Dr. Song Gao
Assistant Professor, University of Wisconsin-Madison
The World Health Organization (WHO) has declared the Coronavirus Disease 2019 (COVID-19) as a pandemic on 11 March 2020. It remains a big challenge to accurately predict the COVID-19 spread in the US and beyond. Our proposed research aims to bridge the gap between the science of epidemic predictive modeling and risk communication to the general public through the CyberGIS-Enabled Geospatial Hub. The increasingly available human mobility data from location-based services and social media have been utilized to understand multi-scale human movement patterns and heterogeneous social distancing compliance in response to COVID-19 pandemic. In addition, recent advancements in deep learning approaches are now providing fast and near human-level perception from massive datasets to inform public health decision makings. By combining deep learning techniques with multi-source human mobility sensors, we can more accurately predict the spatiotemporal spread trend of COVID-19 and identify potential infectious hotspots and most vulnerable communities in different geographic scales for risk communication, epidemic prevention, and mitigation.
The proposal includes two synergistic research activities: (RA#1): Investigate and develop an advanced deep-learning-enabled spatiotemporal-network epidemic model for predicting the COVID-19 spread in the US and beyond by integrating census demographic statistics, human mobility patterns derived from large- scale mobile phone users, and real-time social media big data, travel restriction policy, weather data, social distancing compliance, and community risk awareness. (RA#2). Develop educational materials for characterization of human mobility patterns from human mobility big data (e.g., mobile phone data and geo-tagged social media data) and prediction of COVID-19 spatiotemporal spread on Jupyter Notebooks.
The Web-based spatial epidemiology applications with data and code developed in this project will be shared on the Geospatial Hub Jupyter Notebooks, which will serve as an educational tool for students in geography, computer sciences, mathematics, and public health, as well as for effectively engaging with multiple stakeholder groups and communities about the science of COVID-19 spread. The results of this project will lead to a better understanding of the geography and spread of COVID-19. Additionally, it is expected that the methods developed in this project can be applied to mitigate the outbreak risks of future respiratory epidemics.
Dr. Peter Kedron
Assistant Professor, Arizona State University
Dr. Joseph Holler
Assistant professor, Middlebury College
This project will advance COVID-19 research by directly assessing the credibility of the research projects of GSI fellows through reproductions and replications. By conducting reproductions in collaboration with students, this project will develop teaching modules, training materials, and a pedagogical model for teaching reproducible research practices. Ensuring and demonstrating that the findings of the program fellows are reproducible will increase the likelihood of this research improving pandemic response and broadly impacting society. The teaching and training materials we produce will provide a pedagogical model other educators can adopt to disseminate reproducible research practices. Directly involving students in reproductions will introduce those students to reproducible research practices, cyberinfrastructure, and geospatial software standards and contribute to workforce development. Reproducing and replicating recent research with students will also accelerate the translation of COVID- 19 research findings into undergraduate and graduate education.
Dr. Clio Andris
Assistant Professor, Georgia Tech
Social networks are rarely analyzed in geographic space (Andris, 2016; Radil, Flint, & Tita, 2010; Sarkar, Sieber, & Sengupta, 2016). As a result, it is difficult to merge geospatial features with features of a dynamic social system of links and nodes. As social network contact tracing becomes more of a needed reality (Park, Choi, & Ko, 2020), it has become apparent that we lack systems to infer patterns from contact tracing data due to issues of privacy, granularity, and reliability of reporting (Dar et al., 2020). One major problem for operations is that spatiotemporal colocation does not always equate with high transmission probabilities. For example, if two individuals cross each other on the highway or while jogging individually at a large park, they may register as being spatiotemporally collocated, but in reality, the probability of transmission is very low. However, if these individuals spent time in a small home together, we can infer a higher probability of transmission (Freeman & Eykelbosh, 2020).
CyberGIS infrastructure and computing technologies allow for the integration of multiple database types and high-performance computing (Wang, 2010) that can be used to improve our ability to perform reliable contact tracing. In this proposed research, we leverage the CyberGIS research framework and frameworks for developing geovisual and analytic methods (Luo & MacEachren, 2014) for analyzing a geolocated network of contacts. The major tasks of this project are: (1) produce an R package that facilitates the visual and numerical analysis of spatial social networks, (2) create an educational tutorial to help guide users through the functionality of the package, and (3) apply this technology to a case study in partnership with the Fire Department of New York City (FDNY).
The R package will contain methods and functions that input Spatio-temporal activities of individuals, non-work risk characteristics of individuals, and a GIS layer of local risk as input. Methods will be created to reshape and pre-process the data to allow for multiple inputs and different researcher and practitioner domains to interact with the package.
The software developed in this project will be used by FDNY to create better policies for managing their teams. Specifically, by determining whether certain workers are more susceptible to infection, the FDNY will implement a strategy to spatially and socially limit interaction with colleagues who may be infected. This policy will be a refined schedule of operations that communicates the risk of interacting at a firehouse, and scheduling pods to limit interaction throughout the employee network. The software developed here will also be integrated with the GSI hub as an example of geospatial software that innovates by accounting for geolocated interpersonal relationships and non-planar network analysis.
Dr. Naomi W. Lazarus
Assistant Professor, California State University
The project will conduct a county-level assessment of COVID-19 in relation to age demographics and comorbidities of the exposed population. Current data and research on the coronavirus have focused on raw case numbers and in identifying hotspots related to positive cases. While this approach is useful to monitor the effectiveness of mitigation strategies, it does not account for the geographical variation in population demographics across the country. This project will address how age and underlying conditions impact the incidence rate and mortality rate of COVID-19, thereby providing some context to the emerging hotspots of the virus.
The causal relationship between coronavirus, age, and underlying conditions will be examined using geographically weighted regression (GWR). It is expected that the results will highlight the degree of change in COVID-19 incidence and mortality in response to the explanatory variables. The project advances the Geospatial Software Institute’s goal of reproducibility and transparency by utilizing spatial statistical software and web applications that are readily available to a wide range of users. Furthermore, it is committed to improving the capacity and capabilities of the geospatial community by producing deliverables that are accessible and user-friendly. The proposed deliverables include, (1) a comprehensive dataset of incidence, mortality, age, and comorbidity; (2) an interactive story map displaying the results of the regression analysis; (3) an instructional notebook outlining the steps undertaken in the analysis, and (4) a research paper that discusses the methodology and findings of the project.
Dr. Xiang Chen
Assistant Professor, University of Connecticut
Many recent COVID-19 studies have relied on mathematical models to simulate future epidemic scenarios. These mathematical models, however, have been largely focused on the macro-scale with relatively large analysis units (e.g., states, counties). When social distancing measures are in effect, however, new cases of infection would be primarily induced by community spread, such as the human movements within and between neighboring cities, towns, and communities. To this end, this project aims to monitor and model COVID-19 infections at the meso-scale, defined as U.S. towns (i.e., county subdivisions). We will implement the project in three phases: (1) we will first develop a meso- scale SEIR model in a pilot study, where the town-level infection data are collected by our COVID-19 monitoring dashboard (covidct.org); we will also calibrate the model parameters with a continuously running travel survey (travel.covid-survey.org); (2) with the calibrated meso-scale model, we aim to simulate future epidemic development across the entire U.S. with different social distancing scenarios; (3) we will monitor the epi curves for every town in the U.S. and distribute modeling results using the Geospatial Hub’s CyberGISX platform. In addition, the project will be incorporated into an open-access GIS course module about CyberGIS and COVID-19, which can be accessed by the public for geospatial training purposes.
Dr. Daniel Goldberg
Associate Professor, Texas A&M University
Proposed study As the spread of COVID-19 gets severe, the importance of providing sufficient access to healthcare facilities has been emphasized, given that spatial accessibility measurement identifies the location with poor access and proposes sites where additional facilities should be provided. Through this project, we aim to tackle the following two objectives. The first objective is to enhance the accuracy of spatial accessibility measurement by incorporating the uncertainties in supply and mobility. In the context of COVID-19, the number of available resources (e.g., ICU beds or ventilator) at each hospital is uncertain, and the travel time between locations varies over time. To address the uncertainty in the inputs (i.e., supply and mobility), we will populate the uncertainty in each data from the historical fluctuation, and we implement the distribution for Monte-Carlo simulation to measure spatial accessibility from the stochastic perspective. The second objective is to expand the approach of spatial accessibility into indoor spaces with a case study of hand sanitizers. As most universities in the United States will reopen their campus for the following semester, they have provided guidelines to prevent the spread of the virus. Considering hand sanitizer is one of them, we will examine how well the proposed locations of hand sanitizers serve the entire space. We will take the hand sanitizer as supply, the capacity of rooms as demand, and the travel time (i.e., walking) as mobility to measure indoor spatial accessibility.
Intellectual merit This project will advance the spatial accessibility measurement from the following aspects. First, objective 1 will increase the accuracy of spatial accessibility by incorporating the uncertainties of the inputs. Second, objective 2 will demonstrate how the application of the spatial accessibility measurement can be extended to indoor spaces. Lastly, the project will foster the reproducibility and replicability of the expected results by providing detailed information through the Jupyter notebook embedded in CyberGISX.
Broader impacts For urban planners and decision-makers, this project will determine probabilistically venerable location to access healthcare facilities. It will propose where additional facilities should be provided. Hence, it has the potential to relieve the mortality rate of COVID-19 and prevent new infectious diseases in the future. For universities, including Texas A&M University, this project will determine whether the currently proposed location of hand sanitizer so that it would raise students’ hygiene by providing additional resources to necessitated location.
Dr. Kenan Li
Research Scientist, University of Southern California
Dr. John Wilson
Professor, University of Southern California
Human mobility is related to the spread of human-transmitted diseases such as the COVID-19 virus and the high spatiotemporal granularity of location services supported by most smart devices provide new opportunities to delineate these relationships. This is important because the traditional human mobility metrics, such as home-dwelling time, traveling distance, and so on, do not directly reflect the individual contact rates and show weak correlations with the spread of the disease. Our research team has initiated an effort to collect anonymous GPS pings from smart devices accounting for around 10% of the total U.S. population since January 2019, and this allows us to further calculate the contact rates of individuals within certain threshold distances of others. In this proposed project, we plan to calculate the daily contact rates that directly control the transmissibility of COVID-19 (Aim 1), develop cutting edge statistical methods to discover the spatial and temporal trends of the daily contact rates and their social determinants and confounders (Aim 2), and build a spatial SEIR model to better predict the virus spread, hospital impacts, and personal protective equipment (PPE) needs (Aim 3). The deliverables will increase the nation’s capacity to simulate the epidemiological dynamics of COVID-19 at fine spatiotemporal scales for knowledge discovery in more reproducible and transparent ways.
Dr. Jayajit Chakraborty
Professor, The University of Texas at El Paso
Persons with disabilities (PwDs) maybe four times more likely to be injured or die than non- disabled individuals during the COVID-19 pandemic, as well as health emergencies caused by climate change and other disasters. Several reports indicate that PwDs are adversely impacted by COVID-19 and face multiple challenges. While recent studies have found greater COVID-19 incidence and fatality rates in counties containing higher proportions of racial/ethnic minorities, people in poverty and other socially disadvantaged groups, geospatial research on the distribution of COVID-19 burdens in the U.S. has paid limited attention to PwDs. The proposed project seeks to address this gap by conducting a comprehensive, comparative, and systematic analysis of the relationship between exposure to COVID-19 and disability characteristics at two geographic scales: the continental U.S. (national) and Harris County, Texas (urban). The research goals are to (1) examine whether PwDs are significantly overrepresented in areas with higher COVID-19 incidence/fatality rates compared to people without disabilities, after controlling for relevant contextual factors; (2) determine whether socially disadvantaged PwDs (based on age, race, ethnicity, poverty, and employment status) are significantly overrepresented in areas with higher COVID-19 incidence/fatality rates compared to other subgroups of PwDs, after controlling for relevant factors; and (3) identify hotspots where significantly greater COVID-19 incidence/fatality rates coincide with higher proportions of PwDs and specific subgroups. In the first phase of this project, relationships between COVID-19 incidence/fatality and various disability characteristics will be analyzed at the county level in the continental U.S. In the second phase, these relationships will be analyzed at the level of ZIP codes in Harris County, Texas, the third-largest U.S. county located in the Greater Houston metropolitan area. Harris County is a particularly suitable study area because it contains both the highest number of PwDs and the highest number of COVID-19 cases in Texas. This research will encompass univariate, bivariate, and multivariate analyses based on several contemporary and cutting-edge geospatial modeling techniques.
The Geospatial Hub represents an ideal forum for disseminating and sharing project outputs via approaches that are consistent with the goals of generating reproducible, transparent, and scalable data and software. It will be used to provide access to brief reports and research articles summarizing analytical findings, relevant educational materials, and computational notebooks representing geospatial data and software modules. This project seeks to advance the goals of the Geospatial Hub in multiple ways. First, it will contribute cutting-edge knowledge and insights on relationships between COVID-19 exposure and various disability characteristics—an understudied and interdisciplinary topic that has significant implications for health equity, social justice, and relevant public policy. Second, this research will build the conceptual and methodological foundations for more geographically detailed analyses of the impacts of COVID-19 on PwDs in other regions, states, and urban areas. Third, this project will benefit the broader geospatial community by providing access to several resources for studying the association between COVID- 19 and disability, including reproducible geospatial data and software. Fourth, project outputs will enhance the Geospatial Hub by adding a disability-focused component that can be used to explore the spatial variability and diversity of PwDs, as well as the exposure patterns of specific disability subgroups to COVID-19 risks and other disasters. Finally, relevant findings and products from this project will also be shared with PwDs and their families, healthcare workers, service providers, decision-makers, and many other stakeholders, thus extending the broader societal impacts beyond the academic community to increase public knowledge and understanding of how PwDs are adversely impacted by various public health emergencies and disasters, including COVID-19.
Dr. Daoqin Tong
Associate Professor, Arizona State University
The COVID-19 pandemic is fueling a range of devastating consequences to the U.S. economy. Soaring unemployment rates have pushed millions of families into food insecurity. Meanwhile, the social distancing requirement is also changing the way that people access food with many people replacing in-store visits with online food shopping and pickup. Not only does this require a reliable computing device or mobile phone, but it also requires internet access, some type of footprint in the formal banking system (e.g., debit or credit card), and in some cases, transportation to/from the store. This shift in consumer behavior may have significant impacts on disadvantaged population groups.
This GSI Fellowship Project combines a large food access spatial dataset and relevant socioeconomic and demographic data to examine how economic stress and the new rules for obtaining food impact healthy food access during COVID-19. The study consists of four tasks. Task 1 focuses on spatial data collection. Task 2 consists of producing a food access map to compare the patterns before and during COVID-19. Task 3 focuses on an examination of the potential factors associated with the food access pattern change. Task 4 explores strategies that could be used to help improve healthy food access in disadvantaged neighborhoods. I plan to conduct the study in Maricopa County. With a population of more than 4 million, the region is among the fastest-growing places in the U.S. and is a location where food insecurity is a particularly pressing issue even before COVID- 19.
The study will be among the first attempts to use geospatial data to examine food access under the challenges brought by the COVID-19 pandemic. The project will help advance the GSI goals. The data processing procedure in this project involves a large flow data set and is generalizable, and easily transferred to different geographic scales for examining the impacts of COVID-19. The integration of geospatial data will also advance knowledge discovery in the field of food access. Food access changes may remain even after the pandemic subsides. Therefore, the research might have profound impacts in food studies. The empirical study in Maricopa county will also provide important context and deepen our understanding of the food access challenges for vulnerable population groups – helping develop and plan more resilient food systems.
Dr. Daniel Block
Professor, Chicago State University
The COVID-19 crisis has put a focus on vulnerabilities and inequities in contemporary food production and distribution systems in the United States. On the production side, COVID-19 outbreaks within the meatpacking sector led to plants closing and temporary shortages and exposed the vulnerability of the workers themselves (Grabell, Perlman, and Yeung 2020). On the distribution side, civil unrest following the killing of George Floyd led to the temporary or long-term closing of many grocery stores, particularly in areas with already low food access (Brown, 2020).
The proposed project involves two spatial data collection projects brought together by coalitions of governments, organizations, and academics in Chicago and suburban Cook County, Illinois. The first is an effort to build a frequently updated food environment map of Chicago and Cook County. This map was originally inspired by the closings of stores in early June 2020 following civil unrest after the killing of George Floyd. The City of Chicago wanted to create a map of up to date (daily or weekly) information on store status. This immediate need was met, but it was agreed by partners that there was a need for the region to have access to an integrated food environment map. A core group consisting of academic, government, non-profit, and business partners is currently working on developing this map. Databases will be provided by multiple agencies and at differing frequencies. The key task will be integrating and maintaining these databases.
A second, but related, project encompasses data brought together and collected to evaluate the adoption of the Good Food Purchasing Program (GFPP) in Chicago and Cook County. GFPP is a metric-based procurement framework that supports institutional food buyers to direct their buying power, make informed decisions, and measure impact towards five core value categories: local economies; environmental sustainability; valued workforce; animal welfare; and nutrition (Center for Good Food Purchasing n.d.). The City of Chicago and Cook County began adopting the GFPP in 2017-2019 and a consortium is currently evaluating the implementation and effects of the program. The increased understanding from this evaluation involves collecting and bringing together multiple datasets on food production and distribution within the Chicago “foodshed” (the area providing Chicago with its local food), as well as data from foodservice contracts with government agencies in Chicago and suburban Cook County.
These are large, connected, collaborative projects involving many of the same people and agencies. Individual datasets are not generally large but are derived from multiple agencies and academics and at varying frequencies. A core question of both projects is: How can innovations in data science help identify and facilitate more sustainable, healthy, and equitable outcomes in urban food systems? Towards this end, I propose to 1) research and work with fellow team members to implement cyberGIS techniques within the two interrelated projects, if they are found to be applicable to the projects; 2) create a white paper with suggestions for how cyberGIS and data science techniques can be integrated into projects involving multiple databases of modest size updated at varying frequencies 3) work with team members (and possibly other fellows) to create a publishable article on the possible application of cyberGIS and data science to “Community Geography” mapping projects.
Dr. Xun Shi
Professor, Dartmouth College
Most current modeling works of COVID-19 use extensions or variants of the classical susceptible- Infected-removed (SIR) model, which is typically at the population level and adopts the top-down strategy. Such models are sensitive to the modeling setting and parameter values that feature subjectivity. Their highly aggregate input data also mask variation across people and space. An alternative approach is to model at the individual level and to resort to a bottom-up strategy. This latter approach takes advantage of increasingly available big data of individuals’ mobilities and high-performance computing capacity, trying to model the individual-level transmission through tracing contacts of individuals. It anticipates the general and meanwhile high-resolution spatiotemporal pattern of the epidemic to emerge from the modeling process. This bottom-up approach has great potential but is not popular yet in epidemic modeling, mainly due to certain bottlenecks in data precision and method/software development.
Those bottlenecks are what this proposed project is to address. In the past several years, we have developed components of the bottom-up approach to epidemic modeling. This project will provide an opportunity for us to further develop and assemble those components into an integrated, complete, and executable procedure to generate reproducible and replicable results. This is also an opportunity to promote this procedure in both research and education domains.
The specific components form a technical series, and they include: 1) disaggregating aggregate data of human mobility to obtain individual-level trajectory data of people’s daily activities, using a Markov Chain Monte Carlo (MCMC) process; 2) tracing contacts among individuals based on their daily activity trajectories, for which the data are obtained from the previous component; 3) building a tree model that represents transmission relationships among actual disease cases in the study area, based on the contact data obtained from the previous component; and 4) modeling/predicting future epidemics based on the epidemic characteristics derived from the tree model built in the previous component, as well as data depicting intervention scenarios. This last component is essentially an agent-based modeling (ABM) process.
We have developed procedures and associated software tools for each of these components, and have fully or partially tested the procedures and tools in a number of studies of communicable diseases, including our recent studies of COVID-19 in Nanchang and Chengdu, two cities in China, and in Boston. In this project, we intend to further develop the tools to make them suit more general data and analyze situations and to be more robust. We have not put these components together and tested the whole procedure is a real-world study. Such integration and test are the goals of the proposed project.
Currently, the tools run in a local desktop environment. Through the proposed project, we expect to migrate the tools and assembled procedure to the cyberGIS environment, so as to make them to be more computationally capable and to be more accessible to researchers and educators. The tools and the procedure should be useful for studies of current COVID-19 pandemics, and for future epidemic incidents of other communicable diseases.
We believe that the integrated procedure and assembled software package for the bottom-up approach to epidemic modeling that we are proposing are highly relevant to the two goals of GSI and can make a considerable contribution to the construction of GSI’s Geospatial Hub.
Dr. Ningchuan Xiao
Professor, Ohio State University
Changes in human mobility due to the outbreak of COVID-19 hold the key to effectively fight the disease spread and to expedite economic recovery. However, understanding such changes is a significant challenge because of the lack of high quality and high spatiotemporal resolution mobility data, as well as effective methodological frameworks. In this project, we propose to develop a computational framework that utilizes a new source of geospatial data, the traffic camera feeds, to explore the impact of COVID-19 on the origin-destination (OD) flows of on-road vehicles. First, we will develop new automatic vehicle detection methods by extending deep learning-based object detection models to identify the motor vehicles present in-camera images and estimate the traffic density captured by the traffic cameras. We will then develop a probabilistic model to estimate the OD flows of on-road vehicles based on the traffic density obtained from traffic cameras. Finally, we will analyze how OD flows change in response to the pandemic as well as the social and economic contexts of such changes. This project is focused on the area around Columbus, OH, and we aim to make the methodological framework generalizable for other areas.
Intellectual merit. The proposed study will advance our understanding of COVID-19 in two directions. First, a computational framework will be developed to fully utilize traffic camera feeds to explore mobility variations that are closely related to the COVID-19 outbreak. Human mobility patterns are directly impacted by the public health crisis. Changes in mobility patterns before, during, and after the outbreak will not only reveal the breadth and depth of the impact but also provide insight on how the economy and society will recover from the pandemic. Second, by linking mobility changes with their social and economic contexts, this research will reveal how the impact of the pandemic varies across different areas with different backgrounds.
Broader impacts. This project will produce several computational notebooks and software modules that will contribute to a broader research and education community of human mobility and public health in general. The human mobility data produced in this research provides a new perspective by using a new data source that directly captures vehicle movements. This data can be further used together with other human mobility measures from social media and mobile devices through data fusion, which will lead to accurate and comprehensive mobility research. Computational methods and software tools developed in this project can be deployed as part of the CyberGISX environment and can be used together with other software packages. In addition, the findings of this project have the potential to inform policymakers and the general public about the full spectrum of how the COVID-19 pandemic will affect the society.
Dr. Andrew J. Greenlee
Associate Professor, University of Illinois at Urbana-Champaign
Prior to the economic and social disruption associated with the COVID-19 pandemic and ensuing public health measures, the United States was experiencing an eviction crisis. As a direct result of the massive economic and social shock we are currently experiencing, an estimated 20- 28 million additional households stare down the direct threat of eviction. A patchwork of national and local eviction moratoria allows many of those households a short respite from their day in court, however, as these moratoria and other forms of financial triage associated with the CARES Act expire, the rate of evictions is likely to skyrocket, with massive implications for renters, landlords, and society in general. There is no central repository for information on eviction filings and completions, and no consistent data source tracks post-eviction outcomes.
This geospatial fellows proposal calls for the development of a novel system for tracking pre eviction and post-eviction residential locations and household characteristics. Drawing upon New York City as a pilot case, I propose combining public data on evictions, property types, and neighborhood characteristics with private data on household residential histories. Merging and validating these data reflect a major innovation in and of itself. I then propose the development of an information portal which allows the public to query information on the characteristics of evicted households, as well as the characteristics of their residential location trajectories before and after eviction. This portal will heavily leverage the capacity and expertise present in the CyberGIS center to produce reproducible code for visualizing complex and numerous longitudinal housing trajectories and sequences of residential location choice and change.
I also propose initiating two specific applications of these data to answer basic social science research questions related to human mobility – a) how different are the residential trajectories and characteristics of households currently facing eviction in the wake of covid from evictions which occurred over the past few years in New York City; and b) based upon how past evicted households have moved following their evictions, given current housing market characteristics and available housing options, where are households facing eviction now likely to move?
These outputs help to affirm the proposed proof of concept and make a major contribution to the general understanding of post-eviction location outcomes. These outputs also help to isolate the myriad effects of covid on New York’s housing market, and the ways in which it interacts with other regional and national housing markets. The notebooks and other outputs from this project should prove useful for housing advocates, public accountability, and to encapsulate complex information for lay-audiences. The notebooks will also form the basis for heuristics in a University of Illinois at Urbana-Champaign graduate workshop on neighborhood analysis.
Dr. Ruby Mendenhall
Associate Professor, University of Illinois at Urbana-Champaign
Despite decades of computer science pipeline programming, the participation of students of color in computer science and information technology disciplines remains alarmingly low. STEM Illinois is a project deeply rooted in the historic mission of land-grant institutions, which is to democratize higher education and to address the world’s most pressing societal challenges. The Youth Citizen (Community) Scientists Mapping subproject will work with 50 to 100 students, mostly students of color whose parents do not have a college degree and who are at risk for dropping out of school. Computer scientists from these backgrounds represent a small number of faculty and industry professionals. This project seeks to evaluate and analyze the impact of COVID-19 on the vulnerable communities where these youth live by using a radical model of intergenerational outreach and education. This model seeks to create a culture of innovation where all members of the community see, understand and feel that they can support their youth in computer science innovation and COVID-19 mapping. This project will provide underrepresented students with computer science activities that include: (1) using spatial analysis to map vulnerable communities or hotspots, (2) attempting to predict the spread and impacts, (3) understanding the impacts and (4) creating tools to decrease the spread and impact of COVID-19 such as data visualizations, infographics, photographs, workbooks, etc. We will seek to have the outcomes of the Youth Citizen (Community) Scientists be disseminated via publications, during Youth TEDxUIUC talks, in the National Geographic magazine, and an exhibit at the National Museum of Smithsonian Institution Traveling Exhibition Service.