[WILBUR] Variable rainfall on Wilbur?

[rgcalsaverini]

I was going to PM waldronate, but in a moment of clarity I decided no to bother him, and instead hope that he will read this thread
The question is, is it possible to load into Wilbur a "rainfall map" so to speak, indicating the density of rain (on a grayscale map perhaps), so it can be used by the precipition-based erosion and river flow calculations?

If not, would you have any interest at all on implementing it? =)

[Jalyha]

I don't know if it's what you're looking for, but I found this post on other fora:

Wilbur has two erosion tools, a "precipiton" based system that simulates rainfall washing soil downhill, and an "incise flow" tool that calculates water flows for the whole map, and erodes rivers based on flow rates. When using the incise flow tool, be sure to use the "fill basins" tool, every single time. I like to run a few passes of the precipiton tool, fill basins, and then run a few passes of the incise tool, using gradually increasing gradients and decreasing widths. This creates river valleys that are wide near the mouth and narrower near the source.

If your rivers aren't going where you want them to, you can use Wilbur's hill and valley tools to get them to go where you like. On one map, for instance, I created a large, high plateau where all the water drained off the plateau on one side. This created a major river feeding into another major river over a HUGE waterfall. I will often switch back and forth between GIMP and Wilbur in order to use the different tools each provides. The curves and levels tools in GIMP are very useful for fine tuning your contours, using curves for instance you can create continental shelves, flat topped mesas, and sharp spiky peaks.

[rgcalsaverini]

Thanks Jalyha, that helps!
But it's not exactly what I was looking for. I't my understanding that the precipition erosion as implemented on Wilbur assumes that it rains the same everywhere (or so I understand) and I wanted to know if it's possible (or will be possible) to fiddle with that.

[waldronate]

Wilbur has no notion of rainfall as such. The precipiton erosion feature works with a portion of height differences rather than any sort of explicit fluid and flow. An agent is randomly dropped onto the surface and it heads downhill, processing the differences in height as it goes. The amount of change is controlled only by the relative heights of points; there is no conservation of mass or energy in the system.

Incise flow and river finding work by using a counting agent: an agent is dropped onto a surface and heads downhill, increasing the count at each spot it touches ("leaving a dropping behind" might be a more memorable way to view this). An agent is dropped onto the surface for every point and their aggregate behavior is similar to what would happen if there were fluid flow. Incise flow then transforms altitude and the count for each point into an amount to be removed. Again, there is no conservation of mass or energy in the system. Fractal Terrains uses a similar algorithm for its river finding, but it does allow for a change in the amount counted for each block (this is where FT's rainfall map comes into play). I have no intention to change this behavior any time soon (certainly not until I get layers implemented in Wilbur).

[rgcalsaverini]

Oh I get it now, thank you, I was curious. Layers? That's cool, I'll be waiting eagerly!

[arch-fiend]

even though i realize this is a bit of an old post. i'm replying with the hope of sparking continued conversation on the matter. while ive only been messing with wilbur for the last few days and i'm just a noob in general, i also realized rather quickly that both erosion filters used by wilbur tend to make broad assumptions about how much its "raining" over an area and that its raining consistently everywhere. while i'm still unsure just how incise flow works, precipiton acts sorta kinda like natural rain fall patters cuz water goes down. it came to my mind however that if you were to say have a basic idea of where rain should be happening over your map then you could make a selection based on that and apply precipiton erosion over those choice areas. for an idea on where rain should be occurring on your map i've been using this tutorial made by pixie, whom divides important areas of rainfall into 7 different category that likely overlap at times.

[waldronate]

As I stated previously, I'm probably not going to be doing anything much in this area until I get layers implemented in Wilbur (that's not going to happen anytime soon). The reason for this is that the erosion system would ideally have a layer for rainfall intensity, surface hardness, water areas, and sediment accumulation. I do understand the interest, but I don't even have enough time to code up that feature.