Hi Norbert,

I'll intersperse answers below ...

"Norbert Kern" <nor### [at] t-onlinede> wrote:
> Interesting approach!
>
> How you implemented self thinning? I didn't implemented it yet.

This is just killing plants which are encroached upon too much. The function
that I use for encroachment is quite arbitrary - see if each local plant
overlaps the plant under test and has higher competitive ability, if so
then add a value related to the overlap. Each plant species has an
encroachment survival value, so I use this, the calculated encroachment and
a random to see if the plant dies. I'm just trying a new test condition
which seems to give a better thinning ...

> How large was the number of plants at the start?

That image had 500 initial seeds. I ran one over the weekend on a larger
patch (15 by 15 units, max plant size 0.3 units) for thirty iterations that
started with 5000 seeds and ended (twelve hours parsing later) with 55000
plants. I'll attach the image.

> How did you define suniness?

Easiest to paste the code:

#declare fBase = function {
  pigment {
    bozo
    colour_map{[0.2 rgb 0][0.7 rgb 1]}
    scale 2
    warp { turbulence 0.15 }
    warp { turbulence 0.15 }
    scale 3
  }

}

#declare Lor = function {1/(1+4*pow((x-y)/z,2))}

#declare Logistic = function (_A) {1/(1+exp(-_A))}

#declare fTerrain = function { fBase(x,y,0).x }

#declare fWet     = function { 1-fTerrain(x,y,z) }

#declare fSun     = function {
  Logistic(
    10*sum(
      I,
      0,
      5,
      cos(I*pi/3)*fTerrain(x+0.1*sin(I*pi/3),y+0.1*cos(I*pi/3),z)
    )
    /(0.1*sum(I,0,5,abs(cos(I*pi/3))))
  )
}

#declare fViability = function (x,y,z,SW,SS,SA)
  { Lor(SW,fWet(x,y,z),SA)*Lor(SS,fSun(x,y,z),SA) }

So, fSun()is function giving 'south'-facing slope value.

In the fViability() function SW is the species 'wetness' preference, SS the
'suniness' preference and SA the species environmental adaptability. This
viability value is randomised slightly for each individual plant.

>
> Norbert Kern

How is your system progressing?

Bye for now,

Mike Andrews.

Post a reply to this message

Attachments:
Download 'ecosystem2.jpg' (207 KB)

Preview of image 'ecosystem2.jpg'