An ingredient-based severe weather environment analysis and pattern-recognition tool built with NiceGUI and powered by Open-Meteo forecast data.
This application evaluates the environmental support for organized severe convection by analyzing instability, kinematics, moisture, and forcing proxies. Rather than issuing forecasts, it provides diagnostic guidance, highlights likely convective outcomes, and identifies common failure modes that may prevent severe weather despite favorable parameters.
The tool is designed to aid situational awareness and reasoning during severe weather setups, with emphasis on transparency and physical consistency.
This is not an official forecast product. Always consult SPC and NWS guidance for operational decisions.
- 24–48 hour environmental analysis using Open-Meteo forecast data
- Ingredient-based diagnostics, including:
- Instability (maximum CAPE, average Lifted Index)
- Low-level and deep-layer shear (approximate 0–1 km, 0–3 km, and deep-layer proxies)
- Storm-relative rotation potential (proxy, with clear limitations)
- Moisture and precipitation signals
- Strict instability gating to avoid false positives in high-shear / low-CAPE or cold-season regimes
- Environmental support classification (Minimal → Extreme), describing how supportive the atmosphere is for severe convection
- Likely convective outcome guidance, based on ingredient combinations rather than deterministic thresholds
- Failure mode identification, explaining why storms may remain weak or fail to initiate
- Clean, grouped summary designed for rapid situational awareness
This tool is intended as a diagnostic aid, not a forecast generator. It mirrors how human forecasters reason through severe weather setups by:
- Emphasizing ingredients over indices
- Separating environmental support from storm realization
- Explicitly communicating uncertainty and conditionality
Outputs describe what the environment may support, not what storms will do.
Try the live demo here:
https://huggingface.co/spaces/Freshwaffle23426/Weather-Risk-Checker
Forecast data is provided by Open-Meteo, which aggregates multiple numerical weather prediction models. Parameter availability and vertical resolution are limited; all kinematic and composite quantities are therefore approximations and treated accordingly.
- Vertical levels are pressure-based proxies and may not correspond exactly to fixed-height layers
- Storm-relative rotation metrics are simplified and not equivalent to operational SRH calculations
- No convective initiation, forcing strength, or storm-scale evolution is explicitly modeled
- Not suitable for operational warning or decision-making
MIT License