key: cord-0916795-s85sseqo authors: Gharakhanlou, Navid Mahdizadeh; Hooshangi, Navid title: Spatio-temporal simulation of the novel coronavirus (COVID-19) outbreak using the agent-based modeling approach (Case study: Urmia, Iran) date: 2020-07-30 journal: Inform Med Unlocked DOI: 10.1016/j.imu.2020.100403 sha: 6915903a6a893961ad5b8e16c0897e259394a771 doc_id: 916795 cord_uid: s85sseqo The ongoing outbreak of the COVID-19 as the current global concern threatens lives of many people around the world. COVID-19 is highly contagious so that it has infected more than 1,848,439 people until 14 April 2020 and killed more than 117,217 people. The main aim of this study is to develop an agent-based model (ABM) that simulates the spatio-temporal outbreak of COVID-19. The main innovation of this research is investigating the impacts of various strategies of school and educational center closures, heeding social distancing, and office closures on controlling the COVID-19 outbreak in Urmia city, Iran. In this research, the outbreak of COVID-19 disease was simulated with the help of ABM so that all agents considered in the ABM along with their attributes and behaviors as well as the environment of the ABM were described. Besides, the transmission of COVID-19 between human agents was simulated based on the SEIRD model, and finally, all control strategies applied in Urmia city along with corresponding actions of each control strategy were implemented in the ABM. The results of the ABM indicated that school and educational center closures in Urmia city, reduced the number of infected people by 4.96% each week on average and 49.61% in total from February 21 until May 10. Heeding social distancing by 30% and 70% of people of Urmia city from March 27, led to decrease the number of infected people by 5.24% and 10.07% each week, on average and 31.46% and 60.44% in total, respectively, and if 30% and 70% of civil servants of Urmia city did not go to work, the number of infected people would be decreased by 3.30% and 5.25% each week, on average and 32.98% and 52.48% in total from February 21 until May 10, respectively. As a result of this research, heeding social distancing by the majority of people is recommended for Urmia city in the current situation. The main advantages of disease modeling are to investigate how the disease is likely to evolve amongst the population of society and also assess the impacts of control strategies on controlling the outbreak of disease. Nowadays, although various precautions and control strategies have been taken throughout the 54 world, the ongoing outbreak of COVID-19 is the major challenge that many countries confront as 55 well as Iran (Bayham and Fenichel 2020; Escalera-Antezana et al. 2020). In that way, since the 56 identification of the first case of COVID-19 in Iran, numerous proceedings and strategies have been 57 implemented in order to control the COVID-19 outbreak that include: closing schools, universities, all 58 educational centers, cinemas, concerts, theaters, national competitions, and sports leagues, as well as 59 implementing the social distancing plan throughout the country. 60 Evaluations of the impacts of such control strategies as well as predictions of the future outbreak 61 of COVID-19 can play important roles in controlling the COVID-19 outbreak (Chatterjee et al. 2020 ; 62 Li et al. 2020 ). Traditional decision-making methods generally use the experience of experts to 63 estimate the efficiency of applied strategies. However, an assessment of COVID-19 with novel 64 approaches is essential to the global debate (Sarkodie and Owusu 2020). Although accurate evaluation 65 of the efficiency of strategies is determined over time, in modern management methods, experts' 66 opinions can be combined with new methods of modeling in order to have a better prediction of future 67 conditions and impacts of control strategies (Hooshangi and Alesheikh 2018) . Therefore, the 68 simulation of the COVID-19 outbreak as well as investigating the effects of diverse strategies on 69 controlling its outbreak have attracted the interests of many scientists and researchers recently. 70 It is obvious that the exact simulation of phenomena is impossible due to their complexities and 71 detail; therefore, in order to simulate phenomena, lots of simplifications must be considered in the 72 model ( ABMs can be used to find optimal strategies in large-scale incidents and to manage the situation. In 88 addition, the use of ABMs regarding their flexibility and bottom-up structure, allows decision-makers 89 to combine space and time (Hooshangi and Alesheikh 2017) . 90 There are three epidemiological modeling approaches including the ordinary differential equation 91 (ODE) model (Li et (Simoes 2012 ) and its effectiveness has been proven. 106 In the field of COVID-19 disease, most research has examined clinical and epidemiological 107 features of COVID-19, whereas spatio-temporal modeling of COVID-19 has been performed just in 108 limited research. Kang et al. (2020) described the spatio-temporal pattern and also measured the 109 spatial association of the early stages of the COVID-19 epidemic in the mainland, China. They 110 explored the spatial epidemic dynamics of COVID-19 by Moran's I spatial statistic with various 111 definitions of neighbors. They revealed that with the exception of medical-care-based connection 112 models, a significant spatial association of COVID-19 infections were indicated in most of the 113 models. Guliyev (2020) utilized spatial panel data models to investigate the propagation power and 114 effects of the COVID-19 disease, examine the factors affecting COVID-19 along with the spatial 115 effects, as well as determine the relationship among the variables including their spatial effects. • The environment of the ABM is made up of a variety of spatial data so that humans' 159 movements as well as interactions were performed in a similar way to reality. 160 • The number of human agents created as well as the process of distributing human agents 161 were performed with regard to the demographic condition of Urmia city. 162 • Human agents were discriminated based on their occupations as well as whether having 163 vehicles or using public transportations so that their movements and interactions were 164 affected by these parameters. 165 • The Susceptible-Exposed-Infected-Recovered-Dead (SEIRD) model was used to simulate 166 the transmission of COVID-19 among human agents. 167 • All strategies applied in the study area along with their exact dates were considered in the 168 The organization of this paper is as follows: materials and methods are described in section 2. In 170 that way, first, the study area is introduced, then, the spatial dataset used in this research are 171 described, and finally, the proposed ABM for modeling the outbreak of COVID-19 is explained in 172 detail. In section 3, the outputs of the ABM and also the evaluation of the ABM are presented. In 173 addition, applying three control strategies of school and educational center closures, heeding social 174 distancing, and office closures are completely explained and their impacts on controlling the COVID-175 19 outbreak are explicitly discussed, and finally, in section 4, the conclusions of this research are 176 presented. 177 must be correctly determined in the simulation. Therefore, in order to dynamics simulation of the 194 COVID-19 outbreak, the actual geographical data of the study area is required. The combination of 195 ABM with GIS allows agents to be situated in their actual geographic locations. In addition, this 196 integration improves the ABMs' ability to correctly simulate agents' behaviors, movements, as well 197 as interactions. 198 In this research, the spatial data utilized consist of residential areas, offices, business areas, 199 schools, roads, as well as the boundary of Urmia city (Fig. 2.A) , and the population density of 200 residential areas (Fig. 2 .B). These spatial data together provide the environment of the proposed ABM 201 in this study (see section 2.3.1). 202 to ABMs that involves a set of cells on which agents are situated (more detail about ABMs can be 215 found in Crooks and Heppenstall (2012) ). In the remainder of this section, the environment of the 216 proposed ABM will be explicitly described as well as the agents considered. 217 2.3.1. The environment of the ABM 218 As it is obvious, the outbreak of a contagious disease such as COVID-19 highly depends on the 219 location of the people as well as their movements. For this reason, the environment of the ABM 220 should encompass the spatial data concerned with people's residential places as well as the places 221 where people are likely to move. Therefore, the environment of the proposed ABM encompasses all 222 constructed areas related to either people's occupations or their living places. The environment of the 223 proposed ABM includes residential areas, schools, offices, and business areas. Besides, the outbreak 224 of contagious disease, particularly COVID-19 is so likely in the areas with a high number of 225 population; therefore, the environment of our ABM also involves the population density of residential 226 areas. Moreover, the movements of people are performed through the roads. All in all, the 227 environment of ABM involves the boundary of Urmia city, residential areas, schools, offices, business 228 areas, the population density of residential areas, and roads. 229 For cells of the ABM's environment, an attribute was defined with regard to each spatial data and 230 its value was correspondingly initialized according to the spatial data layers. In Regarding the issue of COVID-19 outbreak as well as the most important parameters affecting the 244 outbreak of the COVID-19, several attributes as well as behaviors were taken into account for human 245 agents that all of those were shown in Fig. 3 . In the remainder of this section, the process of human 246 agents' distribution, as well as their movements are completely described. 247 2.3.2.1. Distribution of human agents 248 In Urmia city, since the number of households along with the number of their members are known 249 (Table 1) , human agents were distributed as groups (households). In this model, it was assumed that 250 one house was assigned to each household. Therefore, first, house agents were created by the number 251 of households and situated randomly in the cells defined as residential areas; then, in each house, 252 human agents were created by the number of members in the household. It is possible to have a cell 253 with several houses located in it since the distribution of houses was performed randomly. On the 254 other hand, COVID-19 is so likely to be spread in places with a high number of the population rather 255 than a low number; therefore, it is better to distribute human agents, or rather, house agents according 256 to the population density patterns of the study area. To do so, house agents were situated randomly in 257 one of the five considered regions (see Fig. 2 .B) according to the roulette wheel selection method; 258 thus, house agents are so likely to be situated in regions with high population density. 259 It is obvious that people's movements are the cause of the COVID-19 outbreak. In this model, the 264 movements of human agents were considered in two modes. They move either by their personal 265 vehicles or public transportations. In this model, humans based on age were classified into four groups 266 (Table 2) : 5 and less, more than 5 to 24, more than 24 to 64, and more than 64. It was assumed that 267 humans who are more than 24 years old (two last groups) can have personal vehicles, of which 63% are not in any of the first three classes. In this ABM, it was assumed that students were assigned to 273 human agents who are more than 5 and less than 24 years old, and civil servants, as well as self-274 employed employees, were assigned to human agents who are more than 24 years old. Similarly, the 275 number of human agents in each group was determined according to the statistics of human resources 276 in Urmia city (Statistical Center of Iran 2017). 277 The selection of agents' workplaces was performed randomly among the office, business, or 280 school cells on the basis of agents' occupations. During the simulation, for each human agent, a cell 281 was assigned as the workplace and does not change. It should be noted that the same cell can be 282 several human agents' workplaces since the selection of the workplaces was done randomly. In this 283 model, all employees (civil servants and self-employed) and students move between their houses and 284 workplaces twice a day. To simulate the movements of human agents, one day was divided into four 285 time intervals: in the first time interval, employee agents (civil servants and self-employed) who have 286 personal vehicles move from their houses to their workplaces by their own vehicles and in this time 287 interval, there is not any risk of becoming infected for these human agents. On contrary, employee 288 agents who do not have personal vehicles as well as student agents in order to move from their houses 289 to their workplaces, use public transportations and in this case, they are likely to become exposed if 290 there is any exposed human agent on their way to work regarding the R0 1 value of the exposed human 291 agent as well as the COVID-19 transmission probability. It should be noted that the speed of public 292 transportations with regard to their stops in stations was considered 45 km an hour and according to 293 this speed, it was assumed that human agents move about 11.25 km every 15 minutes; therefore, the 294 movements of human agents are updated every 15 minutes and their health status (see section 2.3.3) 295 are investigated every 15 minutes based on the existence of exposed human agents in the same cells 296 with them. It should be noted that in order to transmit the COVID-19 indoors, people should be kept 297 at a distance of fewer than 2 meters for at least 15 minutes (Centers for Disease Control and 298 Prevention (CDC) 2019; World Health Organization (WHO) 2020); therefore, in this ABM, the time 299 interval was considered 15 minutes and the dimension of the cells was taken into account 2 meters. 300 Human agents move between their houses and workplaces through road cells. All employees as well 301 as students move to their own workplace cells in the second time interval and similarly, the 302 probability of becoming exposed is investigated at their workplace cells. In the next time interval (the 303 third time interval), employees and students come back from their workplaces to their houses in the 304 same way described in the first time interval and the risk of disease transmission is investigated for 305 employees who use public transportations as well as students; and in the last time interval, all 306 employees and students move to their houses and in this time interval, the disease transmission is 307 investigated among all members of the households. These processes are explicitly illustrated in Fig. 4 . 308 In our model, human agents at any moment of the simulation are in one of the susceptible, 312 exposed, infected, recovered, or dead states and their states change regarding the SEIRD model 313 (Korolev 2020; Piccolomini and Zama 2020; Shao and Shan 2020). The states of human agents 314 change under certain conditions over time. They will remain susceptible until they coincide with at 315 least one exposed human agent in the same cells. In this case, based on the transmission probability as 316 well as the R0 value of the exposed human agent, COVID-19 might be transmitted from an exposed 317 human to a susceptible human. If the human agent becomes exposed, it starts to transmit the disease 318 and this trend lasts 2-14 days (COVID-19 incubation period) (Centers for Disease Control and 319 Prevention (CDC) 2019). After passing the incubation period which is considered diverse for each 320 human agent, the state of exposed human turns into infected, and in this case, it will be quarantined. 321 In the infected state, human agents cannot move and consequently spread the disease because of the 322 quarantine. In the state of infected, one of two events regarding the age of the human agent happens: 323 first, the infected human agent becomes recovered after 2-4 weeks (World Health Organization 324 (WHO) 2019), or second, the infected human agent dies. In Fig. 5 In order to have an overview of the parameters of our ABM, input parameters of the model along 329 with their values as well as the values assigned to the attributes of human agents were summarized in 330 Table 3 . To initialize these values, previous studies and some of the most authoritative websites were 331 used. For the parameters and attributes whose values were defined in a range, a normal distribution 332 was considered. Moreover, for these parameters and attributes, the mean and standard deviation were 333 also reported. It should be noted that the normal distribution was simply considered for these 334 parameters and attributes due to the novelty of coronavirus as well as the lack of information on what 335 distribution it follows (as a suggestion for future research, an investigation can be made on how these 336 parameters (parameters defined in a range) are distributed; then, the values adjusted for these 337 parameters would be based on the study of their changes in reality.). 338 Qom and returned to Urmia on February 25; therefore, the beginning of our ABM was scheduled for 348 February 25. In addition, one human agent was randomly considered exposed at the beginning of the 349 ABM. Precede the outbreak of COVID-19 from Qom to all cities in Iran, the Iranian government in 350 order to reduce the rate of COVID-19 outbreak closed all schools and educational centers on February 351 21. For this reason, at the beginning of our ABM, it was assumed that all student agents do not move 352 in the ABM as well as about 30% of civil servant agents (those who are teachers in the real world). 353 Similar to the first detected case, the second infected case with COVID-19 disease in the Urmia city 354 was identified on March 2. This person also became infected due to traveling to Qom city and 355 returned to Urmia city on February 28; so, another susceptible human agent was randomly selected 356 among all human agents, and its health status was changed to exposed. In Iran, the New Year imposed the social distancing plan on the population. Regarding this event, only human agents who 361 did not pay attention to social distancing were likely to become exposed. The interval between March 362 31 and April 1 was the official holidays of the New Year and similarly, all human agents were 363 assumed to stay at their houses and there were not any human agents' movements in this interval in 364 the model. It should be noted that in Iran, one day a week is a holiday and these holidays were also 365 considered in the ABM according to the exact dates of them. Similar to other holidays, on these 366 holidays, it was assumed that human agents stay in their houses and do not move. Fig. 6 clearly 367 illustrates all that happened in the study area from the beginning of the epidemic to May 10 (the end 368 of the simulation). 369 It should be noted that in this research, although the values set for the parameters of the number of 386 population by sex, age, and occupation (student, civil-servant, self-employed), percentage of people 387 using public transportations, and percentage of people using personal vehicles according to the 388 characteristics of the study area (Urmia city), all these parameters' values were defined as changeable 389 so that users are able to change them regarding their purposes. In addition, in this ABM, three various 390 control strategies were designed that the parameters assigned to them are changeable, as well; 391 therefore, the proposed ABM is very flexible as well as interactive so that not only it is able to predict 392 the COVID-19 outbreak and consequently, help make decisions, but also the effects of the parameters 393 as well as the impacts of various control strategies can be examined by increasing and decreasing the 394 values of the parameters. In that way, the proposed ABM can be implemented for different scenarios, 395 and consequently, the COVID-19 outbreak and the impacts of control strategies can be investigated in 396 the study area for diverse conditions. them. In addition, there are always some stochastic events in the real world that are not considered in 426 models. The outbreak of COVID-19 is one of the complex phenomena so that diverse parameters 427 affect it. It is obvious that taking into account all parameters affecting is not feasible. On the other 428 hand, disease data always has a number of errors; for instance, people with the disease may not refer 429 to health centers and this case may not be recorded. It is also possible that the place of becoming 430 infected of people was recorded wrongly due to the lack of health facilities or other reasons; for 431 example, people might become infected in the city other than the city where the disease was recorded. 432 The actual observed data (daily infected cases) used in this research was the official data provided by who do not go to work as well as when all civil servants go to work are shown in Fig. 11 . 509 510 Fig. 11 . The average number of infected people in Urmia city for two different values of the percentage of civil 511 servant agents who do not go to work as well as when all civil servants go to work. The results of the model illustrated in Fig. 11 indicate that if instead of 30%, 70% of civil servants 513 did not go to work provided that they stayed in their houses as well as all events happened according 514 to Fig. 6 , until May 10, the total number of infected people in the study area will be reduced by 318.7 515 people (29.10%) as well as 31.87 people (2.91%) each week on average. In addition, if all civil 516 servants went to work, the number of people who will be infected by COVID-19 until May 10, was 517 approximately 1634, instead of 1095 people. Therefore, when the 30% and 70% of civil servants do 518 not go to work, the number of infected people is reduced by 53.89 (3.30%) and 85.76 (5.25%) people 519 a week and 538.91 (32.98%) and 857.61 (52.48%) people in total, respectively. 520 It should be mentioned that the start date as well as the time periods are not consistent in three 521 figures (Fig. 9, Fig. 10, and Fig. 11 ) since the starting dates of applying strategies were different in the 522 study area. In addition, in the last two strategies, the values considered for the percentage of people 523 heeding social distancing as well as the percentage of civil servants who do not go to work were 524 defined as subjectively, without any specific investigation, and only in order to evaluate the impacts 525 of strategies on controlling the outbreak of the COVID-19. In addition, the investigation of the 526 impacts of these strategies was performed only based on the viewpoint of the COVID-19 outbreak. 527 According to the results of the model in Fig. 9, Fig. 10 Elimination of the COVID-19 outbreak requires not moving of all members of society or a very high 534 proportion of the population. The strategy of heeding social distancing not only does not damage the 535 country from the economical viewpoint but also leads to an extreme reduction in the number of 536 infected people and consequently, a decrease in the spread speed of COVID-19 in a society. In the 537 current situation, the heeding social distancing strategy by the majority of people as a suggestion of 538 this research can lead to a remarkable reduction in the number of infected people and consequently, 539 the control of the COVID-19 outbreak in Urmia city. 540 4. Conclusions 541 At present, the epidemic of COVID-19 disease has emerged as the most important global health 542 challenge and its outbreak has reached 210 countries and territories. Despite many efforts made to 543 decrease the speed of COVID-19 outbreak around the world, it still spreads quickly and infects so 544 many people daily. This rapid and widespread outbreak of COVID-19 has caused serious social, 545 economic, cultural, and even political damage in the countries. 546 One of the most valuable aspects of simulation is the explanation of real-world phenomena that 547 simulations of those phenomena either are not feasible in the real world or are costly to be performed. 548 Modeling and simulation of the COVID-19 outbreak in a region as well as investigating the efficiency 549 of control strategies can assist health policymakers in controlling and preventing the COVID-19 550 outbreak. This is the main contribution of this research. 551 In this research, the outbreak of the COVID-19 was simulated in Urmia city with the help of an 552 agent-based model due to its capability in modeling people's movements as well as their interactions 553 that are two main causes of the COVID-19 outbreak. In addition, three control strategies of school and 554 educational center closures, heeding social distancing, and office closures were applied in the model 555 and the impacts of each one on decreasing the speed of COIVD-19 outbreak as well as preventing its 556 outbreak were investigated. 557 The results of the model indicated that school and educational center closures reduced the number 558 of infected people by 4.96% each week, on average, and 49.61% in total in Urmia city from February 559 21 until May 10. Heeding social distancing strategy by 30% and 70% of people of Urmia city from 560 March 27, led to decrease the number of infected people by 5.24% and 10.07% each week, on average 561 and 31.46% and 60.44% in total, respectively, and when 30% and 70% of civil servants of Urmia city 562 did not go to work, from February 21 until May 10, the number of infected people was decreased by 563 3.30% and 5.25% each week, on average and 32.98% and 52.48% in total, respectively. As a result of 564 this research, regarding the current situation of Urmia city, in order to slow down the speed of the 565 COVID-19 outbreak, heeding social distancing by the majority of people is recommended. The results 566 of this research can be helpful to health policymakers in selecting appropriate strategies to decrease 567 the outbreak of the COVID-19 in Urmia city. 568 The outbreak of a disease is a highly complex natural phenomenon, and a specific model cannot 569 be utilized for all regions of the world since not only diverse parameters affect the spread of the 570 disease but also these parameters vary from one place to another. People cultures, level of literacy and 571 awareness of people, the way people interact with each other, available public transportations, urban 572 context, population density, job diversity, variety in factors of age, gender, number of people 573 employed, number of students, number of people having personal vehicles, number of people using 574 public transportations, and etc. are among the parameters that make differences in the way COVID-19 575 spreads in different places. Therefore, it is suggested for future research that the outbreak of COVID-576 19 will be simulated for other regions of the world by taking into account various parameters as well 577 as applying diverse control strategies. 578 An individual-based model of hepatitis A transmission Impact of school closures for COVID-19 on the US health-care 590 workforce and net mortality: a modelling study Unknown Etiology in Wuhan, China: Potential for International Spread Via Commercial Air Travel A hybrid multi-scale model of COVID-19 transmission dynamics to 597 assess the potential of non-pharmaceutical interventions Modelling transmission 602 and control of the COVID-19 pandemic in Australia Healthcare impact of COVID-19 605 epidemic in India: A stochastic mathematical model Estimating the reproductive number and the outbreak size of Novel 608 Coronavirus disease (COVID-19) using mathematical model in Republic of Korea. Epidemiology and 609 Health COVID-19 Epidemiology Committee, Ministry of Health and Medical Education Introduction to agent-based modelling. Agent-based models of 615 geographical systems An agent-based model to evaluate the COVID-19 transmission risks in facilities. 618 Computers in Biology and Medicine How to restart? An agent-based simulation model 621 towards the definition of strategies for COVID-19" second phase Quantitative structure-activity relationship and 625 molecular docking revealed a potency of anti-hepatitis C virus drugs against human corona viruses Toward a 629 containment strategy for smallpox bioterror: an individual-based computational approach. Brookings 630 Institution, CSED Working Paper Clinical features of cases and a cluster of Coronavirus Disease A field indoor air measurement of SARS-CoV-2 639 in the patient rooms of the largest hospital in Iran. Science of The Total Environment Modeling exit strategies from 642 COVID-19 lockdown with a focus on antibody tests Developing an agent-based model for 645 simulating the dynamic spread of Plasmodium vivax malaria: A case study of Sarbaz Determining the spatial effects of COVID-19 using the spatial panel data model. 649 Spatial Statistics Agent-based task allocation under uncertainties in disaster 652 environments: An approach to interval uncertainty Developing an agent-based simulation system for post-656 earthquake operations in uncertainty conditions: a proposed method for collaboration among agents Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. The Lancet The continuing 2019-nCoV epidemic threat of novel coronaviruses 665 to global health-The latest 2019 novel coronavirus outbreak in Wuhan Mathematical modeling of the spread of 671 the coronavirus disease 2019 (COVID-19) considering its particular characteristics: The case of 672 China Universal masking is 675 urgent in the covid-19 pandemic: Seir and agent based models, empirical validation Spatial epidemic dynamics of the COVID-19 outbreak 679 in China Epidemiological modeling of bovine brucellosis in 682 IEEE International Conference on Big Data (Big Data) Covasim: an agent-based model of COVID-19 dynamics and 686 interventions Identification and Estimation of the SEIRD Epidemic Model for COVID-19 Fast and slow dynamics of malaria model with relapse. Mathematical 692 biosciences Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia Mathematical model of infection kinetics and its analysis for COVID-19, SARS and 699 MERS. Infection, Genetics and Evolution A conceptual model for the coronavirus disease 2019 (COVID-19) outbreak in Wuhan, China 703 with individual reaction and governmental action 2019 of Conference. DEVELOPING A CELLULAR 707 AUTOMATA MODEL FOR SIMULATING RAINFALL-RUNOFF PROCESS (CASE STUDY: 708 BABOL CATCHMENT). The International Archives of Photogrammetry, Remote Sensing and 709 Spatial Information Sciences Agent-based simulation tools in 712 computational epidemiology An agent-based approach for modeling dynamics of contagious 716 disease spread Preliminary analysis of COVID-19 spread in Italy with an adaptive 719 SEIRD model The effect of control strategies to reduce social mixing on 723 outcomes of the COVID-19 epidemic in Wuhan, China: a modelling study Cellular automata modelling of SEIRS A spatially explicit agent-728 based modeling approach for the spread of Cutaneous Leishmaniasis disease in central Iran Modeling and analysis of global epidemiology of 732 avian influenza Investigating the cases of novel coronavirus disease (COVID-19) 735 in China using dynamic statistical techniques Beware of asymptomatic transmission: Study on 2019-nCoV prevention and 738 control measures based on extended SEIR model An agent-based/network approach to spatial epidemics. Agent-based models of 741 geographical systems The epidemiological characteristics of an outbreak of 2019 novel 744 Coronavirus diseases (COVID-19)-China COVID-19 battle during the toughest sanctions 747 against Iran Spatial epidemiology of networked metapopulation: An overview. Chinese 750 Updated understanding of the outbreak of 2019 novel coronavirus 753 (2019-nCoV) in Wuhan Modeling epidemics using cellular automata Report of the WHO-China Joint Mission on Coronavirus 759 Disease Clinical Management of Severe Acute Respiratory 763 Infection when Novel Coronavirus (nCoV) Infection Is Suspected: Interim Guidance World Health Organization (WHO), 2020. Coronavirus disease (COVID-19) Pandemic Real-time tentative assessment of the epidemiological characteristics of novel coronavirus infections 771 in Wuhan, China, as at 22 Estimation of the reproductive number 774 of novel coronavirus (COVID-19) and the probable outbreak size on the Diamond Princess cruise 775 ship: A data-driven analysis We would like to express our sincere thanks and appreciation to Dr. Ali Akbari due to his 580 beneficial medical advice as well as the personnel of the Ministry of Health and Medical Education 581 for providing and sharing the COVID-19 notifications. 582Funding information 583 This research did not receive any specific grant from funding agencies in the public, commercial, 584 or not-for-profit sectors. 585 We would like to express our sincere thanks and appreciation to Dr. Ali Akbari due to his beneficial medical advice as well as the personnel of the Ministry of Health and Medical Education for providing and sharing the COVID-19 notifications. The authors declare that there are no conflicts of interest.