Settlements and landscape: analysis techniques

1. A raster DEM of the studied territory. You can obtain an aspect map from it using the r.slope.aspect module.
2. A vector map of sites or buildings dating from a defined period.

Convert the vector map to raster with the command:

v.to.rast input=edifici output=edifici.mask use=val layer=1 value=1 rows=4096

Remember that this conversion is run following the region settings. You can view the region settings with:

g.region -p

Now the resulting raster map becomes our MASK:

r.mask input=edifici.mask maskcats=*

From now on, all subsequent operations will be limited to the cells defined in our MASK, corresponding to built areas or settlements.

If you need to remove the MASK, remember it a raster map like any other one:

g.remove rast=MASK

The simplest analysis we can run is d.histogram. These are the histograms of elevation and aspect for the whole of our study area. Aspect is calculated counterclockwise from east, thus 0 = 360 = E, 90 = N, 180 = W, 270 = S. See http://grass.itc.it/grass62/manuals/html62_user/r.slope.aspect.html for the manual page of r.slope.aspect.

Applying built.areas as MASK results in these histograms:

Aspect values are counted from 0 to 360 in degrees. So it would be better to plot aspect in a polar diagram instead of an histogram.

The d.polar module does exactly what we need. Its complete documentation is at http://grass.itc.it/grass62/manuals/html62_user/d.polar.html .

It's handy to directly save the diagram to an EPS file for subsequent operations (inclusion in a LaTeX file, conversion to other graphic formats with ImageMagick, …). Then you can use this command to get your diagram

d.polar map=aspect.5m@PERMANENT undef=0 eps=all.eps

These are the results:

This simple exploratory analysis leads us to understand that in the study area there's a strong preference for sites exposed S-SE, built mostly around 700 meters above sea level.