Page 26 - NGTU_paper_withoutVideo

Modern Geomatics Technologies and Applications

Geospatial analysis and modeling of COVID-19 incidence rates in Iran

1*
2
Nima Kianfar , Aliasghar Azma , Mohammad Saadi Mesgari 3

Department of GIS, Faculty of Geodesy and Geomatics, K. N. Toosi University of Technology, Tehran, Iran
1

College of Architecture and Civil engineering, Beijing University of Technology, Beijing, China
2

3 Department of GIS, Faculty of Geodesy and Geomatics, K. N. Toosi University of Technology, Tehran, Iran

* nimakianfar@email.kntu.ac.ir

Abstract: The coronavirus pandemic (COVID-19) has become one of the most serious health crisis over the world within
a blink of an eye. The disease was originated from Wuhan, one of China’s provinces in late December. Iran’s first
infected case of COVID-19 was detected on February 19, 2020. Qom province was the epicenter of the disease, which
had the highest incidence rate among other provinces of Iran. In order to illustrate the spatial distribution of COVID-19
incidence rates, we applied Global Moran’s I. To determine the location and intensity of high-risk regions, we employed
Getis-Ord Gi* and Anselin Local Moran’s I hot spot analyses. Moreover, we compiled a variety of 10 environmental,
demographic, and socioeconomic factors as potential explanatory variables to investigate the spatial variability of
COVID-19 incidence rates in Iran. Besides, we implemented global ordinary least squares (OLS) and local
geographically weighted regression (GWR) methods to examine the spatial non-stationary relationships. Qom, Tehran,
and Alborz are the top three provinces regarding high values of COVID-19 incidence. The distribution of incidence
rates across Iran was spatially clustered. Regarding the results of hot spot analysis, five provinces, namely Qom, Tehran,
Alborz, Qazvin, and Markazi were detected in high-high clusters, which made them significantly High-risk regions.
Moreover, provinces located in the center of Iran were the hot spot areas due to their 99% of confidence levels. Two
most uncorrelated explanatory variables were identified to be used in both models, namely the percentage of people over
2
60 and the percentage of urban population. GWR model could explain higher variations, due to its higher adjusted R
and lower AICc, which demonstrated 7% improvement of the model compared to OLS. In conclusion, spatial statistical
information obtained from this modeling could provide general insights to authorities for further targeted policies.

KEYWORDS: Spatial analysis, COVID-19, Iran, incidence rate, OLS, GWR

1. Introduction
In the late December 2019, Wuhan Health Commission in China reported coronavirus to the World Health Organization
[1]. Globally, as of January 10, 2021, a total number of 88,383,771 infected cases and 1,919,126 associated deaths were reported
by WHO [2]. Moreover, COVID-19 transmission rate is way higher than MERS-CoV and SARS-CoV, which makes this virus
extremely dangerous [3, 4]. Iran was among the first countries to report a rapid increase in the number of infected cases and
associated deaths [5]. First COVID-19 infected cases were reported in Qom province on February 19, 2020, imported from
Wuhan, China.
Recent research around the world showed that factors such as population density [6-8], air quality [9], and population
flow [10, 11] could have contribution to the higher levels of COVID-19 incidence rates. Besides, Miller et al. applied descriptive
analysis to examine the spatial distribution of COVID-19 globally on March 17, 2020 [12]. The results proved that China, Italy,
Iran, and Spain experienced the highest prevalence of the disease. In this study, we applied spatial analysis methods, including
Global Moran’s I and Getis-Ord Gi* hot spot analyses to understand the spatial distribution of COVID-19 incidence rates and
specify high and low-risk regions of the disease, respectively. Moreover, a global (OLS) and local (GWR) regression methods
were implemented to identify how well these techniques can examine the distribution of disease incidence rates, based on several
potential explanatory variables.

1