Environmental Engineering Reference
In-Depth Information
Equation 3 is used for the transformation from a continuous scale into another
normalised continuous scale.
x Min
−
Xn
=
(3)
Max Min
−
Integration of the 12 Variables into the Model
The 12 normalised raster variables are integrated using GIS combine operations.
Each pixel of 1 ha of territory is assigned a vector with the 12 natural variables
valued. We obtain n vectors distributed among the 50 million 1-ha pixels in
Spain.
The next step is to order the n vectors using the modulus, to assign a synthetic
value of theoretical environmental quality. The vector modulus equation (Eq. 4) is
used to order the n vectors obtained based on their components.
2
2
2
vv v v i
= ++++
....
...
=
1...12
(4)
1
i
12
where
n
is the vector modulus; and
n
i is a vector component. Then the values
obtained for each of the n Euclidean distances are normalised into m equivalent
classes of theoretical territorial environmental quality required (from 1 to m),
applying the formula (5).
Theoretical natural quality class = ((n-n
MIN
) / (n
MAX
-n
MIN
)) * (m + 0.5)
(5)
where
n
is the vector modulus of each of the n vectors obtained;
n
MIN
is the
minimum vector modulus obtained;
n
MAX
is the maximum vector modulus
obtained. In our example, we use five environmental quality classes (m = 5),
therefore, the higher territorial environmental quality class is obtained when
4.5<
n
<5.5. The uses of 5 to 100 classes are considered sufficient to assist
the SEA and screening models of the two Spanish infrastructure plans: PIT
and PEIT (to distinguish clearly extreme cases of environmental quality in
territory).
Verification of the Model
The goodness of fit is checked against the real model of the five classes of envi-
ronmental quality by integrating our environmental quality map with a map of
protected natural areas in Spain (30% of Spanish territory). We apply a contin-
gency table with the corresponding chi squared test, which enables us to analyse
whether there is a statistically significant distribution of squares with greater
territorial environmental quality inside the protected areas.
