Visual Data Science 2026 - Climate Agriculture Dashboard

An interactive dashboard visualizing global agriculture data and climate patterns across different crops, with a focus on accessibility and colorblind-friendly design.

Overview

This project presents an interactive Vega-Lite visualization dashboard that explores the relationships between:

• Agricultural yield and climate change
• Production trends over time
• Climate regime classifications
• Country-level trajectories

Features

• Interactive Crop Selection: Switch between MAIZE, RICE, SOYBEAN, and WHEAT
• Brush & Filter: Select regions of interest to filter other views
• Climate Regime Visualization: Categorizes countries into 4 regimes based on yield and climate factors
• Continent-based Analysis: Track agricultural trajectories by continent
• Colorblind-Friendly Palettes: All visualizations use accessible color schemes

Accessibility

This project prioritizes accessibility for people with color vision deficiencies. All color palettes are carefully selected to be distinguishable for people with red/green color blindness (deuteranopia and protanopia).

Key accessibility features:

• Okabe-Ito and Paul Tol colorblind-safe palettes
• Centralized palette management via viz/palettes.py
• CSS color variables in assets/palette.css
• Comprehensive testing with Color Oracle recommended

See ACCESSIBILITY.md for detailed information about our accessibility approach, palette choices, and testing procedures.

Getting Started

View the Dashboard

Simply open index.html in a modern web browser. The visualization uses Vega-Lite with embedded data and requires no server setup.

Using the Palette Module (Python)

If you're creating additional visualizations:

from viz.palettes import get_categorical_palette, set_global_palettes

# Set colorblind-friendly palettes globally
set_global_palettes('okabe_ito')

# Or get specific colors
colors = get_categorical_palette('okabe_ito', n=4)

See viz/palettes.py for complete API documentation.

File Structure

.
├── index.html              # Main dashboard (Vega-Lite visualization)
├── viz/
│   └── palettes.py        # Centralized colorblind-safe palette module
├── assets/
│   └── palette.css        # CSS color variables for web styling
├── ACCESSIBILITY.md        # Detailed accessibility documentation
└── README.md              # This file

Color Palettes

Climate Regimes (4 categories)

• Orange (#E69F00) - Vulnerable / Subsistence
• Sky Blue (#56B4E9) - Developing / Low-Input
• Bluish Green (#009E73) - Emerging / Mid-Scale
• Blue (#0072B2) - Efficient / High-Yield

Continents (6 categories)

• Blue (#0072B2) - Africa
• Orange (#E69F00) - Asia
• Bluish Green (#009E73) - Europe
• Vermillion (#D55E00) - North America
• Reddish Purple (#CC79A7) - South America
• Yellow (#F0E442) - Oceania

All palettes are from the Okabe-Ito collection, designed specifically for colorblind accessibility.

Testing for Colorblindness

We recommend testing visualizations with:

• Color Oracle: Real-time screen simulation (https://colororacle.org/)
• Coblis: Online image simulation (https://www.color-blindness.com/coblis-color-blindness-simulator/)
• Chrome DevTools: Built-in vision deficiency emulation

Contributing

When adding new visualizations:

1. Use palettes from viz/palettes.py or assets/palette.css
2. Test with colorblindness simulation tools
3. Ensure colors work in grayscale
4. Update documentation if adding new palette options

Dependencies

• Modern web browser with JavaScript enabled
• Vega, Vega-Lite, and Vega-Embed (loaded via CDN)

For Python palette module:

• Optional: matplotlib, seaborn, plotly, or altair (depending on usage)
• Optional: cycler (for matplotlib color cycles)

License

This project is created for educational purposes as part of Visual Data Science 2026.

References

• Okabe-Ito Palette: Okabe, M. and Ito, K. (2008) Color Universal Design (CUD)
• Paul Tol's Palettes: https://personal.sron.nl/~pault/
• ColorBrewer: https://colorbrewer2.org/
• Vega-Lite: https://vega.github.io/vega-lite/

---

Last Updated: January 18, 2026