This project explores the spatial distribution and impacts of severe convective storm activity across the contiguous United States between 2014 and 2023 through a series of thematic cartographic visualizations. Developed as a large-format statistical poster, the project examines storm frequency, storm-related injuries, and normalized property damage patterns using multiple complementary mapping techniques.

The project was designed to communicate large and spatially uneven environmental datasets in a visually accessible and analytically meaningful way. Particular emphasis was placed on balancing statistical clarity with cartographic readability while exploring how different thematic mapping methods influence the interpretation of geographic patterns and risk.

The project workflow began with acquiring county and state boundary datasets from the U.S. Census Bureau TIGER/Line database alongside severe convective storm records from the NOAA Storm Events Database covering the years 2014–2023. Additional demographic information was obtained from the 2020 Census through the IPUMS NHGIS dataset to support normalization and comparative analysis.

A custom Python workflow developed with AI-assisted scripting support was used to clean, process, and restructure the raw datasets into analysis-ready tables, including County_Population_2020, Storm_Events_Summary, and Convective_Storm_Events. This process involved removing unnecessary attributes, standardizing field structures, converting shorthand NOAA damage values (e.g., 1M, 1B) into full numeric USD values, and preparing records for geospatial joins and statistical aggregation.

All spatial datasets were projected to EPSG:5070 — NAD 1983 Contiguous USA Albers to ensure consistent area representation and improve visualization across the contiguous United States. TIGER/Line shapefiles were imported into ArcGIS Pro, validated for extent and geometry, and exported into simplified state and county feature classes within the project geodatabase. Attribute joins were performed using GEOID identifiers to associate cleaned population data with county polygons, and SQL queries were used to isolate only the contiguous United States for analysis (Alaska, Hawaiʻi and US territories were not included).

Several custom attribute fields were created to support statistical mapping and aggregation workflows. These included a combined FIPS5 county identifier, total injury calculations, injury-event flags, normalized property and crop damage fields converted into full USD values, and total storm damage metrics. Additional event-count fields and summary statistics tables were generated to support thematic mapping and comparative visualization throughout the poster series.