🌍 MicroClimate

Real-time Micro-Climate Forecasting & Anomaly Detection for Hyper-Local Regions

MicroClimate is an AI-powered platform designed to monitor, forecast, and analyze hyper-local weather and climate conditions at a 1–5 km spatial resolution — far more granular than traditional city-level forecasts. By combining live atmospheric data, machine learning models, and anomaly detection, MicroClimate provides actionable insights and early risk signals for communities, planners, and researchers.

---

🚀 Motivation

Most weather platforms provide coarse-grained forecasts that fail to capture neighborhood-level variations caused by urban density, terrain, vegetation, and localized weather phenomena.

MicroClimate bridges this gap by enabling:

• Localized climate intelligence
• Early detection of abnormal weather patterns
• Risk-aware decision making for smart cities, agriculture, and disaster preparedness

---

✨ Key Features

🌡️ Real-Time Weather Intelligence

• Live ingestion of atmospheric data:

  • Temperature
  • Humidity
  • Pressure
  • Wind speed & direction

• Powered by trusted public weather APIs

📍 Hyper-Local Geo Support

• High-precision latitude/longitude-based tracking
• Reverse geocoding to identify the user’s neighborhood or locality
• Dynamic location-based weather updates

📊 Time-Series Forecasting

• Short-term forecasting using ML-based time-series models (e.g., Prophet)
• Visualized trend predictions for upcoming hours/days
• Adaptive to changing patterns over time

🚨 Anomaly Detection Engine

• Detects sudden or unusual climate deviations

• Flags potential risks such as:

  • Sudden temperature spikes/drops
  • Abnormal humidity or pressure changes

• Designed for early warning and monitoring systems

⚠️ Risk Scoring System

• Aggregates anomalies and forecast deviations
• Generates a risk score indicating potential threat levels
• Useful for decision-making and alerting pipelines

🌍 Interactive 3D Visualization

• 3D Earth-based visualization using Three.js
• Intuitive and immersive exploration of climate conditions
• Designed for clarity, education, and insight discovery

---

🧩 System Architecture

🖥 Frontend

• React + Vite for fast and modern UI
• Three.js + React Three Fiber for 3D Earth visualization
• Responsive design for smooth user interaction

⚙️ Backend

• FastAPI for high-performance API services

• ML pipelines for:

  • Time-series forecasting
  • Anomaly detection

• Clean RESTful endpoints for integration

🌐 Data Sources

• OpenWeather API – real-time and historical weather data
• Nominatim Geo API – reverse geocoding (location → address)

🗄 Storage

• SQLite (lightweight local storage)
• Can be extended to PostgreSQL/MySQL for production-scale deployments

🚀 Deployment Stack

• Python (FastAPI, ML models)
• Node.js (React frontend)

---

🛠 Installation & Setup

1️⃣ Clone the Repository

git clone https://github.com/MadhuTiwari-345/MicroClimate.git
cd MicroClimate

---

2️⃣ Backend Setup

cd backend
pip install -r requirements.txt

Create a .env file and add:

OPENWEATHER_API_KEY=your_openweather_api_key
NOMINATIM_API_KEY=your_nominatim_api_key   # optional

Run the backend server:

uvicorn backend.main:app --reload

---

3️⃣ Frontend Setup

cd frontend
npm install
npm run dev

The application will be available at:

http://localhost:5173

---

🧠 Usage

1. Open the frontend application

2. Allow location access for hyper-local detection

3. View:

   • Real-time micro-climate conditions
   • Forecast graphs and trends
   • Anomaly alerts and risk scores

4. Use backend APIs for:

   • Dashboards
   • Smart city tools
   • Research integrations

---

📈 Future Enhancements

• 🔌 IoT Sensor Integration

  • Street/block-level data ingestion
  • Higher accuracy micro-climate mapping

• 🧠 Advanced Deep Learning Models

  • LSTM / Transformer-based forecasting
  • Complex climate pattern recognition

• 🌾 Domain-Specific Modules

  • Agriculture advisories
  • Smart city planning tools
  • Energy optimization insights

• 🚨 Disaster & Emergency Alerts

  • Government and municipal integrations
  • Real-time early warning systems

---

🤝 Contributing

Contributions are welcome and appreciated! You can help by:

• Reporting bugs
• Suggesting features
• Improving documentation
• Submitting pull requests

Please open an issue or start a discussion before major changes.

---