Density of built

Objective of this request:

       Calculation of the density of built in Nantes..

       For that purpose, we are going to use a regular grid.

SQL Instruction :

        -- Creation of the study area (from the layer VEGETATION)

       create table zone as select ST_Envelope(ST_Union(the_geom)) as the_geom from VEGETATION;

       -- Creation of the grid with a step of 200m, on the basis of the layer "Zone"

       create table maillage as select STO_CreateGrid(200,200) from zone;

       -- We compute the intersection between the layer Bati_surface (filter on zone "REG_BAT") and cells from the grid
       create table gridIntersectWithBati as select ST_intersection(a.the_geom,b.the_geom) as the_geom, a.gid from maillage a, BATI_surface b where ST_IsValid(b.the_geom) and ST_intersects(a.the_geom,b.the_geom) and b."LAYER" like 'REG_BAT';

        -- We explode the multiploygon
       create table exploded as select STO_Explode() from gridIntersectWithBati;

       -- We preserve all polygon objects (surfaces)
       create table filterPointsAndLines as select * from exploded where ST_Dimension(the_geom) = 2;

       -- We merge all objects which are in the same cell (via the field "gid")
       create table unionGridBati as select ST_Union(the_geom) as the_geom,gid from filterPointsAndLines group by gid;

       -- We calculate the total area of builiding in each cells
       create table batiDensityPerCell as select ST_Area(the_geom) as area,gid from unionGridBati;

       -- We compute the density (on a scale of 0 to 1) of building with regard to the surface of the cell
       create table density as select a.the_geom, a.gid, b.area/ST_Area(a.the_geom) as density from maillage a,batiDensityPerCell b where a.gid=b.gid;

       -- OPTIONNAL

       -- We remove intermediate layers from the Geocatalog
       drop table gridIntersectWithBati;
       drop table exploded;
       drop table filterPointsAndLines;
       drop table unionGridBati;
       drop table batiDensityPerCell;

Result :

The result is a grid, with for every cell, a value of density of built. To have an interesting view, it is necessary to make a thematic analysis.

Example of characteristics : Interval classification / On the field "density" / Type of interval : Quantiles / Numbers of intervals : 7



From the new layer of density, select cells where the density is superior to 0,4

SQL Instruction :

create table density_sup_04 as select * from density where density>0.4;

Copyright - IRSTV - 2009 -