subduction zone) is that flat edge along the 'bottom' as viewed.

2-3_final_southpole.jpg

[kirkspencer]

Notional plan of action, open to change based on advise and unexpected obstacles.

Basic continental drifting done.

I'm going to take several runs in Wilbur, moving mountains around (within reason) to see how and where rivers form. From this using personal taste I'll finalize mountain locations and where the continental plates are submerged. One particular expectation is the development of bays, by the way.

Separately I'm going to pick out oceanic plates. The obvious will be those formed by spreading rifts (even where interrupted by subduction zones pulling continents across them.) The primary reason for this is opportunities for smaller islands, and possibly japan-like and indonesian-like island systems.

In conjunction with the oceanic plates I'm going to see about either continental plate fracturing (changing 'home plate') or orogenic chains into the continental shelf. Basically I want the chance for some British Isles/Greenland type opportunities.

I'm also going to draft out the climates. The more precise details will undoubtedly tweak the result but I should be able to know what to expect for rainfall quantities in various areas. Which in turn will bring me back around to river development, but this time for a more final effect.

At this point I'll have been bouncing between g.plates, gimp, and wilbur. I'm still trying to decide which system I'll be using to make my final world map. I'm used to Gimp. But I'm going to be doing a lot of scaling and smaller maps, and it's a lot easier if the main map is vector-based. I have CC3 (not +, yet) and inkscape and have poked at both. Choosing one would give me impetus to learn it to at least gimp-levels of skill (advanced beginner). Once I've done that I'll do the long, tedious process of making the maps of the world, starting with a high-level world, then focusing first on where the players will be for further detailing.

Oh, and since I've now got some plaster balls (I've been playing with methods of that for just this purpose) I'll be trying to print gores. Or going to the projector and getting one of the display packages around to see if I can trace then paint directly on the globe - as case may be. That's longer range, though.

Comments, recommendations, and advise in this regard will be read. I may not follow them, but I will definitely read them.

[Charerg]

I'll try to comment on the tectonics a bit after I've read/pondered your post in a bit more detail.

For now, I might offer a small tip:

You can actually clip a raster image into pieces that correspond to the plates you've defined in GPlates. As an example, here's an image with the raster unclipped:

Unclipped Raster.PNG

By choosing whatever feature collection you've saved your plates into within the "reconstructed polygons" slot, you can clip the image (and these pieces will then follow the movements of the plates):

Clipped Raster.PNG

[Charerg]

Ok, so I promised some comments on the tectonics:

1) First off, the pieces that you modelled in GPlates should definitely be considered as "continents" rather than plates. In fact, they should be considered as "pieces of continental crust". The reason being that the tectonic plates themselves change shape constantly (new oceanic crust being created, continental crust being deformed, pieces being sutured into or fractured off the plate), and the shape of the continents (the parts of continental crust above sea level) depends on the sea level (which depends on the average age of extant oceanic crust, amount of sediment in ocean basins, the volume of glaciers...you get the point).

2) If you have a subduction zone, a volcanic arc would develop. The subduction zone itself would be expected to form at the edge of the continental shelf, since by definition, the continental shelf defines the boundary between continental and oceanic crust. That said, if the continental crust is shallow, you might have something like the Sunda Plate: a chain of large islands uplifted at the subduction zone, and a back-arc region of shallow sea/low elevation terrain. This is perhaps a more likely scenario for a relatively young subduction zone. In the case of an older subduction zone (which has experienced volcanism, erosion and uplift for tens of millions of years), it's more likely that the back-arc-region would have experienced sufficient uplift/volcanism/sediment fill to have elevations above sea level.

3) Regarding your model with 3 and 2 separating, you need to keep in mind (depending on how realistic you want your tectonic model to be) that the primary driving force of plate tectonics is slab pull, not slab suction (the latter is considered the weakest, actually).

What this means in practice, is that the usual model of continent-continent collision is as follows:

a) Plate A (with a leading edge of continental crust) begins subducting oceanic crust of Plate B, causing Plate B to be pulled towards Plate A at high speed (think Cretaceous India, about 15-20 cm/year, which holds the speed record of tectonic plates).
b) As the continents close, the collision may evolve into a continent-island arc collision (Australia-New Guinea).
c) As the collision matures, more and more fold belts form at weak points, causing further deformation, eventually evolving into a continent-continent collision (India-Eurasia, or Arabia-Eurasia).
d) Finally, plates A and B are sutured together, and the subduction zone likely "jumps", re-starting the process.

So, if there is/was convergent movement between Plates 2 and 3, this would necessitate that one or the other was being subducted, which would definitely pull the subducted plate towards the subduction zone due to slab pull. An alternative scenario would be a transform fault (something akin to the Alpine Fault in New Zealand, perhaps), with the plates sliding past each other. However, the important thing to remember is that something like the Himalayas is the latest stage of a long process, and wouldn't form in a situation where you have two plates diverging away from each other. If you want a "Himalaya case", you may wish to consider altering the shape of your supercontinent (it doesn't have to be a single monolithic block, you know, the Pangaea had a large ocean between Laurasia and Gondwana).

Hope that was somewhat helpful, and I look forward to the next stage in your project!

[kirkspencer]

Thank you, Charerg.

1) I'm fine having plates (well, submerged continental crust) from the start preceding several of these pieces of continent, but I can't fit it in my head for the ones that start as interior plates. Still, the knowledge that these are mostly 'continent not plate' is very useful.

2) The gplate run is 180 Ma, and some of the plates don't start moving till ~50Ma. I think I've a mix of young and middle aged, and shall ponder them a bit.

3) I've tried, and mostly managed, to make the priority for movement slab pull / ridge push / slab suction. Mostly by making sure the last only happens in conjunction with a justifiable hotspot-rift and in use more for 'turning' than for the actual rotation.

Still, with what you've said reinforcing other things I've been reading recently, I'm restoring that to a transform fault. I'll have to play with the model in gplates to see where it all comes out, but it might well pull that 'tongue' northward. And will definitely kill the mountains built by undercutting. Collision mountains will abide, of course. And looking at existing plates I /think/ I can still have a complex set of subduction-transform faults to make this happen. Just means I need to look hard at the places on Earth where it happens already.

A couple of small points. First, I'm assuming you meant 20 cm, not mm, per year, as that's what I've been reading in multiple sources. On the other hand I could be misreading, so making sure. Second, while I'm aiming for realistic only to a degree. Given my intent of using this as my game world in a few months I can't afford the trip not only down the rabbit hole but through the back warrens. Still, I find it easier if my baseline is as realistic as possible, so I'm not setting it completely aside.

Again, thank you, and I'll undoubtedly ask more after I do some more work on the map.

[Pixie]

Hey there again. I hope you're up for some reading. I saw your post the other day and been thinking about how/if/what to reply.
Something is definitely wrong with your model. It's basically the result of going too fast and not considering properly the implications of each small decision and movement. But I couldn't quite explain it any better that this - until Charerg's reply, which gave me an idea:

So, here's a bit of reasoning based on possibly one of your first bits of continent motion:

From your g.plates screen grab, we can gather the movement of plate 3.
kirkspencer1.jpg
It's moving west by northwest. From there, we can gather that a subduction zone formed there. In gray to black, I identified the possible historical evolution of that boundary. It's curious and odd that subduction happens along the same "line" for plates 1 and 3, but not for 2... and I'll get back to that. But for now, just looking at that, we get that an oceanic plate is moving eastward, reasonably fast and that there is a growing subduction trend in that area. We don't know where is the opposite boundary of that plate. It might be a mid-ocean ridge, or a piece of continental crust, if it goes uninterrupted to the other side of the globe. That opposite side is getting pulled towards the "1 to 3" coast.

I also drew this very raw explanation of the onset of a subduction zone and how that creates slab pull and slab suction.
kirkspencer2.jpg
I hope you see the problem in accommodating the previous "separate, random, bits of subduction" with how an actual subduction zone forms. If you try to imagine this picture from above, the cracks wil expand sideways as well, because if a piece of crust starts to sink, it will pull every piece around it if not separated by a fault (a crack).
One way to harmonize this is to make the subduction boundary initially starting only in that limit of continent 3, and then spreading northwards, eventually to continent 1 (and also southward, but I'll keep to the northern side on this post)

And there's two ways to go at this:
kirkspencer3a.jpgkirkspencer3b.jpg
Plan A is that cracks will form more or less in a north-south direction, and that a piece of ocean won't sink with the sinking slab. Initially this will be a lateral fault, then some underwater horsts and grabbens form and eventually the subduction boundary will expand through this area. You get an ocean-ocean boundary on the backyward of plate 2, which will form a back-arc basin. This might not connect to the coast of block #1, or connect in a strange way. Also, back-arc basins move about over time. Also, given the small length of of the subduction, plate 3 feels very small suction westward.

Plan B has it that the subduction will roughly follow the coast. The bigger difference is that, over time, the direction of the slab pull will be different and this affects the oceanic plate's motion. That motion is what matters the most about this subduction zone, and it's what you are ignoring. Plan B would form a mountainous/volcanic coast, all the way from #3 to #1. In this case, I think it is less likely that blocks #2 and #3 diverge that soon, as they are actually "fusing" on their northern limit.

... The point being - every small mechanism you introduce will have repercussions to all the plates around it. In turn, this might create, alter or dismantle other mechanisms that you had going.

EDIT: you just ninja'ed me with your reply to charerg. I see some of points are "pointless" as you already have "the knowledge". I hope you don't mind that I leave it as it is, kirkspencer, it will probably be educational reading for someone.

[Charerg]

A couple of small points. First, I'm assuming you meant 20 cm, not mm, per year, as that's what I've been reading in multiple sources. On the other hand I could be misreading, so making sure. Second, while I'm aiming for realistic only to a degree. Given my intent of using this as my game world in a few months I can't afford the trip not only down the rabbit hole but through the back warrens. Still, I find it easier if my baseline is as realistic as possible, so I'm not setting it completely aside.

Yes, it should be cm instead of mm, good catch .

[kirkspencer]

Hi Pixie. No, not the knowledge, just a better glimpse of its shadow.

Pondering your comments with Charerg's plus some of my discomfort means I may try version 3 soon. Or wildly readjust what I have.

So you're both aware, it was crisscrossing paths that had me bothered. I've found justification for some but it has felt heavy handed.

Needless to say your comments will be welcome.

[Pixie]

I'm glad you're happy to accommodate my ramblings (feel free not to, anytime...!)

About the crisscrossing, specifically. You should equally be worried about continents reaching a zone where another continent "passed", not only where they actually overlap.

A continent passing over the path of another continent and leaving their motions untouched makes no sense. Here's why:

#1 - Continent passing over the path is being pulled by "slab pull" effect. In this case, the crust ahead of it is already part of its plate. The previous continent couldn't have passed there without subducting that oceanic crust, which would effectively shut down the previous "slab pull", replacing it. Subduction margins don't move that fast, so the two continents would collide close to that area. The direction of their motions would be adjusting continually as the size and orientation of the subduction margin evolved.

or #2 - Continent passing over the path is being sucked by "slab suction" effect. In this case, its "front" is a subduction margin that is receding - this happens at a slower velocity, but it still happens. Once it reached the trail of the first passing continent, it meets oceanic crust which is part of its plate. That oceanic crust is likely to start being subducted as well. This will create a "slab pull" on that continent, forcing it back into that margin - what happens next depends on the geometry of the case.

So, with this in mind, here's how I think you can avoid this: you need to consider the oceanic crust being formed as the pieces break away. And keep a close eye on the subduction zones being formed. Adjust plate motion accordingly, every time subduction adds a new "pull" force to a runaway plate. That's why you don't get a huge number of continents on Earth-like tectonics environment, they keep pulling each other together.

(This, of course, is a much heavier workload for hobbyists world makers like ourselves, but just wait until you get to the climate.... )