1. IntroductionIn the United States, the number of cases of COVID-19 is continuously increasing. Until October 28, 2020, according to John Hopkins Coronavirus Resource Center, there were 8,856,413 confirmed cases in the United States, and 72,183 cases were reported in one day. After a few months of the lockdown in March, most states are reopening, including retail stores, restaurants, and recreation. As a college student, COVID-19 is affecting everyday student life and residence life.According to the CDC, seasonal influenza viruses are expected during the late fall and peak between December to February [1]. There are some explanations about why flu season usually spreads in the winter. First, people spend more time indoors, which increases the chance to closer contact others who might be carrying the virus. Students, for example, would prefer using public transportation, such as buses, instead of walking to class. Second, in the short days of the winter, people may run low on Vitamin D and weaken our immune system [2]. UConn is located in the northeast of the United States; the temperature is low during the fall and winter. Students and the university need to be prepared and preclude the new wave from spreading. Typically, international students and out-of-state students are more vulnerable to get infected due to limited access to testing [3].The government of Connecticut announced the reopening policy phase 2 began on June 17, which up to 50% capacity indoors with 6 feet spacing for restaurants, personal services, libraries, and indoor recreation and up to 25% capacity capped at 100 people for indoor religious gatherings. Phase 3 began on October 8. Restaurants, personal services, and libraries are up to 75% capacity indoors and up to 50% capacity for indoor and outdoor religious gatherings. However, due to the increasing number of cases of COVID-19 in Connecticut, the Connecticut government updated the latest reopening rule, which was phase 2.1 started on November 6. Phase 2.1 is slightly different from the phase 2 version, in which restaurants can accommodate up to 50%, while personal service and libraries can accommodate up to 75% [6].This article focuses on a local scale, which is the UConn main campus. It is important because college campuses are places with high dense population and easily get infected. From a student’s perspective, building spatial models of campus areas are necessary and help us create a safe community. This study article focuses on building a mathematical model, the Susceptible-Infected-Recovered (SIR) model, and estimates the infectious rate and recovery rate at the University of Connecticut (UConn) Storrs. The model generates the number of cases from August 16, when students who live on campus check-in, to September 7. After finding out the parameters using SIR, we use Agent-Based Modeling (ABM) to simulate different cases to predict and evaluate the risks of different places on campus.  UConn, located in Storrs, has approximately 5,000 students living on campus. Such a population would increase the chances of interaction between students in public places such as academic buildings, dining halls, grocery stores, residential halls, and apartments. Before the semester began, UConn had already announced reopening policies. Most of the classes are moving online or distance learning to prevent the spreading of disease. In-person classes require students wearing a mask and maintaining at least six feet of physical distancing from others. Dining halls are switching to take-out and limited dining models. However, for those students who live in residential halls, even though UConn policy requires one person per dorm room, they are still sharing bathrooms. For those who live in apartments or off-campus, students have approximately one to four roommates, which increases the chance of infection. Our primary goal is to extend the SIR model into the spatial form and using QGIS and NetLogo to visualize the spreading. Because the covid-19 disease varies a great deal with places, we consider leveraging this when we estimate covid information for policy-makers to make lockdown or reopening business strategies. We extend the traditional mathematical SIR model into a spatially-explicit model to simulate the spatial dynamics of covid-19 over discrete-time and across discrete space at the Uconn Storrs campus. The spatially-explicit models may provide useful insights into the epidemiological characteristics of the disease and identification of disease hotspots across the campus, thus can inform and guide policy-makers for targeted interventions and targeted reopening the business in specific locations of the campus. This paper focuses on a specific area, rather than a state or a country, with a smaller population size. We are using the data to predict the cases and infection rates in the next few months, evaluating each building’s risk and ranking the score with a higher chance of getting infected. Based on the policies that have been implemented at UConn, we also make some suggestions to the university about forestalling the new wave coming in winter. 2. Data and MethodologyTo simulate the spreading of epidemics, we are building the SIR model. The SIR model was first introduced by Kermack and McKendrick by separating people into three different categories: susceptible (S), infected (I), and recovered (R) [4]. In this case, the population in Storrs is susceptible (S). Individuals who get infected move from susceptible stage to infected stage (I). Eventually, people who were removed from the infected status recovered (R). The SIR model using the parameters β, the infection rate, and γ, the recovery rate, can be presented by the ordinary differential equation (ODE).