Page 115 - NGTU_paper_withoutVideo

Modern Geomatics Technologies and Applications

0.6
Average of Congestion 0.4
0.5

0.3
0.2
0.1
0

Hour

Average of Congestion within CCZ Average of Congestion within PCZ
Average of Congestion out of PCZ

Fig.2. trend of the average of congestion during the time of study on weekends

4.2. Results of Spatial and Spatio-Temporal Analyses

The results of the Moran’s I univariate local statistics test for the average of congestion on weekdays and weekends have
been illustrated in Fig.3 and Fig. 4. As can be seen, the positive and significant spatial autocorrelation among the average of
traffic congestion on weekdays and weekends can be investigated which indicates that the in general, the traffic congestion in
the study area follows a clustered and organized pattern.it is expected that the similar spatial factors might cause the occurrence
of traffic congestion and therefore, spatial traffic prediction models might improve the process of traffic modelling. The maps
make it possible to divide the study area into 4 spatial clusters as follows. The areas labelled as “High-High” refer to the areas
which have high levels of congestion and are surrounded by areas with the high weighted average of congestion as well. In return,
the areas with label “Low-Low” indicates areas which have low levels of congestion and are surrounded by areas with the low
weighted average of congestion.

Spatial Autocorrelation of the
average of congestion (Weekdays)

Fig.3. Spatial autocorrelation of average of traffic congestion of weekdays during the time of study

4