"May the odds be ever in your favor..." Except they never are, and never were.
Every summer, headlines report a "new record high" like it's a curiosity. It isn't. It's a threshold — and for millions of people, crossing it is the difference between a hot day and a fatal one.
The Climate Games is an interactive data-storytelling platform built for the IEEE VGTC SciVis Contest 2026. It takes multi-terabyte atmospheric projection models and turns them into a gamified, arena-style map that asks the question climate averages are designed to hide:
Why are the nations facing the deadliest heat often the ones who did the least to cause it?
We didn't want another dashboard. We wanted something that makes you feel the asymmetry before you read a single statistic, so we built a leaderboard nobody wants to top.
🎮 Play the live app: scivis26.streamlit.app
-
Wet-Bulb Hazard Modeling (
$T_w \ge 28^\circ\text{C}$ ) — Not "how hot," but "can a human body survive this." We use Stull's psychrometric formulation on daily maximum surface temperature (tasmax) and relative humidity (hurs) to model the actual physiological survival ceiling, not just a thermometer reading. -
$P_{95}$ Zonal Aggregation — National averages flatten reality; a mountain range and a monsoon plain don't suffer the same way. We use 95th-percentile spatial zonal reduction across national boundaries so real heat hubs (the Indo-Gangetic Plain, the Sahel) show up instead of getting diluted by cool, uninhabited highlands. - A Story With an Arc, Not Just a Chart — A "Leaderboard Arena," animated temporal scrubbing across SSP2-4.5, SSP3-7.0, and SSP5-8.5, a dramatic "winner" reveal — and then the twist: the countries "winning" this leaderboard are losing the most.
-
The Climate Injustice Explorer — Toggle between three layers of the same story:
-
Physical Threat :
$P_{95}$ annual wet-bulb danger days -
Global Responsibility : per-capita
$CO_2$ emissions (Global Carbon Project) - Adaptive Capacity : GDP per capita, PPP (World Bank)
-
Physical Threat :
-
It's Not Just a Vibe, It's a Correlation — Spearman rank (
$\rho = -0.64$ ) and Pearson log-linear analyses back the pattern with statistics, not just a striking visual.
[ NASA NEX-GDDP-CMIP6 NetCDF ]
│
▼
[ Xarray + Rioxarray Pipeline ] ──► Calculates Stull's Wet-Bulb (Tw)
│
▼
[ 95th Percentile Zonal Reduction ] ──► National Vector Boundary Extraction
│
▼
[ Compressed JSON Asset ] ──► (< 1.2 MB payload)
│
▼
[ Streamlit + Plotly Frontend ] ──► Real-Time Multi-Layer Visual System
Want to inspect the pipeline or tinker with the arena yourself?
1. Clone the repository
git clone https://github.com/your-username/scivis-26.git
cd scivis-262. Install dependencies
pip install -r requirements.txt3. Launch the app
streamlit run app.py- Atmospheric Projections — NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP-CMIP6, CanESM5 model)
- Emissions Data — Global Carbon Project via Our World in Data (per-capita CO₂ emissions)
- Economic Data — World Bank Open Data Catalog (GDP per capita, PPP — indicator
NY.GDP.PCAP.PP.CD)
Aisha Nama · Anjana Sankar Department of Computer Science, Cochin University of Science and Technology (CUSAT)
Built with ❤️, Python, and a stubborn belief that data should make people feel the numbers — for the IEEE VGTC SciVis Contest 2026.