Collect data and map the aggregation of environmental factors (floods, droughts, temperatures, wildfires, rainfall) and the distribution of CommonSpirit Health (CSH) patients. The climate vulnerability indicators are mainly presented for the 24 U.S. states served by CSH, as well as those with actual patient data in Chattanooga, TN, and Omaha, Nebraska.

I am a Data Analyst at the Johns Hopkins University Spatial Science for Public Health Center (SSPHC). In my role, I utilize ArcGIS Pro, SaTScan, and R to perform advanced spatial statistical analyses, assessing public health trends and contributing to impactful research. I continue to collaborate with my advisor on the CHARMED pilot project, applying and expanding my skills in spatial analysis, data integration, and statistical modeling. My work includes analyzing the spatiotemporal correlations of crime rates, heat index, Social Vulnerability Index (SVI), and other socio-economics factors in Baltimore.Additionally, I have collaborated with my supervisors to develop and modify a web-based dashboard for the ECHO project. We are committed to enhancing this tool to make it more user-friendly and feasible. Moving forward, we will continue our collaboration with ECHO to conduct epidemiological studies using spatial analysis techniques.

Spatial Science for Public Health Center / Environmental influences on Child Health Outcomes (ECHO)Project: ECHO Data Analysis Center Atlas• Used ArcGIS Pro to create and modify the ECHO dashboard, a web-based mapping platform that summarizes ECHO’s Five Health Outcomes and related exposure variables.Project: ECHO-COVID-HIV geospatial project• Used non-traditional epidemiological approaches to summarize social determinants of health at the district level; assessed criterion validity on health outcomes, including COVID-19 and HIV.• Implemented an innovative approach to disaggregate area-level data using geostatistical techniques and project it onto hexagonal units, thereby establishing more precise units for neighborhood-level analysis.

Project: Evaluating the Co-Influence of Temperature and Vegetation Cover on Crime Rate: A Case Study of Baltimore City• Compiled and refined extensive open-source datasets including crime, satellite, and temperature data, alongside socioeconomic indicators; applied data cleaning techniques in R (multiple imputation).• Clean missing data estimation spatially; adjusted the resolution of the data (linear interpolation, inverse distance weighted method, and population-weighted average method).• Applied advanced statistical analyses to address data complexities, identifying outliers, and examining collinearity (Variance Inflation Factor, stepwise method, Zero-Inflated Poisson models, leave-one-out cross-validation, and Moran's I). • Employed spatiotemporal analysis and crafted visualizations to display daily crime trends and parameter relationships (R-INLA model, plotly, highcharter, and ggplot2).

National Taiwan University College of Public HealthInstitute of Epidemiology and Preventive Medicine Project: Build up the Taiwan Burden of Disease Center (TBDC)• Utilized R to analyze disease prevalence and mortality rates from Taiwan's National Health Insurance Research Database (2000-2015). Calculate DALYs for each disease by age and gender across Taiwan regions, using GBD-recommended disability weights and life tables.• Analyzed annual patient survival compared to the death registry and categorized diseases by diagnostic codes to define the cohort.• Developed R-Shiny dashboards to display health metrics for TBDC.Project: Assessing the Health and Economic Impacts due to Disease and Risk Factors in Taiwan• Applied the GBD-recommended relative risk table to calculate population attributable fractions (PAFs) for each risk factor, incorporating a novel approach of sampling relative risk values under a normal distribution to estimate PAFs at the smallest geographic levels.• Following sampling, these sorted estimates were aggregated to larger geographic units to generate precise point estimates and confidence intervals, enhancing the accuracy and interpretability of health impact assessments across varied spatial scales.Project: Assessing health and economic benefits of air pollution reduction policies in Taiwan• Integrated township-level PM2.5 data into broader county and city-level exposure metrics using a population-weighted average approach and assessed the attributable disease burden from 2000 to 2015 through random sampling based on the normal distribution of relative risks.• Forecasted future PM2.5 concentrations up to 2050 under various policy scenarios using a log-linear model and predicted the rate of non-attributable disease burden for each disease using the ARIMA model.

Project: The hydrological characteristics of Hehuan Mountain watershed and impact assessment of climate variation• Implemented linear models in MATLAB to impute missing rainfall data.• Applied the SWAT-MODFLOW model in ArcGIS to simulate interactions between surface water and groundwater.• Conducted Fourier transform fractal analysis on each hydrological component and rainfall using MATLAB.Project: Investigation of potential sites for geological storage of carbon dioxide in the land and sea areas of central Taiwan and development planning for carbon storage• Established initial and boundary conditions based on reference values from the geological storage site.• Employed TOUGH2 for simulating carbon dioxide storage and leakage phenomena.