🌊 FloodSense — Terrain-Aware Flood Prediction & Simulation Web Platform

Pune Agri Hackathon | Team: AKARI

Overview

FloodSense is a web platform that:

• Ingests Digital Elevation Model (DEM) data of Maharashtra terrain
• Pulls real-time + forecast rainfall data from weather APIs
• Simulates how water flows and accumulates across terrain
• Predicts flood-risk zones for the next 6 / 12 / 24 / 72 hours
• Displays everything on an interactive map with color-coded risk overlays
• Provides early warning alerts for villages, farmland, and roads in flood zones

Target Region

Maharashtra (focus district: Kolhapur or Raigad — historically flood-prone)

Primary Users

Farmers, local government, disaster management teams

Tech Stack

• Frontend: React (Vite) + Leaflet + Recharts
• Backend: Python FastAPI + RichDEM + XGBoost
• Data: SRTM DEM, Open-Meteo API, India WRIS river data

Getting Started

Backend

cd backend
pip install -r ../requirements.txt
uvicorn main:app --reload

Frontend

cd frontend
npm install
npm run dev

Project Structure

floodsense/
├── frontend/          # React + Vite frontend
├── backend/           # FastAPI backend
├── notebooks/         # Jupyter notebooks for EDA & model training
├── requirements.txt   # Python dependencies
├── package.json       # Node dependencies
└── README.md          # This file

---

AKARI | Pune Agri Hackathon FloodSense v1.0 — Terrain-Aware Flood Prediction Platform