Live Visualization
Project Snapshot
- Core Engine: SCAFET Algorithm
- Input Data: ECMWF IFS Daily Wind Forecasts
- Update Frequency: Daily Execution
- Primary Metric: Atmospheric Vorticity & Geometry
Overview
Storm Tracker applies a simplified implementation of SCAFET (Scalable Feature Extraction and Tracking) to daily wind forecasts, identifying and mapping severe storms globally without manual intervention.
The Methodology: SCAFET
Unlike traditional detection pipelines that rely on dataset-dependent intensity thresholds, SCAFET analyzes the underlying geometric structure of atmospheric phenomena using a mathematical shape index.
| Feature Type | Detection Mechanism |
|---|---|
| Cyclonic Storms | Convex cap-shaped surface geometry & vorticity peaks |
| Atmospheric Rivers | Continuous high-moisture spatial corridors |
| Jet Streams | Linear, high-velocity atmospheric bounds |
For complete mathematical definitions, review the Geoscientific Model Development publication or my Ph.D. dissertation.
Operational Workflow
1. Ingestion & Computation Retrieves daily global wind vectors from the ECMWF Data Store and derives global atmospheric vorticity fields.
2. Geometric Extraction Applies shape-index filters to isolate convex atmospheric structures characteristic of active storm centers.
3. Filtering & Tracking Filters low-amplitude noise to extract, rank, and track high-intensity cyclone trajectories.
4. Automated Publishing Renders updated geo-referenced JSON outputs directly to the web dashboard every 24 hours.
Storm Tracker is an experimental academic prototype developed by Dr. Arjun Babu Nellikkattil. It is not an operational safety tool. Refer to official weather entities (e.g., NOAA, ECMWF, or local meteorological services) for safety advisories.