HAB Project - Bioluminescence Forecast System

A full-stack application that provides weekly bioluminescence forecasts for Scripps Pier, La Jolla, using machine learning models and real-time data visualization.

System Architecture

graph TD
    A[Weekly Cloud Scheduler] --> B[ML Model Run]
    B --> C[Save Forecast to Google Drive]
    D[User Visits Website] --> E[Frontend Next.js App]
    E --> F[API Route]
    F --> G[Google Drive API]
    G --> H[Forecast JSON File]
    H --> E
    E --> I[Display Forecast]

Loading

Components

1. Backend ML System

   • Weekly scheduled job triggers forecast generation
   • Python-based ML model using pyEDM library
   • Outputs 7-day forecast in JSON format
   • Saves results to Google Drive

2. Data Storage

   • Google Drive as primary storage
   • JSON file format for forecast data
   • Service Account for secure access
   • Weekly updates maintain data freshness

3. Frontend Application

   • Next.js-based web application
   • Real-time data fetching from Google Drive
   • Responsive design with Tailwind CSS
   • Interactive visualizations
   • Error handling and fallback states

4. API Layer

   • Next.js API routes
   • Google Drive API integration
   • Data validation and transformation
   • Error handling and logging

Setup Guide

1. Backend Setup

   • See /app/README.md for ML model setup
   • Configure Google Cloud Scheduler
   • Set up Google Drive access

2. Frontend Setup

   • See /frontend/README.md for detailed setup
   • Configure environment variables
   • Set up Google Drive API access

3. Deployment

   • Backend: Cloud Platform of choice
   • Frontend: WordPress hosting
   • Configure cross-origin resource sharing (CORS)

Data Flow

1. Weekly Update Process

   Cloud Scheduler
   → Trigger ML Model
   → Generate Forecast
   → Save to Google Drive
   

2. User Request Flow

   User Visit
   → Next.js Frontend
   → API Route
   → Google Drive
   → Display Data
   

Development

Prerequisites

• Node.js 18+
• Python 3.8+
• Google Cloud account
• WordPress hosting

Local Development

# Frontend
cd frontend
npm install
npm run dev

# Backend
cd app
python -m venv venv
source venv/bin/activate  # or `venv\Scripts\activate` on Windows
pip install -r requirements.txt
python forecast.py

License

This project is licensed under the MIT License - see the LICENSE file for details.