Fractals and the landscape

[spudule]

Hi all,
I have a situation that is throughly irking me (I'm so irked right now). I'm studying Landscape Architecture as mentioned in my intro. There's very little number crunching (sadly), but there is one very important bit of number crunching to be done and I'm not happy with the method that has been introduced to me.

The situation is this. You are given a piece of topography, and you are told to house certain uses on this topography.
This could be houses, a road, you name it. So naturally this involves a bit of 'earth juggling'.

Method indicated to me:
Decide your uses for the site and the area that you will need for them.
Using surveyors tools you take cross sections of the site.
You then split up these cross sections into Euclidean forms and work out the area of your cross section by simply adding up these regular areas.
These 'slices' are then turned into 'blocks' by creating further euclidean forms by combining the cross sections.
The 'blocks' you have created are then treated essentially like Lego, you move one block out, you put it somewhere else.
Eventually you end up with a piece of topography that will suit the uses you have for it.

The big problem for me, is the margin for error, I looked at previous years work and seeing tonnes and tonnes of difference in the before and after volumes I thought, as far as mathematical solutions go, this is not ideal.
When I brought this up my lecturer, he said 'don't worry the engineers can handle that'. I was pretty bemused.
Using outmoded mathematics for a calculation you end up leaving to someone else is not cool, why bother charging a client for a job you're going to pay someone else to do!

 So I was wondering what thoughts you had on this. I've leafed through some old maths books and they all propose a euclidean solution. We can build lovely knobbly bits of landscape using fractal mathematics. So I'm hoping in my quest for perfection to find a fractal method for modelling the topography more accurately.
 
 

[kram1032]

afaik, nice structures where done with both hyperbolic (especially some special kinds of bridges) and fractal (especially 3D-Lindenmeyer systems) geometry...

if you can somehow combine certain physical laws like statics and regional climate with a quite complex extension of lindenmeyer systems you could, maybe, find some nice looking and at the same time highly functional architectures. Functional for both the owners and the environment.

Though, in the end it'll be hard...

Though don't forget: splitting things up in blocks, changing the blocks and splitting the whole set into new blocks is kind of a fractal anyway