I am an architectural designer. I've always wanted to understand what our past climate and temperatures were really like — whether they were relatively stable or becoming increasingly extreme.
Using AI, I transformed decades of global weather station historical data into K-line (candlestick) charts and displayed them on a 3D globe. This makes it much easier to compare and analyze past climate patterns.
I also believe this visualization could be very useful for farmers and agricultural professionals, helping them review historical weather trends to better understand past harvests and make future decisions.
Simply search or click on a city, and you'll see long-term trends for temperature, humidity, wind speed, and more — clearly revealing day-night differences and extreme weather events.
🔗 打开交互演示 https://global-weather-k-line.vercel.app/ (Currently showing a curated subset of major cities)
- Smooth 3D globe interaction (rotate, zoom, and click cities)
- Interactive K-line charts for temperature, humidity, and wind speed
- Clear visualization of long-term trends, day-night differences, and extreme weather events
- Lightweight and responsive design
- Frontend: Vanilla JavaScript
- Charts: Lightweight Charts
- 3D Globe: Globe.gl
Public historical weather station data from global meteorological archives.
All data files include a fingerprint:
source: "Kingcat666_weather_kline_v2026"
Note: This is a curated subset for demonstration purposes only. The full dataset is significantly larger.
MIT License
I'd love to hear your thoughts:
- Is this visualization style clear, useful, and visually appealing?
- Which cities or weather indicators would you like to see added?
- Any suggestions or ideas are welcome!
Feel free to leave a comment or reach out to me on X (Twitter).
2026.4.3_feat: add X (Twitter) and Reddit share links to footer——index.html 2024.4.3—add Osaka data