The UFO Explorer is a comprehensive data visualization project developed at Aarhus University to analyze a century of reports from the National UFO Reporting Center. Utilizing a dataset of 147,890 entries, the project focuses specifically on sightings within the United States to investigate patterns involving UFO shapes, sighting frequencies, geographical distributions, and observer counts. The primary motivation for the project was to improve upon existing, limited visualizations by providing a highly interactive dashboard that allows both enthusiasts and experts to perform synoptic tasks, such as comparing temporal and spatial trends across multiple attributes simultaneously.

The development process involved significant data preprocessing and cleaning to ensure reliability. Because the data is crowdsourced and contains free-form text, the authors removed sightings recorded prior to 1940 to account for reporting inconsistencies and filtered out entries with invalid observer counts or misspelled location data. This resulted in a final refined dataset of 117,836 data points. To facilitate geographic mapping, state codes were translated into full names and matched with Federal Information Processing Standard codes, enabling the creation of detailed state- and county-level visualizations.

The final dashboard was implemented using a combination of D3.js and React, a choice made to support complex, high-performance interactivity that exceeded the capabilities of standard plotting libraries. The design follows Ben Shneiderman’s information-seeking mantra of providing an overview first, then allowing for zooming and filtering, and finally offering details on demand. The visual suite includes a treemap to categorize UFO shapes by frequency and area, a choropleth map for spatial analysis, dual histograms for yearly and monthly temporal trends, and stacked bar charts to analyze observer counts and specific sighting features like light changes or alleged abductions. To maintain a clean interface and maximize screen real estate, the team utilized tooltips for specific metrics and a theme-appropriate color palette featuring alien-green and colorblind-safe secondary tones.

A key discovery highlighted in the project is the dramatic shift in reported UFO characteristics over time; for instance, the dominant reported shape was typically a disk before the year 2000, whereas lights became the most reported shape in the decades following. By allowing users to cross-filter data—such as selecting a specific time range to see how it affects the distribution of shapes and locations—the dashboard facilitates a feedback loop of discovery. While the authors acknowledge limitations in data granularity and the binning of large observer groups, the project successfully demonstrates how interactive design can transform a massive, speculative dataset into a tool for identifying meaningful cultural and temporal trends.