Geoscience Reference
In-Depth Information
In order to prevent the multicollinearity, values of explanatory variables were
modi
ed by creating an interaction variable (e.g. Population density).
This method was used to analyse not only the data by a different tool, but to also
analyse different development due to the political and therefore economical regime.
The political change was in 1989. The analyses were done separately for the
1970
2005. Comparison of individual time spans
shows if the political change in 1989 has had an in
1990, 1995
2005, and 2000
-
-
-
fl
uence on the dependence
between road traf
c and all indicators. The new political regime changed ownership
from the state one to a private one of many industrial non-industrial objects,
changed a system of planning, etc.
5 Results
5.1 Correlation Analysis
The data with the strongest in
c development in the cities
were determined from CCAI. From the point of view of frequency of extreme
values (positive and negative) of indicators, indicator N (Population growth) was a
parameter with the highest negative correlation for 50 % of the analysed cities.
Indicator K (Other areas of land use in the core area) was a parameter with the
highest positive correlation in 16 % of the analysed cities.
The
fl
uence to the road traf
first approach of the analysis uses a procedure of weighted sum. Six extreme
values of correlation (3 positive and 3 negative) are chosen for each city where each
of the three values of each indicator is weighted by values 1
3. The most important
indicators are selected by the sum of these values. By this approach, the highest
sum of the weighs has N indicator (population growth), which is followed by I
indicator (Residential area in the core area) and R indicator (Dust emissions from
large sources). The highest sum of the positive weigh has I indicator (Residential
area in the core area). The highest sum of the negative weighs has N indicator
(Population growth).
The second approach uses a sum of correlation coef
-
cient, and sum of negative
values of the correlation coef
cient.
This approach takes into account all values, so the indicators with lower contri-
butions to the
gured in the charts below.
It is shown (Fig. 2 ) that R indicator (Dust emissions from large sources) has the
highest absolute value of the CCAI correlation coef
nal value are also included. This approach is
cient from all processed cities
for which all indicators was known in more than two time spans. These values show
that Dust emissions (R indicator) and Residential area in the core city area
(I indicator) have the highest in
fl
uence on road traf
c intensity, however, their
in
uence can be both negative and positive (see Figs. 3 and 4 ).
Figure 3 presents the Residential area in the core area (indicator I) to have the
highest direct in
fl
fl
uence to the average road traf
c intensity. The positive in
fl
uence
does not occur in all cities.
Search WWH ::




Custom Search