Spatial analysis of population density and its effects during the Covid-19 pandemic in Sanandaj, Iran

ABSTRACT Cities are densely populated centers that have struggled with many issues throughout history. Undoubtedly, the Covid-19 pandemic is the newest and one of the most critical challenges, which has caused many problems related to urban functions since its start in 2020. As the most vital factor in the spread and incidence of this virus relates to the contact with infected people, increased communication and face-to-face connections between people can probably increase the spread of the virus. This article seeks to answer the question whether population density and building density in urban areas affect the spread of Covid-19 and relating incidents. Using official statistics of Covid-19 patients from the beginning of its occurrence in Sanandaj, Iran (March 2020), to the end of 2020, its relationship with the two variables of population and residential density at the neighbourhood level was examined. The results show that the correlation between infection rate and population density per hectare as well as dwellings per hectare in the neighbourhoods is significant at the 0.01 level. This indicates that, with increasing population and residential density in the urban areas of Sanandaj, the incidence of Covid-19 has also increased.


Introduction
The Covid-19 pandemic is one of the most challenging diseases of recent centuries, affecting all countries and individuals at various levels. Transmission of the virus can occur in certain situations, particularly in indoor, crowded, and inadequately ventilated environments, where the infected person(s) spend longer times with others, such as in restaurants, at choir repetitions and performances, fitness classes, night parties, in offices and/or at places of worship (WHO 2020). Theoretically, population density increases a person's exposure to the infection, resulting in an increase in the number of reproductions of the virus (also called the R number by epidemiologists), leading to more outbreaks. Due to the rapid spread of the Covid-19 disease worldwide, social distance is the first factor that is advanced as an explanation of the transmissibility of pandemic viruses (Kadi and Khelfaoui 2020).
The Covid-19 crisis has caused irreparable harm, particularly in the cities of developing countries. The incidence and mortality rate may be lower than in developed countries for various reasons, but the problems and consequences are very tangible. Numerous factors and strategies have been proposed by international and specialized institutions, including the World Health Organization, to reduce transmission and infection rates. These include social distance, quarantine, and personal isolation at home, all of which can be effective in reducing the rate of disease transmission. Research and recommendations of experts and competent institutions have identified handwashing, maintaining social distance, and quarantine of infected people at home, as well as the use of masks as the most effective ways to reduce the transmission of the Covid-19 disease. However, other factors can also contribute in different ways. As cities became centers of wealth, finance, and activities, people were pulled into urban areas since the advent of industrialization (Eltarabily and Elghezanwy 2020). This resulted in the concentration of populations and due to the density on this scale, cities have always been vulnerable to the initiation and spread of pandemic diseases. The currently ongoing Covid-19 epidemic can be treated as an urban incident which causes challenges throughout the world (Liu 2020).
In terms of urban planning, spatial aspects of density and compactness are among the factors that can play a direct as well as an indirect role in the spread of the virus, the ways in which it is transmitted, and in the increase in the number of patients. Therefore, defining the extent of the relationship between spatial factors and the transmission of the Covid-19 disease can help policymakers to consider better and more conservative urban development patterns for future prospects of the city. Experimental observations show that the extent of epidemics generally is directly related to the initial density of susceptible individuals (those who did not develop immunity) and that it increases its severity (nonlinearly) in the course of time (Li, Richmond, and Roehner 2018). The number of new infections is also strongly correlated with the number of contacts of susceptible individuals, with significant differences in epidemic size observed among populations with different densities (Maybery 1999). Recent studies show that high population density in cities, as it reduces the distance between people (Acuto 2020), contributed to the current prevalence of and mortality rate related to the Corona virus (Coşkun, Yıldırım, and Gündüz 2021;Hong et al. 2021;Kadi and Khelfaoui 2020;Lee et al. 2021;Nakada and Urban 2020;Wheaton and Kinsella Thompson 2020). Population density and spatial compactness can thus be influential factors causing increases in epidemics. This means that densely populated cities are far more vulnerable to the Covid-19 pandemic. Therefore, it may be necessary to consider specific criteria related to the risk of infectious diseases in the development and design of cities along with common socio-spatial criteria (Gandy 1999).
However, the analysis of a study on the relationship between density, the incidence of Covid-19, and mortality rates in 913 US metropolises has shown different results. In denser cities, the death toll from the Covid-19 was lower, but conversely, in less densely populated cities, the mortality rate was higher. As the study indicates, this was more likely due to better access to healthcare services and the "likelihood of greater adherence to social distancing advisories or orders in compact counties" . According to Littman, residential crowding (the number of people per square unit of interior space) within the city is more closely associated with Covid-19 than population density (the number of people per surface area). For him, the high number of confirmed cases in metropolitan areas such as New York, Chicago and Seattle stands in closer relation to their global connections than to their density (Litman 2020). These cities are major centers of travel, trade, tourism, and migration. It means that "connectivity matters more than density in the spread of the Covid-19 pandemic" (Hamidi, Sabouri, and Ewing 2020, 495). The results of Teller's research also confirm these findings (Teller 2021). Some studies placed emphasis on slums and found that the highest density in such settlements is an important factor in increasing the transmission of the disease due to the lack of healthcare services and due to non-compliance with health protocols (Baker et al., 2020;Durizzo et al. 2021;Sahasranaman and Jensen 2021). However, the results of some studies also indicate that density is not the only factor in increasing the incidence of the disease (Hamidi and Zandiatashbar 2021;Khavarian-Garmsir, Sharifi, and Moradpour 2021). Rather, the large number of commercial activities and the density of transportation infrastructure (Li et al. 2021) as well as demographic and socio-economic characteristics of urban population may play a more effective role in increasing the incidence of the disease (Gargiulo et al. 2020;Li et al. 2021). Furthermore, through lockdown and stay-athome policies, the role of population density in the spread of the disease will diminish (Sun et al. 2020). One study demonstrated that the correlation between density and infection levels is opaque; contrary to expectations, it was observed that the denser cities were also wealthier, enabling them to channel considerable resources to respond to the pandemic -and so reduce their infection rates (UN-Habitat 2021).
Nevertheless, population density seems to be an essential factor in the Covid-19 epidemic in cities. The results of a study on American cities show that the Covid-19 pandemic has led to a further decline in housing demand in densely populated neighborhoods (Liu and Su 2020). Although it is too early to conclude whether reconfiguration of housing and the environment can be effective in controlling this pandemic (Webster 2021), there has been debates about the need to reconsider the nature of urban spaces (Martínez and Short 2021) and the appropriate population density for cities (Desai 2020).
According to the above background, it can be concluded that population density resulting from the spatial compactness of urban forms has caused different responses depending on each urban context. Therefore, we cannot set a general rule. It is necessary to examine each urban context according to its own conditions regarding its effects on the transmission of and mortality rate due to the Covid-19 pandemic. Accordingly, this study aims to analyze the pattern of morbidity and mortality in Sanandaj spatially by identifying their relationship with population density, specifically regarding the spatial compactness of the city's built-up areas. Our conceptual framework was based on the hypothesis that congestion leads to more exposure to the Covid-19 virus, and more exposure will increase the rate of infection, which in turn leads to an increase in the mortality rate.

Data and methods
The study area of this research is Sanandaj city in western Iran. This city is the capital of Kurdistan province, and most of its inhabitants are Kurds. Sanandaj located at the geographical coordinates of E46.999° and N35.311° and its altitude varies from 1450 to 1540 meters. According to the official census of 2017, the city has a population of 412,767 people. Its total area comprises about 4888 hectares, of which 4282 hectares constitute the central area of the city, and another 606 hectares belong to four separate urban areas.
The data used for the analysis include spatial, demographic, and other statistical data relating to Covid-19 patients. Maps and spatial data are among the secondary data received from the Sanandaj Housing and Urban Development Department based on the latest updates. Demographic data was also obtained from the Statistics Center and data on the number of patients with the Covid-19 disease was received from the Iranian Ministry of Health and Medical Education.

Data preparation
Since the primary purpose of this study is to investigate the relationship between population density regarding the spatial compactness of the city's built-up areas and the prevalence of the Covid-19 disease, and given the importance of the role of social distance and population congestion in the epidemic, this article attempts to test the existence or non-existence of such a relationship. For analytical purposes, the data was first prepared and the required information was generated from there. The official spatial boundaries of the city's neighbourhood units, including 68 neighbourhoods and four separate urban areas, form the basis of the analysis. However, because most neighbourhood units include vacant and undeveloped land, a new area is drawn for each neighbourhood to increase the accuracy of the results, including only built-up areas ( Figure 1). Furthermore, due to the lack of sufficient data for four separate urban sections, these areas were ignored, and only the main area of Sanandaj considered for the study. After the removal of detached urban areas and vacant lands, the total area of the city of Sanandaj reached 1905 hectares. The built-up area of each neighborhood was specified and listed separately. Based on the obtained area for each neighborhood, population density per hectare and housing units per hectare was also calculated. The data concerning the Covid-19 patients includes official data recorded by the Iranian government, which was provided to the research team in an Excel file with fields including gender, age, occupation, address, and underlying diseases of individuals. However, only the frequency and spatial distribution of patients were used in this study.
The file concerning Covid-19 patients contains 8048 cases that were recorded since the beginning of the official Covid-19 tests in Sanandaj, in the period from March 2020 to the end of 2020. Accordingly, the data has been recorded based on the addresses in the file. It was necessary to convert these addresses into spatial statistics by specifying them on the map. In a timeconsuming and accurate process, all records were classified, based on their location addresses. The share of each of the 68 neighborhoods determined -and its quantity added to -the relevant attribute table. Out of the total number of records, 5922 people lived in the main area of Sanandaj city and 2126 people lived in four separate areas or outside Sanandaj city limits, which caused them to be removed from the list. After the data screening, we calculated the number of Covid-19 patients in each neighborhood per 1000 people and the number of patients per hectare of the built area as well as the further data required for completion of the analysis (Table 1).

Data analysis
Due to the type of data and the nature of the problem, using two spatial analysis methods, ArcGIS 10.8 and IBM SPSS Statics 25 software packages were applied. First, the obtained quantities were plotted on the map using ArcGIS software and then the correlation coefficient between the indices were calculated, using the Pearson correlation model in the SPSS software.

Population distribution pattern in the built-up area
The spatial structure of the city influences the pattern of population distribution in Sanandaj. The city has a relatively monocentric spatial structure with a primary central core, bordered by northern and southern parts. The southern part includes mainly modern neighborhoods and suburbs that have developed over the past three decades. Several settlements have furthermore rapidly developed around the city during the last decade; three have been annexed to the city limits and are considered as separate urban areas of Sanandaj. As mentioned, these separate urban areas were excluded from our study due to the explained reasons. As shown in Figure 2, the population density is higher in the northern parts of the city. Only two neighborhoods in the southern part of the city have a density of over 200 people per hectare due to the student dormitories in one neighborhood and high-rise apartments in the other neighborhood.
In addition to the population density per hectare, two further indicators, namely dwellings per hectare and family size, can also reveal a population distribution pattern. Figure 3 shows the frequencies of these two indicators on the map. The density of dwellings per hectare is similar to the population distribution pattern, but the family size in the neighborhoods is relatively uniform. The family size in 59 of the 68 neighborhoods is between 3 and 3.5. In two neighborhoods, it is less than 3, and in 7 neighborhoods it is more than 3.5, mainly due to the presence of student and military dormitories as well as prisons in these urban districts.

Population density and the Covid-19 pandemic
The relationship between population distribution patterns and density in the built-up areas on the one hand and the rate of the spread of the the Covid-19 pandemic on the other was studied according to spatial and Pearson correlation coefficient analysis. The spatial analysis results show that there is not much correlation between the incidence rate concerning the population and the incidence rate per hectare of the built-up area of the neighborhoods. Regarding the incidence rate per hectare, this rate is less than 1 in 6 neighborhoods and in 21 neighborhoods it is greater than 4 ( Figure 4).
The Pearson correlation analysis results indicate a relationship of the Covid-19 incidence rate per hectare with three variables: family size, population density per hectare, and dwellings per hectare. This relationship is positive for the variables of population density and dwellings per hectare. According to the values obtained, the association is significant at the level of 99% and the intensity of the relationship is relatively high. Furthermore, the relationship between the Covid-19 incidence rate per hectare and the variable of family size is significant at the level of 95% and the direction of the relationship is inverse. This means that with the increase in the family size, not only has the incidence of this disease not increased per unit area, but it has decreased ( Table 2).
The results concerning the relationship between the Covid-19 incidence rate per 1000 people with the three variables of family size, population density per hectare, and the number of dwellings per hectare, are different. According to the findings, the number of the Covid-19 incidence rate per 1000 people has no significant relation to family size, but at the level of 99% confidence, this variable related to the two variables of population density and the number of dwellings per hectare. It implies that in the neighborhoods where the population density and the number of homes per hectare are higher, the disease incidence has not necessarily increased but even decreased (Table 3).

Discussion and findings
Epidemic disease outbreaks generally increase through people's close interactions and through increased population density and connectivity. Cities are centers of high population density. The latter generally also depends on the density of buildings and the density of housing units. The World Health Organization and the results of many studies emphasize that, as the population density decreases and social distance increases, the pandemic of the Covid-19 will also decrease. However, cities themselves are inherently centers of population concentration and consequently, maintaining social distancing will hardly be possible in the long run. The Covid-19 pandemic has sparked discussions about rethinking the ideal form and structure of cities, and many have sought to provide a model for post-Corona urban life. New paradigms such as new urbanism and the smart city encourage congestion, while many believe that overcrowding is a factor that puts cities at risk in the face of pandemics such as the Covid-19 crisis. Many cities in developing countries are struggling with rapid population growth and population density. Sanandaj is one of the cities in Iran that has experienced rapid population growth over the past few decades and has many economic, physical, and social problems. On the one hand, natural factors, including topographical features, have limited the city's physical development, but on the other hand, poverty and economic problems have led to the development of slums in many parts of the city,   especially in the north and northeast. These areas have the highest population and building density. The analysis demonstrated that in Sanandaj, as the population density per hectare and the density of housing units per hectare increase, the incidence of Covid-19 per hectare also increases. On the other hand, the study of the relation between the incidence per 1000 people and the two variables of population density and housing unit density shows that a large population and the number of housing units do not increase the incidence. However, with increasing density of these two indicators per unit area, the incidence rate does rise. This indicates that the neighborhoods' building density and physical density contributed to the epidemic of the Covid-19 virus.
Another noteworthy point is that the neighborhoods with the highest incidence rates per hectare are primarily residential and are far from the city's central business district. More importantly, although restrictions were imposed on specific city activities, commercial and administrative activities never went into lockdown. The neighborhoods harboring these activities mainly include the suburbs and slums to the north and northeast of the city and the city's southern neighborhoods, containing residential complexes and high-rise buildings. However, these two parts of the city are physically different from each other. The buildings in the northern neighborhoods are often one or two stories high, but the narrow and organically patterned streets have led to a high population and housing unit density. In the city's southern neighborhoods, which are part of relatively new developments, the streets are wide, and there are many open spaces. Yet, the housing units are mainly in the form of residential complexes and high-rise apartments. In fact, despite the restrictions and the need to follow the protocols, it can be said that in the neighborhoods with higher population and building density, the risk of Covid-19 nevertheless remains higher.

Conclusion
Regarding the global Covid-19 pandemic, many studies have been conducted in various fields to identify this disease's consequences in different dimensions. In urban planning, various studies likewise attempted to determine the city's interrelationship and its functions in relation to the Covid-19 virus. This study has attempted to answer whether or not higher density in cities increases the incidence of Covid-19. Based on the findings, the answer is "yes" for Sanandaj. However, many factors play a role in the urban spread of the virus, and our conclusions therefore are not definitive yet.
Moreover, we only have tried to approach the issue from an urban planning perspective and have looked at the challenge through a different lens. This research was done using official Iranian data. As was mentioned, the built-up area of the neighborhoods was used in the analysis to achieve the most accurate results possible. Furthermore, we only used the two variables of population density per hectare and housing unit density per hectare in the analysis. The analysis did not include physical, economic, and social differences between neighborhoods. Therefore, other studies should be performed to complete this research by considering these factors as well as indicators such as social interactions, behavioral patterns, adherence to protocols, economic status, and level of employment.

Disclosure statement
No potential conflict of interest was reported by the author(s).