# Thread: Fleshing Out a Setting - River Police?

1. ## Fleshing Out a Setting - River Police?

Okay, so here I am, trying to map out a new setting for a story I may or may not ever get around to writing. So I go through making a heightmap and then exporting to Wilbur, then using it to make the rivers and such. But what I really have to ask is, how trustworthy is Wilbur's river generation?

In the image attached, I wish to draw you to the massive super-river in the center of the continent. It reminds me of the Mississippi, but the problem is, this is a significant continent, and I just wonder if this is at all realistic. In other words, I ask the River Police if it is realistic to have a massive super-river winding through vast plains (or possibly desert; haven't really worked out climates yet) like this?

2. Well... climate is important, and Wilbur doesn't care about that.

Also, because of the way the flow accumulation algorithm works, low-gradient areas tend to make laser-straight stream courses. A real river would meander, but there's none of the dynamic feedback effects in the Wilbur model.

Also, the straight flow accumulation model often creates a lot of long parallel stream segments. In reality a main stream segment would tend to pirate together a lot of those. Deposition effects can produce yazoo tributaries, which parallel the main stream for awhile, but not usually great rafts of them.

For the most part, Wilbur will give you a good sense of where streams will lie given the terrain, but you have to give the terrain a bit of a grain of salt. Erosion is only of limited help: incise flow uses the same flow accum algo and can sometimes compound the areas, especially on flat areas. Doing a few long and short -distance precipiton runs and some small-amplitude magnitude noise can make it look better. I'm trying to perfect a method of selecting flat areas and digging a meandering seed trench, but it's slow going. Definitely, use the Wilbur rivers, but be ready to play with them.

One thing I do is to look at the river tex between noising and incise flow. Use the noise brush where the rivers look too straight, add another bout of absolute magnitude noise and fill basins. Look at the river tex again. If it looks good, apply incise flow. Otherwise rinse and repeat. This can be really slow, but probably a good deal faster than a dynamic river model with feedbacks.

3. Another thing is to save a white-rivers-on-black-background copy of the river texture and blur it a bit in ps or gimp. I'll have more to say on that later, but for now good night.

4. Wilbur's flow computations are 100% topologically accurate. The basin fill operation causes the surface to be connected without pits to the edges of the maps and the texture operation that "finds rivers" drops a bit of "water" on every pixel, runs it "downhill" until it finds a low spot and then lets you select a threshold on the total "flow" to generate the final results.

Making the surface connected is what causes those funny straight lines because they fill in big basins or flat areas at the edges of the raster. The flow then goes straight across the flat planes generated during the fill basins operations.

How you interpret those results is up to you, though. Wilbur doesn't know anything about rainfall or temperature so those items aren't included in the result. Wilbur therefore assumes uniform temperature and rainfall across the entire terrain. Those assumptions are not correct in any "real" sense but are reasonably accurate for a relatively local scale.

So is your map "realistic"? Certainly, within the bounds of Wilbur's model. Is it reasonable to have a river that large? The Amazon basin is certainly very large but it's not 3/4 enclosed by very high mountain ranges (it has one high mountain range on a single side on about 1/4 sides). If the basin of interest is equatorial and catches incoming wet winds then it might be plausible. If the winds hit the back side of the mountains then glaciers might well feed significant rivers flowing into a desert basin (see the Taklimakan basin here on earth), but the outflow channel would be more likely a potentiality than an actuality.

5. From a River Police standpoint it looks accurate in a technical sense. From an aesthetics standpoint it looks awfully straight. What I would do is to take what Wilbur gives me and use that as a reference point and draw my own by hand. That way I can insure that there is plenty of meander and less parallel rivers. Depending on scale, when you go to apply a river basin you might be overwhelmed with darker green basins and not enough color variation for plains or other things.

6. Okay, thanks for all the help, guys. Immensely detailed explanations are always good.

From a River Police standpoint it looks accurate in a technical sense. From an aesthetics standpoint it looks awfully straight. What I would do is to take what Wilbur gives me and use that as a reference point and draw my own by hand. That way I can insure that there is plenty of meander and less parallel rivers. Depending on scale, when you go to apply a river basin you might be overwhelmed with darker green basins and not enough color variation for plains or other things.
That was my plan, I just wondered how well I could trust Wilbur as a reference.

Again, thanks to Su_Liam, Waldronate, and Ascension for the help. Now I've got to figure out climates. After that, I'll try and churn out an actual map of this continent.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•