Page 108 - NGTU_paper_withoutVideo

Modern Geomatics Technologies and Applications

3.2 Global Moran’s I evaluation
Prior to the implementation of hot spot analysis, Global Moran’s I analysis was applied (based on the feature locations
and attributed values) to reveal the spatial distribution of COVID-19 incidence rates across the world. As can be seen, Fig. 2
illustrates the results of spatial autocorrelation (Global Moran’s I) analysis, proving that the CIR values of COVID-19 were
clustered throughout the world at the time of this study. The Moran’s I value of 0.49 and the z-score of 14.48, derived from the
analysis, described that there was less than 1% likelihood that the clustered pattern could be the result of random chance.

Fig. 2. Spatial autocorrelation (Global Moran’s I) of COVID-19 incidence rates

3.3 Getis-Ord Gi* hot spot analysis evaluation
Getis-Ord Gi* hot spot analysis was used to find out which countries were located in hot spot regions and which of
them were low risk areas. This method had been applied to model other disease outbreaks [15, 27]. As is presented in Fig. 3,
the clusters of high values (hot spots) with a 99% of confidence levels were located in all parts of Europe and northern Africa.
Almost all regions in the Europe continent, investigated as the hot spot areas with highest rates of incidence. Three countries of
Africa, namely Libya, Algeria, and Tunisia, were also considered as high risk regions due to the high confidence level of 99%
obtained by the analysis. However, other countries of Africa were the least dangerous areas. A large number of African
countries, especially those situated in central and eastern parts of the continent were cold spots due to their 99% of confidence
5