Skip to main content
Zhanyang Liu profile picture

Zhanyang Liu

Political Science • University of Wisconsin–Madison

← Back to Projects

Tornado Event Visualization Dashboard

Streamlit Altair Collaborative Project Python Data Visualization

A collaborative Streamlit dashboard project that visualizes tornado patterns across the United States, featuring interactive analysis of frequency, size distribution, and geographic intensity patterns.

Project Overview

This project was developed as part of a collaborative effort to create an interactive narrative around tornado events in the United States. I contributed multiple modules to a comprehensive Streamlit dashboard that allows users to explore tornado data through various analytical lenses.

The dashboard combines temporal analysis, geographic visualization, and statistical insights to provide a comprehensive view of tornado patterns, making complex meteorological data accessible through intuitive interactive visualizations.

My Contributions

Monthly Frequency Analysis:

Developed interactive charts showing seasonal tornado patterns and monthly occurrence rates across different years.

Size Distribution Visualization:

Created visualizations analyzing tornado intensity distributions using the Enhanced Fujita Scale, revealing patterns in storm severity.

State-wise Intensity Mapping:

Built geographic visualizations showing tornado intensity patterns across different states, highlighting regional variations in storm characteristics.

Interactive Plot Development:

Utilized Altair to create responsive, interactive visualizations that allow users to filter and explore data dynamically.

Technical Implementation

The project leverages Streamlit for the web interface and Altair for creating sophisticated interactive visualizations. The dashboard processes historical tornado data from the Storm Prediction Center, applying data cleaning and transformation techniques to ensure accurate analysis.

Each module was designed with modularity in mind, allowing for easy integration into the larger dashboard while maintaining individual functionality. The collaborative nature of the project required careful coordination of data pipelines and consistent visualization styling.

Key Insights

The dashboard reveals significant seasonal patterns in tornado activity, with peak occurrences during spring months in the central United States. State-wise analysis shows clear geographic clustering of high-intensity tornadoes in traditional "Tornado Alley" regions.

Interactive features allow users to discover correlations between tornado frequency, intensity, and geographic factors, providing valuable insights for both educational and research purposes.