Coastal Threat Alert System

A hackathon-oriented idea for detecting potentially dangerous coastal conditions using multiple data sources and machine learning.

Why

Climate risk is local and fast. A system that only looks at one sensor misses the story the others are telling. The goal was to combine signals into an early-warning layer rather than another dashboard nobody checks.

Inputs

• Tide gauges for water-level anomalies.
• Weather stations for wind and pressure shifts.
• Satellite data (NASA / NOAA) for large-scale context.
• Environmental signals aggregated over time.

AI component

The system uses machine-learning concepts to identify abnormal patterns across these inputs and surface them as early warnings, with a Streamlit dashboard for human review.

Prototype stack

Python for data and modeling, Streamlit for the interface, n8n for automation, and Supabase for storage and alerts.