Is there something like a depth buffer in matplotlib?

I think it's difficult to do such things in matplotlib, since it's a 2D plot library with some 3D plot ability. I recommend some real 3D plot library, such as visvis, mayavi, vpython. For example, in visvis, you can create a 3D curve like:

import numpy as np
import visvis as vv
app = vv.use()

f = vv.clf()
a = vv.cla()

angle = np.linspace(0, 6*np.pi, 1000)
x = np.sin(angle)
y = np.cos(angle)
z = angle / 6.0
vv.plot(x, y, z, lw=10)

angle += np.pi*2/3.0
x = np.sin(angle)
y = np.cos(angle)
z = angle / 6.0 - 0.5
vv.plot(x, y, z, lc ="r", lw=10)

app.Run()

mayavi:

import numpy as np
from mayavi import mlab

angle = np.linspace(0, 6*np.pi, 1000)
x = np.sin(angle)
y = np.cos(angle)
z = angle / 6.0

mlab.plot3d(x, y, z, color=(1,0,0), tube_radius=0.1)

angle += np.pi*2/3.0
x = np.sin(angle)
y = np.cos(angle)
z = angle / 6.0 - 0.5
mlab.plot3d(x, y, z, color=(0,0,1), tube_radius=0.1)

mlab.axes()
mlab.show()

You could try to use the fill_between function to color the regions manually perhaps. See the random walkers example:

http://matplotlib.sourceforge.net/users/recipes.html

http://matplotlib.sourceforge.net/users/recipes-6.py